I did my PhD in Atomic, Molecular, and Optical (AMO) physics, and despite "optical" being part of that I realized midway that I didn't know enough about how regular cameras worked!
It didn't take very long to learn, and it turned out to be extremely important in the work I did during the early days at Waymo and later at Motional.
I wanted to pass along this fun video from several years ago that discusses HDR: https://www.youtube.com/watch?v=bkQJdaGGVM8 . It's short and fun, I recommend it to all HN readers.
Separately, if you want a more serious introduction to digital photography, I recommend the lectures by Marc Levoy from his Stanford course: https://www.youtube.com/watch?v=y7HrM-fk_Rc&list=PL8ungNrvUY... . I believe he runs his own group at Adobe now after leading a successful effort at Google making their pixel cameras the best in the industry for a couple of years. (And then everyone more-or-less caught up, just like with most tech improvements in the history of smartphones).
Try capturing fire with a non-Sony phone and a Sony phone. At least Samsung doesn't color correct blackbodies right and the flame looks nothing like reality.
I think Canon makes at least some of their sensors, and Nikon designs theirs and makes it at a third party I forget the name of that isn't Sony or Samsung but they still do use Sony stuff in a lot of their cameras.
I don't know about Pentax, Panasonic or OMD (formerly Olympus)
I think folks here have some idea how expensive chip fabs are. That's why only Canon is able to make their own sensors.
Sony makes sensors for pretty much everyone else. But it's well known that other folks e.g. Nikon have been able to get better signal-to-noise with Sony-made sensors than Sony themselves. I think Panasonic used to make their own sensors but with some recent re-org, that got spun out.
It's been widely rumored that Leica uses Sony sensors, but this gets repeatedly denied by people claiming inside information. We know that Leica was getting 24MP CMOS sensors from CMOSIS in the 2012 timeframe, but CMOSIS has since been acquired by OSRAM, and there hasn't been any verifiable information since then, whether confirming or denying a continued business relationship.
This gets to a gaming rant of mine: Our natural vision can handle these things because our eyes scan sections of the scene with constant adjustment (light-level, focus) while our brain is compositing it together into what feels like a single moment.
However certain effects in games (i.e. "HDR" and Depth of Field) instead reduce the fidelity of the experience. These features limp along only while our gaze is aimed at the exact spot the software expects. If you glance anywhere else around the scene, you instead percieve an unrealistically wrong coloration or blur that frustratingly persists no matter how much you squint. These problems will remain until gaze-tracking support becomes standard.
So ultimately these features reduce the realism of the experience. They make it less like being there and more like you're watching a second-hand movie recorded on flawed video-cameras. This distinction is even clearer if you consider cases where "film grain" is added.
It's crazy that post is 15 years old. Like the OP and this post get at, HDR isn't really a good description of what's happening. HDR often means one or more of at least 3 different things (capture, storage, and presentation). It's just the sticker slapped on advertising.
Things like lens flares, motion blur, film grain, and shallow depth of field are mimicking cameras and not what being there is like--but from a narrative perspective we experience a lot of these things through tv and film. Its visual shorthand. Like Star Wars or Battlestar Galactica copying WWII dogfight footage even though it's less like what it would be like if you were there. High FPS television can feel cheap while 24fps can feel premium and "filmic."
Often those limitations are in place so the experience is consistent for everyone. Games will have you set brightness and contrast--I had friends that would crank everything up to avoid jump scares and to clearly see objects intended to be hidden in shadows. Another reason for consistent presentation is for unfair advantages in multiplayer.
> Things like lens flares, motion blur, film grain, and shallow depth of field are mimicking cameras and not what being there is like
Ignoring film grain, our vision has all these effects all the same.
Look in front of you and only a single plane will be in focus (and only your fovea produces any sort of legibility). Look towards a bright light and you might get flaring from just your eyes. Stare out the side of a car or train when driving at speed and you'll see motion blur, interrupted only by brief clarity if you intentionally try to follow the motion with your eyes.
Without depth of field simulation, the whole scene is just a flat plane with completely unrealistic clarity, and because it's comparatively small, too much of it is smack center on your fovea. The problem is that these are simulations that do not track your eyes, and make the (mostly valid!) assumption that you're looking, nearby or in front of whatever you're controlling.
Maybe motion blur becomes unneccessary given a high enough resolution and refresh rate, but depth of field either requires actual depth or foveal tracking (which only works for one person). Tasteful application of current techniques is probably better.
> High FPS television can feel cheap while 24fps can feel premium and "filmic."
Ugh. I will never understand the obsession this effect. There is no such thing as a "soap opera effect" as people liek to call it, only a slideshow effect.
The history behind this is purely a series of cost-cutting measures entirely unrelated to the user experience or artistic qualities. 24 fps came to be because audio was slapped onto the film, and was the slowest speed where the audio track was acceptable intelligible, saving costly film paper - the sole priority of the time. Before that, we used to record content at variable frame rates but play it back at 30-40 fps.
We're clinging on to a cost-cutting measure that was a significant compromise from the time of hand cranked film recording.
> Look in front of you and only a single plane will be in focus (and only your fovea produces any sort of legibility). Look towards a bright light and you might get flaring from just your eyes. Stare out the side of a car or train when driving at speed and you'll see motion blur, interrupted only by brief clarity if you intentionally try to follow the motion with your eyes.
The problem is the mismatch between what you’re looking at on the screen and what the in-game camera is looking at. If these were synchronised perfectly it wouldn’t be a problem.
> Ugh. I will never understand the obsession this effect.
All of it (lens flares, motion blur, film grain, DoF, tone mapping, and exposure, frame rate) are artistic choices constrained by the equipment we have to collect and present it. I think they'll always follow trends. In my entire career following film, photography, computer graphics, and game dev the only time I've heard anyone talk about how we experience any of those things is when people say humans see roughly equivalent of a 50mm lens (on 35mm film).
Just look at the trend of frame size. Film was roughly 4:3, television copied it. Film started matting/cropping the frame. It got crazy with super wide-screen to where some films used 3 projectors side-by-side and most settled on 16:9. Then television copied it. Widescreen is still seen as more "filmic." I remember being surprised working on a feature that switched to Cinemascope's aspect ratio and seeing that was only 850 pixels tall--a full frame would be about twice that.
To me, high frame rate was always just another style. My only beef was with motion-smoothing muddying up footage shot at different frame rates.
The problem is that it just doesn't work on modern, fast displays. Without motion smoothing on a fast and bright screen, 24fps/30fps goes from "choppy" to "seizure inducing and unwatchable". Older sets would just naturally smooth things out.
Even on my LCD TV, smooth motion like credits at certain speeds are extremely uncomfortable to look at at these frame rates.
I consider it borderline irresponsible to continue using these framerates, forcing users into frame interpolation and horrible artifacts, a decision the manufacturer might even have made for them. 120 Hz is finally becoming the norm for regular content (with monitors going to 500+ nowadays), we should at least be able to get to 60 Hz as the lower bound for regular content delivery.
Going further down for artistic value, e.g. for stop motion or actual slide shows is less of a problem in my opinion. It is not as disturbing, and if regular content was appropriately paced there would be no need for interpolation to mess with it...
> Just look at the trend of frame size.
Frame size is different from the other parameters, as it is solely a physical practicality. Bigger is better in all directions, but a cinema screen needs to fit in the building - making a building much taller is less economical than making it wider, and making it whatever it isn't right now adds novelty.
The content needs to be made for the screen with the appropriate balance of periphery and subject to not be completely wrong, so screen technology and recording technology tends to align. Economy of scale causes standardization on lenses and image circles, and the choice of aspect ratio within that circle on the film, forming a feedback loop that enforces the parameters for almost all content.
If some technology somewhere else in the stack causes a change, some will follow for the novelty but others will simply follow the reducing cost, and soon all content aligns on the format, and the majority of home TV sets will be shaped to fit the majority content it can receive.
> The problem is that it just doesn't work on modern, fast displays.
I'm very confused by this. From what I've seen it's been getting a lot better (since transitioning from CRTs). At least for television, frame-rate matching is becoming more of a thing. Higher frame rates really help. Calling everything fps for simplicity; 120 divides evenly by 24, 30, and 60. Lower values wont match and cause issues.
Similarly, (maybe back in the 90s?) projectors in theaters would double-expose each frame to reduce the flicker in between frames. With digital, they no longer have to advance the film between frames.
> smooth motion like credits at certain speeds are extremely uncomfortable to look at at these frame rates.
I think scrolling credits are the most difficult use case: white on black with hard text and no blur. DLP projectors (common 10+ years ago) drive me nuts displaying R G and B separately.
Outside of credits, cinematographers and other filmmakers do think about these things. I remember hearing a cinematographer talk about working on space documentaries in Imax. If you panned too quickly, the white spaceship over a black star field could jump multiple feet each frame. Sure films shot today are optimized for the theater, but the technology gap between theater and home is nowhere near as crazy as CRT vs acetate.
> Frame size is different from the other parameters, as it is solely a physical practicality.
I'm still struggling to see how its that different. Widescreen meant a lower effective resolution (it didn't have to--it started with Cinerama and Cinemascope), but was adopted for cost and aesthetic reasons.
> If some technology somewhere else in the stack causes a change…and soon all content aligns on the format, and the majority of home TV sets will be shaped to fit the majority content it can receive.
And the industry and audiences are really attached to 24fps. Like you say, home televisions adopted film's aspect ratio and I've also seen them adopt much better support for 24fps.
As kind of an aside, I wonder if the motion blur is what people are attached to more than the actual frame rate. I assume you're talking about frame rates higher than 30? Sure, we have faster films and brighter lights, but exposure time is really short. I saw the Hobbit in theaters in both high frame rate and 24fps and the 24fps one looked weird to me, too--I meant to look it up, but I assume they just dropped frames making the blur odd.
I'm with you on depth of field, but I don't understand why you think HDR reduces the fidelity of a game.
If you have a good display (eg an OLED) then the brights are brighter and simultaneously there is more detail in the blacks. Why do you think that is worse than SDR?
That's not what it means since 2016 or so when consumer TVs got support for properly displaying brighter whites and colors.
It definitely adds detail now, and for the last 8-9 years.
Though consumer TVs obviously still fall short of being as bright at peak as the real world. (We'll probably never want our TV to burn out our vision like the sun, though, but probably hitting highs at least in the 1-2000nit range vs the 500-700 that a lot peak at right now would be nice for most uses.
Right. Just like the article, HDR is too vague to mean anything specific and a label that's slapped onto products. In gaming, it often meant they were finally simulating light and exposure separately--clipping highlights that would have previously been shown. In their opinion, reducing the fidelity. Same with depth of field blurring things that used to not have blur.
It's HDR at the world data level, but SDR at the rendering level. It's simulating the way film cannot handle real-life high dynamic range and clips it instead of compressing it like "HDR" in photography.
> Instead of compressing it like "HDR" in photography
That's not HDR either, that's tone mapping to SDR. The entire point of HDR is that you don't need to compress it because your display can actually make use of the extra bits of information. Most modern phones take true HDR pictures that look great on an HDR display.
The “HDR” here is in the sense of “tone mapping to SDR”. Should also be said that even “H” DR displays only have a stop or two of more range, still much less than in a real-world high-contrast scenes
Displays as low as 400nits have been marketed as "HDR".
But nits are only part of the story. What really matters at the end is the range between the darkest and brightest color the display can show under the lighting conditions you want to use it as. 400 nits in a darkened room where blacks are actually black can have much more actual range than 1000nits with very bright "blacks" due to shitty display tech or excessive external illumination.
Hell yeah, this one of many issues I had with the first Avatar movie. The movie was so filled with cool things to look at but none of it was in focus. 10 minutes in I had had enough and was ready for a more traditional movie experience. Impressive yes, for 10 minutes, then exhausting.
this thread is helping me understand why I always thought 3D movies looked _less_ 3D than 2D movies.
That and after seeing Avatar 1 in 3D, then seeing Avatar 2 in 3D over 10 years later and not really noticing any improvement in the 3D made me declare 3D movies officially dead (though I haven’t done side by side comparisons)
I had a similar complaint with the few 3D things I watched when that has been hyped in the past (e.g., when Avatar came out in cinemas, and when 3D home TVs seemed to briefly become a thing 15 years ago). It felt like Hollywood was giving me the freedom to immerse myself, but then simultaneously trying to constrain that freedom and force me to look at specific things in specific ways. I don't know what the specific solution is, but it struck me that we needed to be adopting lessons from live stage productions more than cinema if you really want people to think what they're seeing is real.
Stereo film has its own limitations. Sadly, shooting for stereo was expensive and often corners were cut just to get it to show up in a theater where they can charge a premium for a stereo screening. Home video was always a nightmare--nobody wants to wear glasses (glassesless stereo TVs had a very narrow viewing angle).
It may not be obvious, but film has a visual language. If you look at early film, it wasn't obvious if you cut to something that the audience would understand what was going on. Panning from one object to another implies a connection. It's built on the visual language of still photography (things like rule of thirds, using contrast or color to direct your eye, etc). All directing your eye.
Stereo film has its own limitations that were still being explored. In a regular film, you would do a rack focus to connect something in the foreground to the background. In stereo, when there's a rack focus people don't follow the camera the same way. In regular film, you could show someone's back in the foreground of a shot and cut them off at the waist. In stereo, that looks weird.
When you're presenting something you're always directing where someone is looking--whether its a play, movie, or stereo show. The tools are just adapted for the medium.
I do think it worked way better for movies like Avatar or How to Train Your Dragon and was less impressive for things like rom coms.
These effects are for the artistic intent of the game. Same goes for movies, and has nothing to do with "second hand movies recorded on flawed cameras". or with "realism" in the sense of how we perceive the world.
This is why I always turn off these settings immediately when I turn on any video game for the first time. I could never put my finger on why I didn’t like it, but the camera analogy is perfect
It seems like a mistake to lump HDR capture, HDR formats and HDR display together, these are very different things. The claim that Ansel Adams used HDR is super likely to cause confusion, and isn’t particularly accurate.
We’ve had HDR formats and HDR capture and edit workflows since long before HDR displays. The big benefit of HDR capture & formats is that your “negative” doesn’t clip super bright colors and doesn’t lose color resolution in super dark color. As a photographer, with HDR you can re-expose the image when you display/print it, where previously that wasn’t possible. Previously when you took a photo, if you over-exposed it or under-exposed it, you were stuck with what you got. Capturing HDR gives the photographer one degree of extra freedom, allowing them to adjust exposure after the fact. Ansel Adams wasn’t using HDR in the same sense we’re talking about, he was just really good at capturing the right exposure for his medium without needing to adjust it later. There is a very valid argument to be made for doing the work up-front to capture what you’re after, but ignoring that for a moment, it is simply not possible to re-expose Adams’ negatives to reveal color detail he didn’t capture. That’s why he’s not using HDR, and why saying he is will only further muddy the water.
Arguably, even considering HDR a distinct thing is itself weird an inaccurate.
All mediums have a range, and they've never all matched. Sometimes we've tried to calibrate things to match, but anyone watching SDR content for the past many years probably didn't do so on a color-calibrated and brightness calibrated screen - that wouldn't allow you to have a brightness slider.
HDR on monitors is about communicating content brightness and monitor capabilities, but then you have the question of whether to clip the highlights or just map the range when the content is mastered for 4000 nits but your monitor manages 1000-1500 and only in a small window.
This! Yes I think you’re absolutely right. The term “HDR” is in part kind of an artifact of how digital image formats evolved, and it kind of only makes sense relative to a time when the most popular image formats and most common displays were not very sophisticated about colors.
That said, there is one important part that is often lost. One of the ideas behind HDR, sometimes, is to capture absolute values in physical units, rather than relative brightness. This is the distinguishing factor that film and paper and TVs don’t have. Some new displays are getting absolute brightness features, but historically most media display relative color values.
Absolute is also a funny size. From the perspective of human visual perception, an absolute brightness only matters if the entire viewing environment is also controlled to the same absolute values. Visual perception is highly contextual, and we are not only seeing the screen.
It's not fun being unable to watch dark scenes during the day or evening in a living room, nor is vaporizing your retinas if the ambient environment went dark in the meantime. People want good viewing experience in the available environment that is logically similar to what the content intended, but that is not always the same as reproducing the exact same photons as the directors's mastering monitor sent towards their their eyeballs at the time of production.
Indeed. For a movie scene depicting the sky including the Sun, you probably wouldn't want your TV to achieve the same brightness as the Sun. You might want your TV to become significantly brighter than the rest of the scenes, to achieve an effect something like the Sun catching your eye.
Of course, the same thing goes for audio in movies. You probably want a gunshot or explosion to sound loud and even be slightly shocking, but you probably don't want it to be as loud as a real gunshot or explosion would be from the depicted distance.
The difference is that for 3+ decades the dynamic range of ubiquitous audio formats (like 16 bit PCM in audio CDs and DVDs) has provided far more dynamic range than is comfortably usable in normal listening environments. So we're very familiar with audio being mastered with a much smaller dynamic range than the medium supports.
This brings up a bunch of good points, and it tracks with what I was trying to say about conflating HDR processing with HDR display. But do keep in mind that even when you have absolute value images, that doesn’t imply anything about how you display them. You can experience large benefits with an HDR workflow, even when your output or display is low dynamic range. Assume that there will be some tone mapping process happening and that the way you map tones depends on the display medium and its capabilities, and on the context and environment of the display. Using the term “HDR” shouldn’t imply any mismatch or disconnect in the viewing environment. It only did so in the article because it wasn’t very careful about its terms and definitions.
The term "HDR" arguably makes more sense for the effect achieved by tone mapping multiple exposures of the same subject onto a "normal" (e.g. SRGB) display. In this case, the "high" in "HDR" just means "from a source with higher dynamic range than the display."
> but your monitor manages 1000-1500 and only in a small window.
Owning a display that can do 1300+ nits sustained across a 100% window has been the biggest display upgrade I think I have ever had. It's given me a tolerance for LCD, a technology I've hated since the death of CRTs and turned me away from OLED.
There was a time I would have said i'd never own a non OLED display again. But a capable HDR display changed that logic in a big way.
Too bad the motion resolution on it, especially compared to OLED is meh. Again, at one point, motion was the most important aspect to me (its why I still own CRTs) but this level of HDR...transformative for lack of a better word.
Motion resolution? Do you mean the pixel response time?
CRTs technically have quite a few artifacts in this area, but as content displayed CRTs tend to be built for CRTs this is less of an issue, and in many case even required. The input is expecting specific distortions and effects from scanlines and phosphor, which a "perfect" display wouldn't exhibit...
The aggressive OLED ABL is simply a thermal issue. It can be mitigated with thermal design in smaller devices, and anything that increases efficiency (be it micro lens arrays, stacked "tandem" panels, quantum dots, alternative emitter technology) will lower the thermal load and increase the max full panel brightness.
(LCD with zone dimming would also be able to pull this trick to get even brighter zones, but because the base brightness is high enough it doesn't bother.)
> Motion resolution? Do you mean the pixel response time?
I indeed meant motion resolution, which pixel response time only partially affects. It’s about how clearly a display shows motion, unlike static resolution which only reflects realistically a still image. Even with fast pixels, sample and hold displays blur motion unless framerate and refresh rate is high, or BFI/strobing is used. This blur immediately lowers perceived resolution the moment anything moves on screen.
> The input is expecting specific distortions and effects from scanlines and phosphor, which a "perfect" display wouldn't exhibit...
That's true for many CRT purists, but is not a huge deal for me personally. My focus is motion performance. If LCD/OLED matched CRT motion at the same refresh rate, I’d drop CRT in a heartbeat, slap on a CRT shader, and call it a day. Heresy to many CRT enthusiasts.
Ironically, this is an area in which I feel we are getting CLOSE enough with the new higher refresh OLEDs for non HDR retro content in combination with: https://blurbusters.com/crt-simulation-in-a-gpu-shader-looks... (which hopefully will continue to be improved.)
> The aggressive OLED ABL is simply a thermal issue.
Theoretically, yes and there’s been progress, but it’s still unsolved in practice. If someone shipped an OLED twice as thick and full of fans and heatsinks, I’d buy it tomorrow. But that’s not what the market wants, so obviously it's not what they make.
> It can be mitigated with thermal design in smaller devices, and anything that increases efficiency (be it micro lens arrays, stacked "tandem" panels, quantum dots, alternative emitter technology) will lower the thermal load and increase the max full panel brightness.
Sure, in theory. But so far the improvements (like QD-OLED or MLA) haven’t gone far enough. I already own panels using these. Beyond that, much of the tech isn’t in the display types I care about, or isn’t ready yet. Which is a pity, because the tandem based displays I have seen in usage are really decent.
That said, the latest G5 WOLEDs are the first I’d call acceptable for HDR at high APL, for the preferences I hold with very decent real scene brightness, at least in film. Sadly, I doubt we’ll see comparable performance in PC monitors until many years down the track and monitors are my preference.
> What is your use case? Retro video games? PC games? Movies?
All of the above! The majority of my interest largely stems from the fact that for whatever reason, I am INCREDIBLY sensitive to sample and hold motion blur. Whilst I tolerate it for modern gaming because I largely have no choice, CRT's mean I do not for my retro gaming, which I very much enjoy. (I was very poor growing up, so most of it for me is not even nostalgia, most of these games are new to me.)
Outside of that, we have a "retro" corner in our home with a 32" trinitron. I collect laserdisc/VHS and we have "retro video" nights where for whatever reason, we watch the worst possible quality copies of movies we could get in significantly higher definition. Much the same as videogames, I was not exposed to a lot of media growing up, my wife has also not seen many things because she was in Russia back then, so there is a ton for us to catch up on very slowly and it just makes for a fun little date night every now and again.
Sadly though, as I get ready to take on a mortgage, it's likely most of my CRT's will be sold, or at least the broadcast monitors. I do not look forward to it haha.
> Outside of that, we have a "retro" corner in our home with a 32" trinitron.
A 32” Trinny. Nice. I have the 32” JVC D-series which I consider my crown jewel. It’s for retro gaming and I have a laserdisc player but a very limited selection of movies. Analog baby.
> Sadly though, as I get ready to take on a mortgage, it's likely most of my CRT's will be sold
Mortgage = space. You won’t believe the nooks and crannies you can fit CRTs into. Attic. Shed. Crawl space. Space under basement stairs. Heck, even the neighbors house. I have no less than 14 CRTs ferreted away in the house. Wife thinks I have only 5. Get creative. Don’t worry about the elements, these puppies were built to survive nuclear blasts. Do I have a sickness? Probably. But analog!!!
Speaking of laser disc it's wild how vivid colors are on that platform. My main example movie is Star Trek First contact and everything is very colorful. DVD is muddy. Even a Blu-ray copy kinda looks like crap. A total side note is the surround sound for that movie is absolutely awesome especially the cube battle scene.
I would love one of these however I have never seen one in my country. Super jealous haha! The tubes they use apparently were an american made tube, with most of the JVCs that were released in my country using different tubes than those released in the US market.
That being said, I do own two JVC "broadcast" monitors that I love. A 17" and a 19". They are no D-series real "TV" but.
Adams adjusted heavily with dodging and burning, even working to invent a new chemical process to provide more control when developing. He was great at determining exposure for his process as well. A key skill was having a vision for what the image would be after adjusting. Adams talked a lot about this as a top priority of his process.
> It's even more incredible that this was done on paper, which has even less dynamic range than computer screens!
I came here to point this out. You have a pretty high dynamic range in the captured medium, and then you can use the tools you have to darken or lighten portions of the photograph when transferring it to paper.
Indeed so. Printing on paper and other substrates is inherently subtractive in nature which limits the gamut of colors and values that can be reproduced. Digital methods make the job of translating additive to subtractive media easier vs. the analog techniques available to film photographers. In any case, the image quality classic photography was able to achieve is truly remarkable.
Notably, the dodging and burning used by photographers aren't obsolete. There's a reason these tools are included in virtually every image-editing program out there. Manipulating dynamic range, particularly in printed images, remains part of the craft of image-making.
> The claim that Ansel Adams used HDR is super likely to cause confusion
That isn't what the article claims. It says:
"Ansel Adams, one of the most revered photographers of the 20th century, was a master at capturing dramatic, high dynamic range scenes."
"Use HDR" (your term) is vague to the point of not meaning much of anything, but the article is clear that Adams was capturing scenes that had a high dynamic range, which is objectively true.
where my interpretation is colored by the experience of making high quality prints and viewing them under different conditions, particularly poor illumination quality but you could also count "small handheld game console", "halftone screened and printed on newsprint" as other degraded conditions. In those cases you might imagine that the eye can only differentiate between 11 tones so even if an image has finer detail it ought to connect well with people if colors were quantized. (I think about concept art from Pokémon Sun and Moon which looked great printed with a thermal printer because it was designed to look great on a cheap screen.)
In my mind, the ideal image would look good quantized to 11 zones but also has interesting detail in texture in 9 of the zones (extreme white and black don't show texture). That's a bit of an oversimplification (maybe a shot outdoors in the snow is going to trend really bright, maybe for artistic reasons you want things to be really dark, ...) but Ansel Adams manually "tone mapped" his images using dodging, burning and similar techniques to make it so.
Correct. I didn’t say that sentence was the source of the conflation, I said it was the source of the Ansel Adams problem. There are other parts that mix together capture, formats, and display.
Edit: and btw I am objecting to calling film capture “HDR”, I don’t think that helps define HDR nor reflects accurately on the history of the term.
That’s a strange claim because the first digital HDR capture devices were film scanners (for example the Cineon equipment used by the motion picture industry in the 1990s).
Film provided a higher dynamic range than digital sensors, and professionals wanted to capture that for image editing.
Sure, it wasn’t terribly deep HDR by today’s standards. Cineon used 10 bits per channel with the white point at coding value 685 (and a log color space). That’s still a lot more range and superwhite latitude than you got with standard 8-bpc YUV video.
They didn’t call that “HDR” at the time, and it wasn’t based on the idea of recording radiance or other absolute physical units.
I’m certain physicists had high range digital cameras before Cineon, and they were working in absolute physical metrics. That would be a stronger example.
You bring up an important point that is completely lost in the HDR discussion: this is about color resolution at least as much as it’s about range, if not moreso. I can use 10 bits for a [0..1] range just as easily as I can use 4 bits to represent quantized values from 0 to 10^9. Talking about the range of a scene captured is leaving out most of the story, and all of the important parts. We’ve had outdoor photography, high quality films, and the ability to control exposure for a long time, and that doesn’t explain what “HDR” is.
And "larger dynamic range" by Rea & Jeffrey (1990):
> With γ = 1 there is equal brightness resolution over the entire unsaturated image at the expense of a larger dynamic range within a given image. Finally, the automatic gain control, AGC, was disabled so that the input/output relation would be constant over the full range of scene luminances.
As I remember it, Paul Debevec had borrowed Greg Ward’s RGBE file format at some point in the late 90s and rebranded it “.hdr” for his image viewer tool (hdrView) and code to convert a stack of LDR exposures into HDR. I can see presentations online from Greg Ward in 2001 that have slides with “HDR” and “HDRI” all over the place. So yeah the term definitely must have started in the late 90s if not earlier. I’m not sure it was as there in the early 90s though.
> Variability is a qualitative setting that indicates how much light levels vary in the zone, i.e. the dynamic range of light landing on surfaces.
> By the nature of the situation being modeled, the user knows whether to expect a high degree of variability in the lighting or a low one.
Given those two phrases, 'a high or low degree of variability in the lighting' translates as 'a high or low degree of dynamic range' — or would be likely to, given human abbreviation tendencies, in successive works and conversations.
It certainly was called HDR when those Cineon files were processed in a linear light workflow. And film was the only capture source available that could provide sufficient dynamic range, so IMO that makes it “HDR”.
But I agree that the term is such a wide umbrella that almost anything qualifies. Fifteen years ago you could do a bit of superwhite glows and tone mapping on 8-bpc and people called that look HDR.
Do you have any links from 1990ish that show use of “HDR”? I am interested in when “HDR” became a phrase people used. I believe I remember hearing it first around 1996 or 97, but it may have started earlier. It was certainly common by 2001. I don’t see that used as a term nor an acronym in the Cineon docs from 1995, but it does talk about log and linear spaces and limiting the dynamic range when converting. The Cineon scanner predates sRGB, and used gamma 1.7. https://dotcsw.com/doc/cineon1.pdf
This 10 bit scanner gave you headroom of like 30% above white. So yeah it qualifies as a type of high dynamic range when compared to 8 bit/channel RGB, but on the other hand, a range of [0 .. 1.3] isn’t exactly in the spirit of what “HDR” stands for. The term implicitly means a lot more than 1.0, not just a little. And again people developing HDR like Greg Ward and Paul Debevec were arguing for absolute units such as luminance, which the Cineon scanner does not do.
HDR is not referring to the scene’s range, and it doesn’t apply to film. It’s referring superficially but specifically to a digital process that improves on 8 bits/channel RGB images. And one of the original intents behind HDR was to capture pixels in absolute physical measurements like radiance, to enable a variety of post-processing workflows that are not available to film.
"High dynamic range" is a phrase that is much older than tone mapping. I see uses of "dynamic range" going back to the 1920s and "high dynamic range" to the 1940s:
You might argue that "HDR" the abbreviation refers to using tone mapping to approximate rendering high dynamic range imagery on lower dynamic range displays. But even then, the sentence in question doesn't use the abbreviation. It is specifically talking about a dynamic range that is high.
Dynamic range is a property of any signal or quantifiable input, including, say sound pressure hitting our ears or photons hitting an eyeball, film, or sensor.
> But even then, the sentence in question doesn’t use the abbreviation
Yes it does. Why are you still looking at a different sentence than the one I quoted??
HDR in this context isn’t referring to just any dynamic range. If it was, then it would be so vague as to be meaningless.
Tone mapping is closely related to HDR and very often used, but is not necessary and does not define HDR. To me it seems like your argument is straw man. Photographers have never broadly used the term “high dynamic range” as a phrase, nor the acronym “HDR” before it showed up in computer apps like hdrView, Photoshop, and iPhone camera.
Oh, sorry, you're right. Mentioning the abbreviation is a red herring. The full quote is:
"But what if we don't need that tradeoff? What if I told you that analog photographers captured HDR as far back as 1857? Ansel Adams, one of the most revered photographers of the 20th century, was a master at capturing dramatic, high dynamic range scenes. It's even more incredible that this was done on paper, which has even less dynamic range than computer screens!"
It seems pretty clear to me that in this context the author is referring to the high dynamic range of the scenes that Adams pointed his camera at. That's why he says "captured HDR" and "high dynamic range scenes".
> It seems pretty clear to me that in this context the author is referring to the high dynamic range of the scenes that Adams pointed his camera at.
Yes, this is the problem I have with the article. “HDR” is not characterized solely by the range of the scene, and never was. It’s a term of art that refers to an increased range (and resolution) on the capture and storage side, and it’s referring to a workflow that involves/enables deferring exposure until display time. The author’s claim here is making the term “HDR” harder to understand, not easier, and it’s leaving out of the most important conceptual aspects. There are some important parallels between film and digital HDR, and there are some important differences. The differences are what make claiming that nineteenth century photographers were capturing HDR problematic and inaccurate.
To further complicate the issue, "high dynamic range" is a phrase that will come up across a few different disciplines, not just related to the capture & reproduction of visual data.
The digital process of tonemapping, aka. 'what Apple calls Smart HDR processing of SDR photos to increase perceptual dynamic range', can be applied to images of any number of channels of any bit depth — though, if you want to tonemap a HyperCard dithered black-and-white image, you'll probably have to decompile the dithering as part of creating the gradient map. Neither RGB nor 8-bit are necessary to make tonemapping a valuable step in image processing.
That’s true, and it’s why tonemapping is distinct from HDR. If you follow the link from @xeonmc’s comment and read the comments, the discussion centers on the conflation of tonemapping and HDR.
That said, the entire reason that tonemapping is a thing, and the primary focus of the tonemapping literature, is to solve the problem of squeezing images with very wide ranges into narrow display ranges like print and non-HDR displays, and to achieve a natural look that mirrors human perception of wide ranges. Tonemapping might be technically independent of HDR, but they did co-evolve, and that’s part of the history.
> It seems like a mistake to lump HDR capture, HDR formats and HDR display together, these are very different things.
These are all related things. When you talk about color, you can be talking about color cameras, color image formats, and color screens, but the concept of color transcends the implementation.
> The claim that Ansel Adams used HDR is super likely to cause confusion, and isn’t particularly accurate.
The post never said Adams used HDR. I very carefully chose the words, "capturing dramatic, high dynamic range scenes."
> Previously when you took a photo, if you over-exposed it or under-exposed it, you were stuck with what you got. Capturing HDR gives the photographer one degree of extra freedom, allowing them to adjust exposure after the fact.
This is just factually wrong. Film negatives have 12-stops of useful dynamic range, while photo paper has 8 stops at best. That gave photographers exposure latitude during the print process.
> Ansel Adams wasn’t using HDR in the same sense we’re talking about, he was just really good at capturing the right exposure for his medium without needing to adjust it later.
There's a photo of Ansel Adams in the article, dodging and burning a print. How would you describe that if not adjusting the exposure?
> Film negatives have 12-stops of useful dynamic range
No, that’s not inherently true. AA used 12 zones, that doesn’t mean every negative stock has 12 stops of latitude. Stocks are different, you need to look at the curves.
But yes most modern negatives are very forgiving. FP4 for example has barely any shoulder at all iirc.
I agree capture, format and display are closely related. But HDR capture and processing specifically developed outside of HDR display devices, and use of HDR displays changes how HDR images are used compared to LDR displays.
> The post never said Adams used HDR. I very carefully chose the words
Hey I’m sorry for criticizing, but I honestly feel like you’re being slightly misleading here. The sentence “What if I told you that analog photographers captured HDR as far back as 1857?” is explicitly claiming that analog photographers use “HDR” capture, and the Ansel Adams sentence that follows appears to be merely a specific example of your claim. The result of the juxtaposition is that the article did in fact claim Adams used HDR, even if you didn’t quite intend to.
I think you’re either misunderstanding me a little, or maybe unaware of some of the context of HDR and its development as a term of art in the computer graphics community. Film’s 12 stops is not really “high” range by HDR standards, and a little exposure latitude isn’t where “HDR” came from. The more important part of HDR was the intent to push toward absolute physical units like luminance. That doesn’t just enable deferred exposure, it enables physical and perceptual processing in ways that aren’t possible with film. It enables calibrated integration with CG simulation that isn’t possible with film. And it enables a much wider rage of exposure push/pull than you can do when going from 12 stops to 8. And of course non-destructive digital deferred exposure at display time is quite different from a print exposure.
Perhaps it’s useful to reflect on the fact that HDR has a counterpart called LDR that’s referring to 8 bits/channel RGB. With analog photography, there is no LDR, thus zero reason to invent the notion of a ‘higher’ range. Higher than what? High relative to what? Analog cameras have exposure control and thus can capture any range you want. There is no ‘high’ range in analog photos, there’s just range. HDR was invented to push against and evolve beyond the de-facto digital practices of the 70s-90s, it is not a statement about what range can be captured by a camera.
> The sentence “What if I told you that analog photographers captured HDR as far back as 1857?” is explicitly claiming that analog photographers use “HDR” capture,
No, it isn't. It's saying they captured HDR scenes.
> The result of the juxtaposition is that the article did in fact claim Adams used HDR
You can't "use" HDR. It's an adjective, not a noun.
> Film’s 12 stops is not really “high” range by HDR standards, and a little exposure latitude isn’t where “HDR” came from.
The Reinhard tone mapper, a benchmark that regularly appears in research papers, specifically cites Ansel Adams as inspiration.
"A classic photographic task is the mapping of the potentially high
dynamic range of real world luminances to the low dynamic range
of the photographic print."
> Perhaps it’s useful to reflect on the fact that HDR has a counterpart called LDR that’s referring to 8 bits/channel RGB.
8-bits per channel does not describe dynamic range. If I attach an HLG transfer function on an 8-bit signal, I have HDR. Furthermore, assuming you actually meant 8-bit sRGB, nobody calls that "LDR." It's SDR.
> Analog cameras have exposure control and thus can capture any range you want.
Sorry man, you seem really defensive, I didn’t mean to put you on edge. Okay, if you are calling the scenes “HDR” then I’m happy to rescind my critique about Ansel Adams and switch instead to pointing out that “HDR” doesn’t refer to the range of the scene, it refers to the range capability of a digital capture process. I think the point ultimately ends up being the same either way. Hey where is HDR defined as an adjective? Last time I checked, “range” could be a noun, I think… no? You must be right, but FWIW, you used HDR as a noun in your 2nd to last point… oh and in the title of your article too.
Hey it’s great Reinhard was inspired by Adams. I have been too, like a lot of photographers. And I’ve used the Reinhard tone mapper in research papers, I’m quite familiar with it and personally know all three authors of that paper. I’ve even written a paper or maybe two on color spaces with one of them. Anyway, the inspiration doesn’t change the fact that 12 stops isn’t particularly high dynamic range. It’s barely more than SDR. Even the earliest HDR formats had like 20 or 30 stops, in part because the point was to use physical luminance instead of a relative [0..1] range.
8 bit RGB does sort-of in practice describe a dynamic range, as long as the 1 bit difference is approximately the ‘just noticeable difference’ or JND as some researchers call it. This happens to line up with 8 bits being about 8 stops, which is what RGB images have been doing for like 50 years, give or take. While it’s perfectly valid arithmetic to use 8 bits values to represent an arbitrary amount like 200 stops or 0.003 stops, it’d be pretty weird.
Plenty of people have called and continue to call 8 bit images “LDR”, here’s just three of the thousands of uses of “LDR” [1][2][3], and LDR predates usage of SDR by like 15 years maybe? LDR predates sRGB too, I did not actually mean 8 bit sRGB. LDR and SDR are close but not quite the same thing, so feel free to read up on LDR. It’s disappointing you ducked the actual point I was making, which is still there even if you replace LDR with SDR.
What is confusing about the sentence about analog cameras and exposure control? I’m happy to explain it since you didn’t get it. I was referring to how the aperture can be adjusted on an analog camera to make a scene with any dynamic range fit into the ~12 stops of range the film has, or the ~8 stops of range of paper or an old TV. I was just trying to clarify why HDR is an attribute of digital images, and not of scenes.
You opened this thread arguing that Ansel Adams didn't "use HDR." I linked you to a seminal research paper which argues that he tone mapped HDR content, and goes on to implement a tone mapper based on his approach. This all seems open and shut.
> I’m happy to rescind my critique about Ansel Adams
Great, I'm done.
> and switch instead to pointing out that “HDR” doesn’t refer to the range of the scene
"Surprisingly, daytime shots with high dynamic range may also suffer from lack of light."
"In low light, or in very high dynamic range scenes"
"For high dynamic range scenes we use local tone mapping"
You keep trying to define "HDR" differently than current literature. Not even current— that paper was published in 2016! Hey, maybe HDR meant something different in the 1990s, or maybe it was just ok to use "HDR" as shorthand for when things were less ambiguous. I honestly don't care, and you're only serving to confuse people.
> the aperture can be adjusted on an analog camera to make a scene with any dynamic range fit into the ~12 stops of range the film has, or the ~8 stops of range of paper or an old TV.
You sound nonsensical because you keep using the wrong terms. Going back to your first sentence that made no sense:
> Analog cameras have exposure control and thus can capture any range you want
You keep saying "range" when, from what I can tell, you mean "luminance." Changing a camera's aperture scales the luminance hitting your film or sensor. It does not alter the dynamic range of the scene.
Analog cameras cannot capture any range. By adjusting camera settings or attaching ND filters, you can change the window of luminance values that will fit within the dynamic range of your camera. To say a camera can "capture any range" is like saying, "I can fit that couch through the door, I just have to saw it in half."
> And I’ve used the Reinhard tone mapper in research papers, I’m quite familiar with it and personally know all three authors of that paper. I’ve even written a paper or maybe two on color spaces with one of them.
I'm sorry if correcting you triggers insecurities, but if you're going to make an appeal to authority, please link to your papers instead of hand waving about the people you know.
Hehe outside is “HDR content”? To me that still comes off as confused about what HDR is. I know you aren’t, but that’s what it sounds like. A sunny day has a high dynamic range for sure, but the acronym HDR is a term of art that implies more than that. Your article even explains why.
Tone mapping doesn’t imply HDR. Tone mapping is always present, even in LDR and SDR workflows. The paper you cited explicitly notes the idea is to “extend” Adams’ zone system to very high dynamic range digital images, more than what Adams was working with, by implication.
So how is a “window of luminance values” different from a dynamic range, exactly? Why did you make the incorrect and obviously silly assumption that I was suggesting a camera’s aperture changes the outdoor scene’s dynamic range rather than what I actually said, that it changes the exposure? Your description of what a camera does is functionally identical. I’m kinda baffled as to why you’re arguing this part that we both understand, using hyperbole.
I hope you have a better day tomorrow. Good luck with your app. This convo aside, I am honestly rooting for you.
> Hehe outside is “HDR content”? To me that still comes off as confused about what HDR is.
"Surprisingly, daytime shots with high dynamic range may also suffer from lack of light."
That's from, "Burst photography for high dynamic range and low-light imaging
on mobile cameras," written by some of the most respected researchers in computational photography. It has 342 citations according to ACM.
First sentence: "Tone mapping is a technique used in image processing and computer graphics to map one set of colors to another to approximate the appearance of high-dynamic-range (HDR) images in a medium that has a more limited dynamic range."
> Why did you make the incorrect and obviously silly assumption that I was suggesting a camera’s aperture changes the outdoor scene’s dynamic range rather than what I actually said, that it changes the exposure?
Because you keep bumbling details like someone with a surface level understanding. Your replies are irrelevant, outdated, or flat out wrong. It all gives me flashbacks to working under engineers-turned-managers who just can't let go, forcing their irrelevant backgrounds into discussions.
It's cool that you studied late 90s 3D rendering. So did I. It doesn't make you an expert in computational photography. Please stop confusing people with your non-sequiturs.
What does the lack of light quote prove? That’s a statement about color resolution, not range, and it uses “high dynamic range” and not “HDR content”. I think you’ve missed my point and are not listening.
Yes tone mapping is used on HDR images. It just doesn’t imply HDR. SDR gamma is tone mapping, for example, which the Wikipedia link you sent explains. Your claim is that Adams use of tone mapping is evidence that he is capturing “HDR content”. The paper you sent doesn’t use that language, it doesn’t ever say Adams was doing tone mapping, it says they develop a tone mapping method inspired by Adams’ zone system that extends the idea into higher dynamic range.
You’re using your own misunderstanding and mis-interpretation of my comments as evidence that they’re wrong. Hey I totally might be wrong about a lot of things, and sure maybe I’m completely non-sensical, but you certainly haven’t convinced me of that. I haven’t had trouble speaking with other people about HDR imaging, people who are HDR experts. All I’m getting out of this so far is that some people react very badly to any hint of critique.
From my perspective, I’m also only hearing bumbling errors, errors like that HDR is an adjective, that LDR doesn’t exist and nobody uses it, that using “range” is incorrect when I say it but not when you do and “window of luminance values” is better, and that Ansel Adams was doing HDR imaging.
Ben, we’re having a bona-fide miscommunication, and I wanted to fix it but I’m failing, and it feels like you’re determined not to fix it or find any common ground. In another environment we’d probably be having a friendly, productive and enlightening conversation. I’m sure there are some things I could learn from you.
But the article even shows Adams dodging/burning a print, which is 'adjusting the exposure' in a localised fashion of the high dynamic range of the film, effectively revealing detail for the LDR of the resulting print that otherwise wouldn't have been visible.
If I look at one of the photography books in my shelf, they are even talking about 18 stops and such for some film material, and how this doesn't translate to paper and all the things that can be done to render it visible in print and how things behave at both extreme ends (towards black and white). Read: Tone-mapping (i.e. trimming down a high DR image to a lower DR output media) is really old.
The good thing about digital is that it can deal with color at decent tonal resolutions (if we assume 16 bits, not the limited 14 bit or even less) and in environments where film has technical limitations.
No, Adams, like everyone who develops their own film (or RAW digital photos) definitely worked in HDR. Film has much more DR than photographic paper, as noted by TFA author (and large digital sensors more than either SDR or HDR displays) especially if you’re such a master of exposure as Adams; trying to preserve the tonalities when developing and printing your photos is the real big issue.
Good question. I think it depends. They are kind of different concepts, but in practice they can overlap considerably. RAW is about using the camera’s full native color resolution, and not having lossy compression. HDR is overloaded, as you can see from the article & comments, but I think HDR capture is conceptually about expressing brightness in physical units like luminance or radiance, and delaying the ‘exposure’ until display time. Both RAW and HDR typically mean using more than 8 bits/channel and capturing high quality images that will withstand more post-processing than ‘exposed’ LDR images can handle.
HDR on displays is actually largely uncomfortable for me. They should reserve the brightest HDR whites for things like the sun itself and caustics, not white walls in indoor photos.
As for tone mapping, I think the examples they show tend way too much towards flat low-local-contrast for my tastes.
HDR is really hard to get right apparently. It seems to get worse in video games too.
I'm a huge fan of Helldivers 2, but playing the game in HDR gives me a headache: the muzzle flash of weapons at high RPMs on a screen that goes to 240hz is basically a continuous flashbang for my eyes.
For a while, No Mans' Sky in HDR mode was basically the color saturation of every planet dialed up to 11.
The only game I've enjoyed at HDR was a port from a console, Returnal. The use of HDR brights was minimalistic and tasteful, often reserved for certain particle effects.
For a year or two after it launched, The Division 2 was a really, really good example of HDR done right. The game had (has?) a day/night cycle, and it had a really good control of the brightness throughout the day. More importantly, it made very good use of the wide color gamut available to it.
I stopped playing that game for several years, and when I went back to it, the color and brightness had been wrecked to all hell. I have heard that it's received wisdom that gamers complain that HDR modes are "too dark", so perhaps that's part of why they ruined their game's renderer.
Some games that I think currently have good HDR:
* Lies of P
* Hunt: Showdown 1896
* Monster Hunter: World (if you increase the game's color saturation a bit from its default settings)
Some games that had decent-to-good HDR the last time I played them, a few years ago:
* Battlefield 1
* Battlefield V
* Battlefield 2042 (If you're looking for a fun game, I do NOT recommend this one. Also, the previous two are probably chock-full of cheaters these days.)
I found Helldivers 2's HDR mode to have blacks that were WAY too bright. In SDR mode, nighttime in forest areas was dark. In HDR mode? It was as if you were standing in the middle of a field during a full moon.
> I have heard that it's received wisdom that gamers complain that HDR modes are "too dark", so perhaps that's part of why they ruined their game's renderer.
A lot of people have cheap panels that claim HDR support (read: can display an HDR signal) but have garbage color space coverage, no local dimming, etc. and to them, HDR ends up looking muted.
For what it's worth, I have a monitor with VA panel that has no local dimming. It has a maximum brightness of a bit shy of 380 nits, and a real contrast ratio of ~3000:1, so HDR stuff ends up looking fine in dim rooms.
I would much rather have this monitor than one the "zoned" backlights that I've seen. They inevitably put nasty, nasty halos around medium-to-high-contrast parts of the picture. IME, any local dimming scheme that's less fine-grained than per-pixel dimming is simply not good enough.
A lot of this is poor QA. When you start to do clever things like HDR you have to test on a bunch of properly calibrated devices of different vendors etc. And if you're targeting Windows you have to accept that HDR is a mess for consumers and even if their display supports, their GPU supports it, they might still have the drivers and color profiles misconfigured. (and many apps are doing it wrong or weird, even when they say they support it)
Also (mostly) on Windows, or on videos for your TV: a lot of cheap displays that say they are HDR are a range of hot garbage.
Most "HDR" monitors are junk that can't display HDR. The HDR formats/signals are designed for brightness levels and viewing conditions that nobody uses.
The end result is a complete chaos. Every piece of the pipeline doing something wrong, and then the software tries to compensate for it by emitting doubly wrong data, without even having reliable information about what it needs to compensate for.
What we really need is some standards that everybody follows. The reason normal displays work so well is that everyone settled on sRGB, and as long as a display gets close to that, say 95% sRGB, everyone except maybe a few graphics designers will have a n equivalent experience.
But HDR, it's a minefield of different display qualities, color spaces, standards. It's no wonder that nobody gets it right and everyone feels confused.
HDR on a display that has peak brightness of 2000 nits will look completely different than a display with 800 nits, and they both get to claim they are HDR.
We should have a standard equivalent to color spaces. Set, say, 2000 nits as 100% of HDR. Then a 2000 nit display gets to claim it's 100% HDR. A 800 nit display gets to claim 40% HDR, etc. A 2500 nit display could even use 125% HDR in it's marketing.
It's still not perfect - some displays (OLED) can only show peak brightness over a portion of the screen. But it would be an improvement.
DisplayHDR standard is supposed to be it, but they've ruined its reputation by allowing HDR400 to exist when HDR1000 should have been the minimum.
Besides, HDR quality is more complex than just max nits, because it depends on viewing conditions and black levels (and everyone cheats with their contrast metrics).
OLEDs can peak at 600 nits and look awesome — in a pitch black room. LCD monitors could boost to 2000 nits and display white on grey.
We have sRGB kinda working for color primaries and gamma, but it's not the real sRGB at 80 nits. It ended up being relative instead of absolute.
A lot of the mess is caused by the need to adapt content mastered for pitch black cinema at 2000 nits to 800-1000 nits in daylight, which needs very careful processing to preserve highlights and saturation, but software can't rely on the display doing it properly, and doing it in software sends false signal and risks display correcting it twice.
There's a pretty good video on YouTube (more than one, actually) that explains how careless use of HDR in modern cinema is destroying the look and feel of cinema we used to like.
Everything is flattened, contrast is eliminated, lights that should be "burned white" for a cinematic feel are brought back to "reasonable" brightness with HDR, really deep blacks are turned into flat greys, etc. The end result is the flat and washed out look of movies like Wicked. It's often correlated to CGI-heavy movies, but in reality it's starting to affect every movie.
The washed out grey thing was an error that became a style!
Because HDR wasn’t natively supported on most displays and software, for a long time it was just “hacked in there” by squashing the larger dynamic range into a smaller one using a mathematical transform, usually a log function. When viewed without the inverse transform this looks horribly grey and unsaturated.
Directors and editors would see this aesthetic day in, day out, with the final color grade applied only after a long review process.
Some of them got used to it and even liking it, and now here we are: horribly washed out movies made to look like that on purpose.
The transition from Avengers to the later movies is very noticeable, and one of the worst offenders since source material really speaks against the choice.
What you said, it's definitely become a style. But, also, a lot of these movies that look like ass on Joe Public's OOGLAMG $130 85" Black Friday TV in his brightly-lit living room actually look awesome if your entire setup is proper, real HDR devices and software, your screen has proper OLED or local dimming, is calibrated to within an inch of its life etc, and you view them in a dark home theater.
True! But I think that movies adapted for TV must be made for the average "good quality" screen, not the state of the art that almost nobody owns. At least they should look decent enough in a good quality (but not top-notch) setup.
Also, the YouTube video I'm thinking of singles out Wicked as seen in movie theaters. The image "as intended" looks washed out and without contrast.
I've found this especially a problem with those AI systems trying to add HDR to existing images/videos. The worst instance I've seen, was playing one of the recent Spongebob platformer games, and having his eyes glow like giant suns in the menu screen. I have a TV capable of a fairly high maximum brightness, and it was dimming the rest of the image just to make sure Spongebob's eyes lit up my living room like it was midday
It feels like to some photographers/cinematographers/game designers, HDR is a gimmick to make something look more splashy/eye catching. The article touches on this a bit, with some of the 2000s HDR examples in photography. With the rise of HDR TVs, it feels like that trend is just happening again.
HDR is when you’re watching a dark film at night, looking at the subtle nuances between shades of dark and black in the shadows on the screen, making out the faint contours the film director carefully curated, and the subtitles gently deposit 40W of light into your optical nerves with “♪”.
As a photographer, I get the appeal of (this new incarnation of) HDR content, but the practical reality is that the photos I see posted in my feeds go from making my display looking normal to having photos searing my retinas, while other content that was uniform white a second prior now looks dull gray.
It's late night here so I was reading this article in dark mode, at a low display brightness - and when I got to the HDR photos I had to turn down my display even more to not strain my eyes, then back up again when I scrolled to the text.
For fullscreen content (games, movies) HDR is alright, but for everyday computing it's a pretty jarring experience as a user.
While it isn't touched on in the post, I think the issue with feeds is that platforms like Instagram have no interest in moderating HDR.
For context: YouTube automatically edits the volume of videos that have an average loudness beyond a certain threshold. I think the solution for HDR is similar penalization based on log luminance or some other reasonable metric.
I don't see this happening on Instagram any time soon, because bad HDR likely makes view counts go up.
As for the HDR photos in the post, well, those are a bit strong to show what HDR can do. That's why the Mark III beta includes a much tamer HDR grade.
> YouTube automatically edits the volume of videos that have an average loudness beyond a certain threshold.
For anyone else who was confused by this, it seems to be a client-side audio compressor feature (not a server-side adjustment) labeled as "Stable Volume". On the web, it's toggleable via the player settings menu.
I didn't realize this was a thing until just now, but I'm glad they added it because (now that I think about it) it's been awhile since I felt the need to adjust my system volume when a video was too quiet even at 100% player volume. It's a nice little enhancement.
Youtube has been long normalizing videos standard feed, switching to a -14 LUFS target in 2019. But LUFS is a global target, and is meant to allow higher peaks and troughs over the whole video, and the normalization does happen on a global level - if you exceed it by 3dB, then the whole video gets it's volume lowered by 3dB, no matter if the part is quiet or not.
The stable volume thing is meant to essentially level out all of the peaks and troughs, and IIRC it's actually computed server-side, I think yt-dlp can download stable volume streams if asked to.
I know they've automatically boosted brightness in dark scenes for a long time too. It's not rare for people to upload a clip from a video game with a very dark scene and it's way brighter after upload than it was when they played or how it looks in the file they uploaded.
Yes, and I love it! Finally, the volume knob control was pulled from producers all about gimmicks to push their productions.
There are still gimmicks, but at least they do not include music so badly clipped as to be unlistenable... hint: go get the DVD or Blu-Ray release of whatever it is and you are likely to enjoy a not clipped album.
It is all about maximizing the overall sonic impact the music is capable of. Now when levels are sane, song elements well differentiated and equalized such that no or only a minor range of frequencies are crushed due to many sounds all competing for them, it will sound, full, great and not tiring!
Thanks audio industry. Many ears appreciate what was done.
Instagram has to allow HDR for the same reason that Firefox spent the past twenty years displaying web colors like HN orange at maximum display gamut rather than at sRGB calibrated: because a brighter red than anyone else’s draws people in, and makes the competition seem lifeless by comparison, especially in a mixed-profiles environment. Eventually that is regarded as ‘garishly bright’, so to speak, and people push back against it. I assume Firefox is already fixing this to support the latest CSS color spec (which defines #rrggbb as sRGB and requires it to be presented as such unless stated otherwise in CSS), but I doubt Instagram is willing to literally dim their feed; instead, I would expect them to begin AI-HDR’ing SDR uploads in order that all videos are captivatingly, garishly, bright.
> I think the solution for HDR is similar penalization based on log luminance or some other reasonable metric.
I completely understand the desire to address the issue of content authors misusing or intentionally abusing HDR with some kind of auto-limiting algorithm similar to the way the radio 'loudness wars' were addressed. Unfortunately, I suspect it will be difficult, if not impossible, to achieve without also negatively impacting some content applying HDR correctly for artistically expressive purposes. Static photos may be solvable without excessive false positive over-correction but cinematic video is much more challenging due to the dynamic nature of the content.
As a cinemaphile, I'm starting to wonder if maybe HDR on mobile devices simply isn't a solvable problem in practice. While I think it's solvable technically and certainly addressable from a standards perspective, the reality of having so many stakeholders in the mobile ecosystem (hardware, OS, app, content distributors, original creators) with diverging priorities makes whatever we do from a base technology and standards perspective unlikely to work in practice for most users. Maybe I'm too pessimistic but as a high-end home theater enthusiast I'm continually dismayed how hard it is to correctly display diverse HDR content from different distribution sources in a less complex ecosystem where the stakeholders are more aligned and the leading standards bodies have been around for many decades (SMPTE et al).
I believe everything could be solved the same way we solved high dynamic range in audio, with a volume control.
I find it pretty weird that all tvs and most monitors hide the brightness adjustment under piles and piles of menus when it could be right there in the remote alongside the sound volume buttons. Maybe phones could have hardware brightness buttons too, at least something as easy as it is on adjusting brightness in notebooks that have dedicated brightness fn buttons.
Such brightness slider could also control the amount of tonemapping applied to HDR content. High brightness would mean no to low tonemapping and low brightness would use a very agressive tonemapper producing a similar image to the SDR content along it.
Also note that good audio volume attenuation requires proper loudness contour compensation (as you lower the volume you also increase the bass and treble) for things to sound reasonably good and the "tone" sound well balanced. So, adjusting the tonemapping based on the brightness isn't that far off what we do with audio.
Another related parallel trend recently is that bad AI images get very high view and like counts, so much so that I've lost a lot of motivation for doing real photography because the platforms cease to show them to anyone, even my own followers.
Why is nobody talking about the standards development? They (OS, image formats) could just say all stuff by default assumes SDR and if a media file explicitly calls for HDR even then it cannot have sharp transitions except in special cases, and the software just blocks or truncates any non conforming images. The OS should have had something like this for sound, about 25-30 years ago. For example a brightness aware OS/monitor combo could just outright disallow anything about x nits. And disallow certain contrast levels, in the majority of content.
Btw, YouTube doesn't moderate HDR either. I saw one video of a child's violin recital that was insanely bright, and probably just by accident of using a bad HDR recorder.
Completely agree. To me, HDR feels like the system is ignoring my screen brightness settings.
I set my screen brightness to a certain level for a reason. Please don’t just arbitrarily turn up the brightness!
There is no good way to disable HDR on photos for iPhone, either. Sure, you can turn off the HDR on photos on your iphone. But then, when you cast to a different display, the TV tries to display the photos in HDR, and it won’t look half as good.
> To me, HDR feels like the system is ignoring my screen brightness settings.
You might be on to something there. Technically, HDR is mostly about profile signaling and therefore about interop. To support it in mpeg dash or hls media you need to make sure certain codec attributes are mentioned in the xml or m3u8 but the actual media payload stays the same.
Any bit or Bob being misconfigured or misinterpreted in the streaming pipeline will result in problems ranging from slightly suboptimal experience to nothing works.
Besides HDR, "spatial audio" formats like Dolby Atmos are notorious for interop isuues
> To me, HDR feels like the system is ignoring my screen brightness settings.
On both Android & iOS/MacOS it's not that HDR is ignoring your screen brightness, but rather the brightness slider is controlling the SDR range and then yes HDR can exceed that, that's the singular purpose of HDR to be honest. All the other purported benefits of HDR are at best just about HDR video profiles and at worst just nonsense bullshit. The only thing HDR actually does is allow for brighter colors vs. SDR. When used selectively this really enhances a scene. But restraint is hard, and most forms of HDR content production are shit. The HDR images that newer iPhones and Pixel phones are capturing are generally quite good because they are actually restrained, but then ironically both of them have horrible HDR video that's just obnoxiously bright.
"On both Android & iOS/MacOS it's not that HDR is ignoring your screen brightness, but rather the brightness slider is controlling the SDR range and then yes HDR can exceed that"
Doesn't this mean HDR is ignoring my brightness setting? Looking at the Mac color profiles, the default HDR has some fixed max brightness regardless of the brightness slider. And it's very bright, 1600 nits vs the SDR max of 600 nits. At least I was able to pick another option capping HDR to 600, but that still allows HDR video to force my screen to its normal full brightness even if I dimmed it.
> Doesn't this mean HDR is ignoring my brightness setting?
Not exactly because it is still being scaled by your brightness setting. As in, start playing an HDR video and then mess with the brightness slider. You will still see the HDR content getting dimmer/brighter.
It's easier to think about in Apple's EDR terms. 0.0-1.0 is the SDR range, and the brightness slider is changing what the nit value is of "1.0" - is it 100 nits? 300 nits? 50 nits? etc... HDR content (in theory) still has that same 0.0-1.0 portion of the range, and it's still being scaled. However it can exceed that 1.0. It's still being scaled, it's still "respecting" that slider. Just the slider wasn't a brightness limit as you're wanting it to be, but a 1.0 alignment point.
The problem comes when HDR content is disrespectful to that. When it just absolutely slams the brightness, pushing all of its content way past that 1.0 value. This is bad content, and unfortunately it's incredibly common in HDR media due in part to the fact that the original HDR specs are very incomplete and in part because it's a new loudness war.
Ah, just tested with https://github.com/dtinth/superwhite and you're correct. I remember it not even scaling with the brightness, but guess I was wrong.
It's not just the HDR content that gets brighter, but SDR content too. When I test it in Chrome on Android, if an HDR image shows up on screen the phone start overriding the brightness slider completely and making everything brighter, including the phone's system UI.
>The only thing HDR actually does is allow for brighter colors vs. SDR.
Not just brighter, but also darker, so it can preserve detail in dark areas rather than crushing them.
> It's not just the HDR content that gets brighter, but SDR content too. When I test it in Chrome on Android, if an HDR image shows up on screen the phone start overriding the brightness slider completely and making everything brighter, including the phone's system UI.
> Not just brighter, but also darker, so it can preserve detail in dark areas rather than crushing them.
It does not get darker, and while PQ allocates more bits to the dark region HLG does not. And, more importantly, neither does the actual display panel which are still typically gamma 2.2-2.4 regardless. So PQ's extra precision in the dark areas is ~never utilized other than as tonemapping input, but the resulting output does not have any increased precision in the darks over SDR.
In fact it actually has less precision in the dark areas as the increased display luminance range means the panels native bit depth need to cover more range.
you are right but at least in my experience it's very easy for a modern iPhone to capture a bad HDR photo, usually because there is some small strong highlight (often a form of specular reflection from a metallic object) that causes everything to be HDR while the photo content wouldn't need it
In beta testing this morning, the Halide “HDR slider” works as intended to solve that. Some of my photos have only needed +0.3 while a couple taken in near-total no-light-pollution darkness have it cranked all the way to max and that still isn’t enough.
In practice the 'HDR' standards are also about wider color gamuts (than sRGB), and (as mentioned in parallel) packed into more bits, in a different way, so as to minimise banding while keeping file sizes in check.
I’m under the impression this is caused by the use of “HDR mode”(s) and poor adaptive brightness implementations on devices. Displays such as the iPad Pro w/ OLED are phenomenal and don’t seem to implement an overactive adaptive brightness. HDR content has more depth without causing brightness distortion.
In contrast, my TV will change brightness modes to display HDR content and disables some of the brightness adjustments when displaying HDR content. It can be very uncomfortably bright in a dark room while being excessively dim in a bright room. It requires adjusting settings to a middle ground resulting in a mixed/mediocre experience overall. My wife’s laptop is the worst of all our devices, while reviews seem to praise the display, it has an overreactive adaptive brightness that cannot be disabled (along with decent G2G response but awful B2W/W2B response that causes ghosting).
Apple’s method involves a good deal of what they call “EDR”, wherein the display gamma is ramped down in concert with ramping the brightness up, so that the brighter areas get brighter while the non-bright areas remain dark due to gamma math; that term is helpful for searching their WWDC developer videos for more details.
It still looks like how they say where its way too bright and makes formerly “bright” whites appear even neutral grey surrounding the hdr content. Personally I find it extremely jarring when it is just a frame that is hdr within the overall screen. It is much nicer when the hdr content is full screen. Imo I wish I could just disable partial screen hdr and keep the full screen implementation because it is that distracting.
Out curiosity — I don’t have Instagram to test and this is a perceptual question anyways — if you enable iOS Settings > Accessibility > Display > Reduce White Point (25%) and increase the screen brightness slightly to compensate for the slight dimness; does that reduce or eliminate the ‘jarring’ness?
That's not inherent to HDR though. BFV (unless I'm confusing it with something else) has a HDR adjustment routine where you push a slider until the HDR white and the SDR white are identical. Same could be done for desktop environments. In my experience, HDR support is very lacking in PCs atm. You can't even play Dolby Vision on Windows, which is the only widely-used HDR format with dynamic metadata.
If you mean https://i.imgur.com/0LtYuDZ.jpeg that is probably the slider GP wants but it's not about matching HDR white to SDR white, it's just about clamping the peak HDR brightness in its own consideration. The white on the left is the HDR brightness according to a certain value in nits set via the bottom slider. The white on the right is the maximally bright HDR signal. The goal of adjusting the slider is to find how bright of an HDR white your display can actually produce, which is the lowest slider value at which the two whites appear identical to a viewer.
Some games also have a separate slider https://i.imgur.com/wenBfZY.png for adjusting "paper white", which is the HDR white one might normally associate with matching to SDR reference white (100 nits when in a dark room according to the SDR TV color standards, higher in other situations or standards). Extra note: the peak brightness slider in this game (Red Dead Redemption 2) is the same knob as the brightness slider in the above Battlefield V screenshot)
No; Windows 10 barely supports it, and of my hundreds of Steam games, exactly none have any code making use of it; seemingly only AAA mega-budget games have the pipeline bandwidth, but e.g. Dune Imperium and Spaceways have no reason to use it and so don’t bother. Windows 11 focused on improving wide color support which is much more critical, as any game shipping for Mac/Win already has dealt with that aspect of the pipeline and has drabified their game for the pre-Win11 ICC nightmares. Even games like Elite Dangerous, which would be a top candidate for both HDR and wide color, don’t handle this yet; their pipeline is years old and I don’t envy them the work of trying to update it, given how they’re overlaying the in-game UI in the midst of their pipeline.
(It doesn’t help that Windows only allows HDR to be defined in EDID and monitor INF files, and that PC monitors start shutting off calibration features when HDR is enabled because their chipsets can’t keep up — just as most modern Sony televisions can’t do both Dolby Vision and VRR because that requires too much processing power for their design budget.)
Nothing you're written disproves that no one wants it, if anything the fact that nothing really supports it implies that it's not a valuable feature that people are clamoring for.
Given how popular it is in modern console AAA gaming, which many people do play on PCs, one can reasonably expect that any studio desiring to target high contrast gaming — a valuable niche for stealth games, where a bright torch should absolutely stand out brightly against the darkest shadows, as well as star exploration and anything with landscapes or specular highlights — would like to be targeting HDR on PC, if only to simplify their cross-platform build and test pipelines, but cannot due to platform barriers that don’t exist for consoles.
I want it. And, I'm hardpressed to imagine a multimedia consumer that doesn't care about color and brightness ranges, unless they don't understand what that is. Every movie is recorded and many phones now capture in a range that can't be encoded in 8-bits.
I experience the same thing you do — but my take on it is different. Being hit with HDR images (and videos on YouTube), while unsettling, makes me then realize how just damned dull the SDR world I had been forced to succumb to has been.
Let the whole experience be HDR and perhaps it won't be jarring.
This seems more like a "your feeds" problem than an HDR problem. Much in the same way people screencap and convert images willy nilly. I suggest blocking non HDR content
> A big problem is that it costs the TV, Film, and Photography industries billions of dollars (and a bajillion hours of work) to upgrade their infrastructure. For context, it took well over a decade for HDTV to reach critical mass.
This is also true for consumers. I don't own a single 4k or HDR display. I probably won't own an HDR display until my TV dies, and I probably won't own a 4k display until I replace my work screen, at which point I'll also replace one of my home screens so I can remote into it without scaling.
Pretty much any display you can buy today will be HDR capable, though that doesn't mean much.
I think the industry is strangling itself putting "DisplayHDR 400" certification on edgelit/backlit LCD displays. In order for HDR to look "good" you either need high resolution full array local dimming backlighting (which still isn't perfect), or a panel type that doesn't use any kind of backlighting like OLED.
Viewing HDR content on these cheap LCDs often looks worse than SDR content. You still get the wider color gamut, but the contrast just isn't there. Local dimming often loses all detail in shadows whenever there is something bright on the screen.
HDR marketing on monitors almost seems like a scam. Monitors will claim HDR compatibility when what they actually means is they will take the HDR data stream and display it exactly the same as SDR content because they don't actually have the contrast and brightness ability of a proper HDR monitor.
Few things are in absolutes. Yes most consumers wont have every screen hdr nor 4k, but most consumers use a modern smartphone and just about every modern smartphone from the past half decade or more has hdr of some level.
I absolutely loathe consuming content on a mobile screen, but its the reality is the vast majority are using phone and tablets most the time.
Funny enough HDR content works absolutely perfect as long as it stays on device that has both HDR-recording and display tech, aka smartphones.
The problem starts with sending HDR content to SDR-only devices, or even just other HDR-standards. Not even talking about printing here.
This step can inherently only be automated so much, because it's also a stylistic decision on what information to keep or emphasize.
This is an editorial process, not something you want to emburden casual users with. What works for some images can't work for others. Even with AI the preference would still need to be aligned.
I have to think you are are the 1-3% outlier though. Everyone I know has an HDR screen even my friend who never buys anything new, but he did run out and buy an HDR tv to replace his old one that he gave to his son.
I honestly do not know if I have any screen that supports HDR. At least I've never noticed any improved image quality when viewing HDR video content and compare the image on my M3 Macbook Pro screen vs. an old external IPS monitor. Maybe my eyes are just broken?
To demonstrate some contrast (heh) with another data point from someone closer to the other extreme, I’ve owned a very HDR-capable monitor (the Apple Pro Display XDR) since 2020, so that’s 5 years now. Content that takes full advantage of it is still rare, but it’s getting better slowly over time.
I have a screen which is "HDR" but what that means is when you turn the feature on it just makes everything more muted, it doesn't actually have any more dynamic range. When you turn HDR on for a game it basically just makes most things more muddy grey.
I also have a screen which has a huge gamut and blows out colors in a really nice way (a bit like the aftereffects of hallucinogens, it has colors other screens just don't) and you don't have to touch any settings.
My OLED TV has HDR and it actually seems like HDR content makes a difference while regular content is still "correct".
The cheap displays adding broken HDR400 support destroyed so much public opinion on HDR. Not actually providing a wider range but accepting the HDR signal would at least have been a minor improvement if the tone mapping weren't completely broken to the point most people just associate HDR with a washed out picture.
>The cheap displays adding broken HDR400 support destroyed so much public opinion on HDR.
It's funny because the display I have that does this was a relatively premium Odyssey G7 which at $700 isn't at all a cheap monitor. (I like it, it's just not at all HDR, or at least not perceivably so compared to Apple devices and an OLED TV)
The only HDR monitors I have actually enjoyed (which, admittedly, is a slightly higher bar than "most of this price aren't completely miscalibrated") have been $1500+ which is a bit ridiculous considering far cheaper TVs do far better with HDR (though tend to suck as monitors). In the Odyssey line this would be the Neo G9 options, but I've never went home with one because they lack flat variants for the ones with good HDR.
Don't feel like you have to. I bought a giant fancy TV with it, and even though it's impressive, it's kinda like ultra-hifi-audio. I don't miss it when I watch the same show on one of my older TVs.
If you ever do get it, I suggest doing for a TV that you watch with your full attention, and watching TV / movies in the dark. It's not very useful on a TV that you might turn on while doing housework; but very useful when you are actively watching TV with your full attention.
I totally love HDR on my OLED TV, and definitely miss it on others.
Like a lot of things, it’s weird how some people are more sensitive to visual changes. For example:
- At this point, I need 120hz displays. I can easily notice when my wife’s phone is in power saver mode at 60hz.
- 4k vs 1080p. This is certainly more subtle, but I definitely miss detail in lower res content.
- High bitrate. This is way more important than 4k vs 1080p or even HDR. But it’s so easy to tell when YouTube lowers the quality setting on me, or when a TV show is streaming at a crappy bitrate.
- HDR is tricky, because it relies completely on the content creator to do a good job producing HDR video. When done well, the image basically sparkles, water looks actually wet, parts of the image basically glow… it looks so good.
I 100% miss this HDR watching equivalent content on other displays. The problem is that a lot of content isn’t produced to take advantage of this very well. The HDR 4k Blu-ray of several Harry Potter movies, for example, has extremely muted colors and dark scenes… so how is the image going to pop? I’m glad we’re seeing more movies rely on bright colors and rich, contrasty color grading. There are so many old film restorations that look excellent in HDR because the original color grade had rich, detailed, contrasty colors.
On top of that, budget HDR implementations, ESPECIALLY in PC monitors, just don’t get very bright. Which means their HDR is basically useless. It’s impossible to replicate the “shiny, wet look” of really good HDR water if the screen can’t get bright enough to make it look shiny. Plus, it needs to be selective about what gets bright, and cheap TVs don’t have a lot of backlighting zones to make that happen very well.
So whereas I can plug in a 4k 120hz monitor and immediately see the benefit in everything I do for normal PC stuff, you can’t get that with HDR unless you have good source material and a decent display.
I don't either see a point of having 4K TV vs 1080p TV. To me is just marketing, I have at my house both a 4K and a 1080p and from a normal viewing distance (that is 3/4 meters) you don't see differences.
Also in my country (Italy) TV transmissions are 1080i at best, a lot are still 570i (PAL resolution). Streaming media can be 4K (if you have enough bandwidth to stream it at that resolution, which I don't have at my house). Sure, if you download pirated movies you find it at 4K, and if you have the bandwidth to afford it... sure.
But even there, sometimes is better a well done 1080p movie than an hyper compressed 4K one, since you see compression artifacts.
To me 1080p, and maybe even 720p, is enough for TV vision. Well, sometimes I miss the CRT TVs, they where low resolution but for example had a much better picture quality than most modern 4K LCD TV where black scenes are gray (I know there is OLED, but is too expensive and has other issues).
I feel the same way. To be honest even the laptop retina screen is excess. I sometimes go back to a 2012 non retina macbook pro and to be honest at normal laptop viewing angles, you can’t really discern pixels. Biggest difference is display scaling but I have my retina scaled at what the old display would be anyhow because otherwise its too small.
Kind of crazy no one thought of this aspect and we just march on to higher resolution and the required hardware for that.
For TVs under ~80" I feel like you'd have to be sitting abnormally close to your TV for it to matter much. At the same time I think the cost difference between producing 1080p and 4k panels is so low it probably doesn't matter. Like you say, things like the backlight technology (or lack thereof) make a much bigger difference in perceived quality but that's also where the actual cost comes in.
I agree about 4k vs non-4k. I will say going OLED was a huge upgrade, even for SDR content. HDR content is hit-or-miss...I find some of it is tastefully done but in many cases is overdone.
My own movie collection is mostly 2-4GB SDR 1080p files and looks wonderful.
Jokes aside, when a 4k TV has a good upscaler, it's hard to tell the difference between 1080 and 4k. Not impossible; I certainly can, but 1080 isn't distracting.
Well, if it was for me I wouldn't even have a TV at home since I wouldn't use it (and thus I would also save the TV tax that we have to pay in our country if we own one). But my girlfriend watches a lot of TV, and I often watch it with her of course, thus I have two TVs.
This is also true for consumers. I don't own a single 4k or HDR display. I probably won't own an HDR display until my TV dies, and I probably won't own a 4k display until I replace my work screen, at which point I'll also replace one of my home screens so I can remote into it without scaling.
People in the HN echo chamber over-estimate hardware adoption rates. For example, there are millions of people who went straight from CDs to streaming, without hitting the iPod era.
A few years ago on HN, there was someone who couldn't wrap their brain around the notion that even though VCRs were invented in the early 1960's that in 1980, not everyone owned one, or if they did, they only had one for the whole family.
Normal people aren't magpies who trash their kit every time something shiny comes along.
> A few years ago on HN, there was someone who couldn't wrap their brain around the notion that even though VCRs were invented in the early 1960's that in 1980, not everyone owned one, or if they did, they only had one for the whole family.
Point of clarification: While the technology behind the VCR was invented in the '50s and matured in the '60s, consumer-grade video tape systems weren't really a thing until Betamax and VHS arrived in 1975 and 1976 respectively.
Early VCRs were also incredibly expensive, with prices ranging from $3,500 to almost $10,000 after adjusting for inflation. Just buying into the VHS ecosystem at the entry level was a similar investment to buying an Apple Vision Pro today.
Exactly my point. But people on HN, especially the person I referenced, don't understand that we didn't just throw stuff away and go into debt to buy the latest gadgets because we were told to.
>there are millions of people who went straight from CDs to streaming, without hitting the iPod era
Who?
There was about a decade there where everyone who had the slightest interest in music had an mp3 player of some kind, at least in the 15-30 age bracket.
I imagine this depends a LOT on your specific age and what you were doing in the 00's when MP3 player usage peaked.
I finished high school in 2001 and didn't immediately go to college, so I just didn't have a need for a personal music player anymore. I was nearly always at home or at work, and I drove a car that actually had an MP3 CD player. I felt no need to get an iPod.
In 2009, I started going to college, but then also got my first smartphone, the Motorola Droid, which acted as my portable MP3 player for when I was studying in the library or taking mass transit.
If you were going to school or taking mass transit in the middle of the '00s, then you were probably more likely to have a dedicated MP3 player.
I don't know if I count, but I never owned a dedicated MP3 player[1], I listened to MP3s on my computer, but used CDs and cassettes while on the move, until I got an android phone that had enough storage to put my music collection on.
1: Well my car would play MP3s burned to CDs in its CD player; not sure if that counts.
> AI cannot read your mind, so it cannot honor your intent.
This. I can always tell when someone "gets" software development when they either understand (or don't) that computers can't read minds or infer intent like a person can.
Just one other thing. In Analog you also have compensating developers, which will exhaust faster in darker areas (or lighter if you think in negative), and allow for lighter areas more time to develop and show, and hence some more control of the range. Same but to less degree with stand development which uses very low dilutions of the developer, and no agitation. So dodging and burning is not the only way to achieve higher dynamic range in analog photos.
About HDR on phones, I think they are the blight of photography. No more shadows and highlights. I find they are good at capturing family moments, but not as a creative tool.
For analog photos the negative also has more dynamic range than your screen or photopaper without any of that. Contrast is applied in the darkroom by choice of photopaper and enlarger timing and light levels, or after scanning with contrast adjustment applied in post processing. It is really a storage medium of information more than how the final image ought to look.
Slide film has probably a third the dynamic range of negative film and is meant as the final output fit for projection to display.
> About HDR on phones, I think they are the blight of photography. No more shadows and highlights.
HDR is what enables you to capture both the darkest shadow detail and the brightest highlight detail.
With SDR, one or both are often simply just lost. It might come down to preference — if you're an "auto" shooter and like the effect of the visual information at the edges of the available dynamic range being truncated, SDR is for you.
Some people prefer to capture that detail and have the ability to decide whether and how to diminish or remove it, with commensurately more control over the artistic impact. For those folks, HDR is highly desirable.
So, HN, are HDR monitors worth it? I remember ~10 years ago delaying my monitor purchase for the HDR one that was right around the corner, but never (in my purchasing scope) became available. Time for another look?
The utility of HDR (as described in the article) is without question. It's amazing looking at an outdoors (or indoors with windows) scene with your Mk-1 eyeballs, then taking a photo and looking at it on a phone or PC screen. The pic fails to capture what your eyes see for lighting range.
HDR general desktop usage: No. In fact you'll probably actively dislike it to the point of just turning it off entirely. The ecosystem just isn't ready for this yet, although with things like the "constrained-high" concepts ( https://www.w3.org/TR/css-color-hdr-1/#the-dynamic-range-lim... ) this might, and hopefully does, change & improve to a more pleasing result
Also this is assuming an HDR monitor that's also a good match for your ambient environment. The big thing nobody really talks about wiith HDR is that it's really dominated by how dark you're able to get your surrounding environment such that you can push your display "brightness" (read: SDR whitepoint) lower and lower. OLED HDR monitors, for example, look fantastic in SDR and fantastic in HDR in a dark room, but if you have typical office lighting and so you want an SDR whitepoint of around 200-300 nits? Yeah, they basically don't do HDR at all anymore at that point.
The difference is absolutely stunning in some games.
In MS Flight Simulator 2024, going from SDR to HDR goes from looking like the computer game it is to looking life-like. Deeper shadows with brighter highlights makes the scene pop in ways that SDR just can't do.
EDIT: You'll almost certainly need an OLED monitor to really appreciate it, though. Local dimming isn't good enough.
I use HDR for general usage, Windows ruins non-HDR content when HDR is enabled due to their choice of sRGB tf. Luckily every Linux DE has chosen to use the gamma 2.2 tf, and looks fine for general usage.
I use a mini-led monitor, and its quite decent, except for starfields, & makes it very usable even in bright conditions, and HDR video still is better in bright conditions than the equivalent SDR video.
Windows HDR implementation is janky as hell. For months after I got my monitor I couldn't take screenshots because they'd all appear completely blown out, like you cranked the brightness to 300%.
Eventually I did some digging and found there's a setting in Snipping Tool that just... makes screenshots work on HDR displays.
It also seems to add another layer of Your Desktop Trying To Sort Its Shit Out when launching a game that's full screen. Sometimes it's fine, but some games like Balatro will appear fine at first, but then when you quit back to the desktop everything is washed out. Sleeping my PC and waking it back up seems to resolve this.
I recently played through Armored Core VI, and it supports HDR, but whenever I adjust my volume the screen becomes washed out to display the volume slider. Screenshots and recordings also appear washed out in the resulting file.
Agree. Wide gamut and HDR is janky as hell on Windows. I have multi-mon with one SDR and one HDR and that plays havoc with things. But even Microsoft apps aren't updated. I'm pretty certain even Explorer still doesn't support HDR or wide gamut for thumbnails so everything looks either under or oversaturated in the previews. If you open stuff in the default Photos app there is a "flash of content" where it displays it in the wrong profile before it maps it to the correct one, too.
I always thought the "Your Desktop Trying To Sort Its Shit Out" part was a necessary evil, but other platforms don't suffer from this (at least from what I can tell); the state of HDR on Windows is very disappointing, even just adjusting the TF to gamma 2.2 would make it substantially better. Watching all your non hdr content's blacks become gray is terrible. I assume the washed out appearance comes from it giving up on doing SDR->HDR for the desktop.
My brother got an OLED monitor & was telling me how bad his experience was on Windows, & he recently switched to Linux & does not have the issues he was complaining about before. Ofc, downsides to hdr on Linux (no hdr on chromium, hdr on Firefox is unfinished) atm, but the foundation seems better set for it.
Every few years I turn on that HDR toggle for my desktop PC, and it never lasts longer than a day or two.
Top tip: If you have HDR turned on for your display in Windows (at least, MacOS not tested) and then share your screen in Teams, your display will look weirdly dimmed for everyone not using HDR on their display—which is everyone.
A lot of monitors that advertise HDR support really shouldn't. Many of them can decode the signal but don't have the hardware to accurately reproduce it, so you just end up with a washed out muddy looking mess where you're better off disabling HDR entirely.
As others here have said, OLED monitors are generally excellent at reproducing a HDR signal, especially in a darker space. But they're terrible for productivity work because they'll get burned in for images that don't change a lot. They're fantastic for movies and gaming, though.
There are a few good non-OLED HDR monitors, but not many. I have an AOC Q27G3XMN; its a 27" 1440p 180hz monitor that is good for entry-level HDR, especially in brighter rooms. It has over 1000 nits of brightness, and no major flaws. It only has 336 backlight zones, though, so you might notice some blooming around subtitles or other fine details where there's dark and light content close together. (VA panels are better than IPS at suppressing that, though.) It's also around half the price of a comparable OLED.
Most of the other non-OLED monitors with good HDR support have some other deal-breaking flaws or at least major annoyances, like latency, screwing up SDR content, buggy controls, etc. The Monitors Unboxed channel on YouTube and rtngs.com are both good places to check.
I think an OLED is basically an absolute necessity for HDR content.
My current monitor is an OLED and HDR in games looks absolutely amazing. My previous was an IPS that supported HDR, but turning it on caused the backlight to crank to the max, destroying black levels and basically defeating the entire purpose of HDR. Local dimming only goes so far.
Modern mini-led monitors are very good. The “local” dimming is so local that there isn’t much light bleed even in the worst-case situations (cursor over black background makes it particularly apparent).
The advantage of LEDs is they’re brighter. For example, compare two modern Asus ProArt displays: their mini-LED (PA32UCXR) at 1600 nits and their OLED (PA32DC) at 300ish nits. The OLED is 20% more expensive. These two monitors have otherwise comparable specs. Brightness matters a lot for HDR because if you’re in a bright room, the monitor’s peak brightness needs to overpower the room.
Plus for color managed work, I think LED monitors are supposed to retain their calibration well. OLEDs have to be frequently recalibrated.
And so-called micro-LEDs are coming soon, which promise to make “local” so small that it’s imperceptible. I think the near-term future of displays is really good LEDs.
> My previous was an IPS that supported HDR, but turning it on caused the backlight to crank to the max, destroying black levels and basically defeating the entire purpose of HDR
Yeah, that's kind of what I meant when I said that most monitors that advertise HDR shouldn't.
The AOC monitor is the third or fourth one I've owned that advertised HDR, but the first one that doesn't look like garbage when it's enabled.
I haven't gone oled yet because of both the cost and the risk of burn-in in my use case (lots of coding and other productivity work, occasional gaming).
Avoid cranking up the OLED brightness over 70% for static content and absolutely never drive SDR-Reds into the HDR range by using fake HDR-modes when having brightness high.
I have a LG 2018 OLED that has some burnt in Minecraft hearts because of that, not from Minecraft itself, but just a few hours of minecraft Youtube video in those settings from the built in youtube client, but virtually no other detectable issues after excessive years of use with static content.
You only see them with fairly uniform colors as a background where color banding would usually be my bigger complaint.
So burn-ins definitely happen, but they are far from being a deal breaker over the obvious benefits you get vs other types of displays.
And driving everything possible in dark mode (white text on dark bg) on those displays is even the logical thing to do. Then you dont need much max brightness anyway and even save some energy.
> A lot of monitors that advertise HDR support really shouldn't. Many of them can decode the signal but don't have the hardware to accurately reproduce it,
You can lie as much as you like as an advertiser. At the beginning, HD meant 1080p. Then, 1080p became FullHD, only that 720p to became HD and to make HD digital television possible. (the original SD television is worse than analog)
Yes but with asterisks; Best way I can describe it:
You know the 0-10 brightness slider you have to pick at the start of a game? Imagine setting it to 0 and still being able to spot the faint dark spot. The dynamic range of things you can see is so much expanded.
Early HDR screens were very limited (limited dimming zones, buggy implementation) but if you get one post 2024 (esp the oled ones) they are quite decent.
However it needs to be supported at many layers: not just the monitor, but also the operating system, and the content. There are not many games with proper HDR implementation; and even if there is, it may be bad and look worse — the OS can hijack the rendering pipeline and provide HDR map for you (Nvidia RTX HDR) which is a gamble: it may look bleh, but sometimes also better than the native HDR implementation the game has).
But when everything works properly, wow it looks amazing.
> You know the 0-10 brightness slider you have to pick at the start of a game? Imagine setting it to 0 and still being able to spot the faint dark spot. The dynamic range of things you can see is so much expanded.
Note that HDR only actually changes how bright things can get. There's zero difference in the dark regions. This is made confusing because HDR video marketing often claims it does, but it doesn't actually. HDR monitors do not, in general, have any advantage over SDR monitors in terms of the darks. Local dimming zones improve dark contrast. OLED improves dark contrast. Dynamic contrast improves dark contrast. But HDR doesn't.
My understanding is that on the darker scenes (say, 0 to 5 in the brightness slider example), there is difference in luminance value with HDR but not SDR, so there is increased contrast and detail.
This matches my experience; 0 to 5 look identically black if I turn off HDR
You may have a monitor that only enables local dimming zones when fed an HDR signal but not when fed an SDR one, but that would be unusual and certainly not required. And likely something you could change in your monitors controls. On things like an OLED, though, there's no difference in the darks. You'd see a difference between 8bit and 10bit potentially depending on what "0 to 5" means, but 10-bit SDR is absolutely a thing (it predates HDR even)
But like if you can't see a difference between 0 to 5 in a test pattern like this https://images.app.goo.gl/WY3FhCB1okaRANc28 in SDR but you can in HDR then that just means your SDR factory calibration is bad, or you've fiddled with settings that broke it.
> You know the 0-10 brightness slider you have to pick at the start of a game? Imagine setting it to 0 and still being able to spot the faint dark spot
0-10 is so old century. Now it should be 0-2 (0 - default, 1 dark, 2 light,) /s
Honestly, I find the extended brightness FAR less important than the extended color gamut. I have a ~300 nit VA monitor that I'm generally quite happy with and that looks fantastic with well-built HDR renderers.
Given that monitors report information about their HDR minimum and maximum panel brightness capabilities to the machine they are connected to, any competently-built HDR renderer (whether that be for games or movies or whatever) will be able to take that information and adjust the picture appropriately.
I've been thinking of moving out of the Apple ecosystem but after seeing Severance on my iPhone Pro screen I feel like I want the keep the option to have the same HDR experience for movies specifically. With HDR support landing in Linux just a month ago I'm inclined to spend on a good monitor. However, I have IPS HDR 600 monitor but I never felt that the screen was as glorious as the iPhone screen.
I'd also be interested in hearing whether it makes sense to look into OLED HDR 400 screens (Samsung, LG) or is it really necessary to get an Asus ProArt which can push the same 1000 nits average as the Apple XDR display (which, mind you, is IPS).
I think it depends on the screen and also what you use it for. My OLED is unusable for normal work in HDR because it's designed around only a small portion of the screen being at max brightness - reasonable for a game or movie, but the result is that a small window with white background will look really bright, but if I maximize it, it'll look washed out, grey not white.
Also the maximum brightness isn't even that bright at 800 nits, so no HDR content really looks that different. I think newer OLEDs are brighter though. I'm still happy with the screen in general, even in SDR the OLED really shines. But it made me aware not all HDR screens are equal.
Also, in my very short experiment using HDR for daily work I ran into several problems, the most serious of which was the discovery that you can no longer just screenshot something and expect it to look the same on someone else's computer.
> ...the most serious of which was the discovery that you can no longer just screenshot something and expect it to look the same on someone else's computer.
To be pedantic, this has always been the case... Who the hell knows what bonkers "color enhancement" your recipient has going on on their end?
But (more seriously) it's very, very stupid that most systems out there will ignore color profile data embedded in pictures (and many video players ignore the same in videos [0]). It's quite possible to tone-map HDR stuff so it looks reasonable on SDR displays, but color management is like accessibility in that nearly noone who's in charge of paying for software development appears to give any shits about it.
[0] A notable exception to this is MPV. I can't recommend this video player highly enough.
HDR when it works properly is nice, but nearly all HDR LCD monitors are so bad, they're basically a scam.
The high-end LCD monitors (with full-array local dimming) barely make any difference, while you'll get a lot of downsides from bad HDR software implementations that struggle to get the correct brightness/gamma and saturation.
IMHO HDR is only worth viewing on OLED screens, and requires a dimly lit environment. Otherwise either the hardware is not capable enough, or the content is mastered for wrong brightness levels, and the software trying to fix that makes it look even worse.
Apple's Displays yes. But I got a Phillips 4k OLED recently, and I'm already regretting that decision. I need to turn it off every 4 hours to refresh the pixels. Sometimes an entire line of pixels is brighter than the rest. I wiped it with a cloth while pixel refresh was running, and then saw burned in streaks in the direction of the wipe.
And thats now that all the LEDs are still fresh. I can't imagine how bad it will be in a few years.
Also, a lot of Software doesn't expect the subpixel arrangement, so text will often look terrible.
I've had an OLED TV since 2017 and the answer is a resounding yes... if you get an OLED and use it for movies or full screen gaming. Anything else is basically pointless.
For desktop work, don't bother unless your work involves HDR content.
Dunno if it's just my screen or setup but on windows I have a Dell U4025QW and HDR on the desktop just looks strange, overly dull. Looks good in games but I have to manually turn it on and off each time on the screen.
On my Macbook Pro only activates when it needs to but honestly I've only seen one video [1] that impressed me with it, the rest was completely meh. Not sure if its because it's mostly iPhone photography you see in HDR which is overall pretty meh looking anyway.
[1] https://www.youtube.com/watch?v=UwCFY6pmaYY I understand this isn't a true HDR process but someone messing with it in post, but it's the only video I've seen that noticeably shows you colors you can't see on a screen otherwise.
For the love of gods, read the reviews though. Most HDR displays will make the picture worse not better because they have only implemented enough to put a sticker on the box.
You have to spend really good money to get a display which does HDR properly.
Note for Firefox users - view the page in Chrome to see more of what they are talking about. I was very confused by some of the images, and it was a world of difference when I tried again in Chrome. Things began to make a lot more sense - is there a flag I am missing in Firefox on the Mac?
HDR support in Chrome (Android) looks still broken for me. For one, some of the images on the blog have a posterization effect, which is clearly wrong.
Second, the HDR effect seems to be implemented in a very crude way, which causes the whole Android UI (including the Android status bar at the top) to become brighter when HDR content is on screen. That's clearly not right. Though, of course, this might also be some issue of Android rather than Chrome, or perhaps of the Qualcomm graphics driver for my Adreno GPU, etc.
For what it's worth, your comment has me convinced I just "can't see" HDR properly because I have the same page side-by-side on Firefox and Chrome on my M4 MBP and honestly? Can't see the difference.
edit: Ah, nevermind. It seems Firefox is doing some sort of post-processing (maybe bad tonemapping?) on-the-fly as the pictures start out similar but degrade to washed out after some time. In particular, the "OVERTHROW BOXING CLUB" photo makes this quite apparent.
That's a damn shame Firefox. C'mon, HDR support feels like table stakes at this point.
edit2: Apparently it's not table stakes.
> Browser support is halfway there. Google beat Apple to the punch with their own version of Adaptive HDR they call Ultra HDR, which Chrome 14 now supports. Safari has added HDR support into its developer preview, then it disabled it, due to bugs within iOS.
at which point I would just say to `lux.camera` authors - why not put a big fat warning at the top for users with a Firefox or Safari (stable) browser? With all the emphasis on supposedly simplifying a difficult standard, the article has fallen for one of its most famous pitfalls.
"It's not you. HDR confuses tons of people."
Yep, and you've made it even worse for a huge chunk of people. :shrug: Great article n' all just saying.
Does anyone else find the hubris in the first paragraph writing as off-putting as I do?
"we finally explain what HDR actually means"
Then spends 2/3rds of the article on a tone mapping expedition, only to not address the elephant in the room, that is the almost complete absence of predictable color management in consumer-grade digital environments.
UIs are hardly ever tested in HDR: I don't want my subtitles to burn out my eyes in actual HDR display.
It is here, where you, the consumer, are as vulnerable to light in a proper dark environment for movie watching, as when raising the window curtains on a bright summer morning.
(That brightness abuse by content is actually discussed here)
Dolby Vision and Apple have the lead here as a closed platforms, on the web it's simply not predictably possible yet.
Best hope is the efforts of the Color on the Web Community Group from my impression.
> Does anyone else find the hubris in the first paragraph writing as off-putting as I do?
> "we finally explain what HDR actually means"
No. Because it's written for the many casual photographers we've spoken with who are confused and asked for an explainer.
> Then spends 2/3rds of the article on a tone mapping expedition, only to not address the elephant in the room, that is the almost complete absence of predictable color management in consumer-grade digital environments.
That's because this post is about HDR and not color management, which is different topic.
It's about HDR from the perspective of still photography, in your app, on iOS, in the context of hand-held mobile devices. The post's title ("What Is HDR, Anyway?"), content level and focus would be appropriate in the context of your company's social media feeds for users of your app - which is probably the audience and context it was written for. However in the much broader context of HN, a highly technical community whose interests in imaging are diverse, the article's content level and narrow focus aren't consistent with the headline title. It seems written at a level appropriate for novice users.
If this post was titled "How does Halide handle HDR, anyway?" or even "How should iOS photo apps handle HDR, anyway?" I'd have no objection about the title's promise not matching the content for the HN audience. When I saw the post's headline I thought "Cool! We really need a good technical deep dive into the mess that is HDR - including tech, specs, standards, formats, content acquisition, distribution and display across content types including stills, video clips and cinematic story-telling and diverse viewing contexts from phones to TVs to cinemas to VR." When I started reading and the article only used photos to illustrate concepts best conveyed with color gradient graphs PLUS photos, I started to feel duped by the title.
(Note: I don't use iOS or your app but the photo comparison of the elderly man near the end of the article confused me. From my perspective (video, cinematography and color grading), the "before" photo looks like a raw capture with flat LUT (or no LUT) applied. Yet the text seemed to imply Halide's feature was 'fixing' some problem with the image. Perhaps I'm misunderstanding since I don't know the tool(s) or workflow but I don't see anything wrong with the original image. It's what you want in a flat capture for later grading.)
> It's about HDR from the perspective of still photography, in your app, on iOS, in the context of hand-held mobile devices.
It's from the perspective of still photography, video, film, desktop computing, decades of research papers, and hundreds of years of analog photography, condensed into something approachable.
> However in the much broader context of HN, a highly technical community whose interests in imaging are diverse, the article's content level and narrow focus aren't consistent with the headline title. It seems written at a level appropriate for novice users.
"On-Topic: Anything that good hackers would find interesting. That includes more than hacking and startups. If you had to reduce it to a sentence, the answer might be: anything that gratifies one's intellectual curiosity"
To be clear, I didn't submit the post, and I never submit my posts. I don't care if my posts make a splash here, and kind of dread when they do because anything involving photography or video attracts the most annoying "well actually" guys on the Internet.
> When I saw the post's headline I thought "Cool! We really need a good technical deep dive into the mess that is HDR - including tech, specs, standards, formats, content acquisition, distribution and display across content types including stills, video clips and cinematic story-telling and diverse viewing contexts from phones to TVs to cinemas to VR."
The post is called, "What is HDR," and the introduction explains the intended audience. That audience is much larger than "people who want to read about ITU-R Recommendation BT.2100." But if you think people are interested in a post like that, by all means write it.
> However in the much broader context of HN, a highly technical community whose interests in imaging are diverse, the article's content level and narrow focus aren't consistent with the headline title. It seems written at a level appropriate for novice users.
That is hardly the fault of the authors though. The article seems entirely appropriate for its intended audience, and they can’t control who posts it on a site like HN.
Maybe my response was part of the broader HDR symptom—that the acronym is overloaded with different meanings depending on where you're coming from.
On the HN frontpage, people are likely thinking of one of at least three things:
HDR as display tech (hardware)
HDR as wide gamut data format (content)
HDR as tone mapping (processing)
...
So when the first paragraph says we finally explain what HDR actually means, it set me off on the wrong foot—it comes across pretty strongly for a term that’s notoriously context-dependent. Especially in a blog post that reads like a general explainer rather than a direct Q&A response when not coming through your apps channels.
Then followed up by The first HDR is the "HDR mode" introduced to the iPhone camera in 2010. caused me to write the comment.
For people over 35 with even the faintest interest in photography, the first exposure to the HDR acronym probably didn’t arrive with the iPhone in 2010, but HDR IS equivalent to Photomatix style tone mapping starting in 2005 as even mentioned later. The ambiguity of the term is a given now. I think it's futile to insist or police one meaning other the other in non-scientific informal communication, just use more specific terminology.
So the correlation of what HDR means or what sentiment it evokes in people by age group and self-assesed photography skill might be something worthwhile to explore.
The post get's a lot better after that. That said, I really did enjoy the depth. The dive into the classic dodge and burn and the linked YouTube piece. One explainer at a time makes sense—and tone mapping is a good place to start. Even tone mapping is fine in moderation :)
Just out of curiosity since your profile suggests your from an older cohort.
Do you actively remember the Pixelmatix tone mapping era, or where you already old enough to see this as a passing fad, or was this a more niche thing than I remember?
I was old enough to see it as the passing fad it was.
Niche, style points first kind of thing for sure.
Meta: old enough that getting either a new color not intended, or an additional one visible on screen and having the machine remain able to perform was a big deal.
I missed the MDA/EGA/CGA/Hercules era and jumped right into glorious VGA. Only start options for some DOS-games informed you about that drama in the mid 90s not having any idea what that meant otherwise.
It is a fun era NOW. I love the pre VGA PC and earlier systems, like Apple 2, Atari. Am building a CGA system with a K2 CPU to go hacking on to see what was possible. I have, as do many, unfinished business :)
Back then, it was fun at times, bit was also limiting in ways sometimes hard to fathom ways.
Things are crazy good now, BTW. Almost anything is a few clicks away. The CRT is old, panels so damn good..
> "The first HDR is the "HDR mode" introduced to the iPhone camera in 2010."
Yeah, I had a full halt and process exception on that line too. I guess all the research, technical papers and standards development work done by SMPTE, Kodak, et al in the 1990s and early 2000s just didn't happen? Turns out Apple invented it all in 2010 (pack up those Oscars and Emmys awarded for technical achievement and send'em back boys!)
This post is written for people who have heard "HDR" and feel confused. That introduction lists two types of HDR people might think about. "The first" means "the first of two types we're going to explain," not "the first research in the chronological history of HDR."
They also make no mention of transfer functions, which is the main mechanism which explains why the images “burn your eyes” – content creators should use HLG (which has relative luminance) and not PQ (which has absolute luminance) when they create HDR content for the web.
In theory PQ specifies absolute values, but in practice it's treated as relative. Go load some PQ encoded content on an iPhone, adjust your screen brightness, and watch the HDR brightness also change. Beyond the iPhone, it would be ridiculous to render absolute values as-is, given SDR white is supposedly 100-nits; that would be unwatchable in most living rooms.
Bad HDR boils down to poor taste and the failure of platforms to rein it in. You can't fix bad HDR by switching encodings any more than you can fix global warming by switching from Fahrenheit to Celsius.
I predict HDR content on the web will eventually be disabled or mitigated on popular browsers similarly to how auto-playing audio content is no longer allowed [1]
Spammers and advertisers haven't caught on yet to how abusively attention grabbing eye-searingly bright HDR content can be, but any day now they will and it'll be everywhere.
This seems like a fairly easy problem to solve from a UX standpoint, even moreso than auto-playing audio/video. Present all pages in SDR by default, let the user click a button on the toolbar or url bar to toggle HDR rendering when HDR content is detected.
They haven't, but influencers certainly have, I get regular still images which are rendered as a video to get the HDR brightness boost in Instagram, etc.
It seems fine to me. Made sense on the first read and matches my experiences with OpenEXR and ProPhoto RGB and pre-Apple monitors.
High dynamic resolution has always been about tone mapping. Post-sRGB color profile support is called “Wide color” these days, has been available for twenty years or more on all DSLR cameras (such as Nikon ProPhoto RGB supported in-camera on my old D70), and has nothing to do with the dynamic range and tone mapping of the photo. It’s convenient that we don’t have to use EXR files anymore, though!
An HDR photo in sRGB will have the same defects beyond peak saturation at any given hue point, as an SDR photo in sRGB would, relative to either in DCI-P3 or ProPhoto. Even a two-bit black-or-white “what’s color? on or off pixels only” HyperCard dithered image file can still be HDR or SDR. In OKLCH, the selected luminosity will also impact the available chroma range; at some point you start spending your new post-sRGB peak chroma on luminosity instead; but the exact characteristic of that tradeoff at any given hue point is defined by the color profile algorithm, not by whether the photo is SDR or HDR, and the highest peak saturation possible for each hue is fixed, whatever luminosity it happens to be at.
I on the other hand never thought or cared about HDR much before but I remember seeing it everywhere. But I feel the article explains well and clearly with examples, for someone like me who isn't much camera literate.
Isn't that the point of the article? That the colloquial meaning of HDR is quite overloaded, and when people complain about HDR, they mean bad tone-mapping? I say this as someone as close to totally ignorant about photography as you can get; I personally thought the article was pretty spectacular.
When I complain about HDR it's because I've intentionally set the brightness of pure white to a comfortable level, and then suddenly parts of my screen are brighter than that. You fundamentally can't solve that problem with just better tone mapping, can you?
Well yeah. I considered adding that caveat but elected not to because it's obvious and doesn't add anything to the conversation, since that's obviously not what's meant when the industry talks about "HDR". Should've remembered this is HN.
The bit about "confused" turns me off right away. The kind of high-pressure stereo salesman who hopes I am the kind of 'audiophile' who prevents me from calling myself an 'audiophile' (wants mercury-filled cables for a more 'fluid' sound) always presupposes the reader/listener is "confused".
But it's not the colloquial meaning, HDR is fairly well defined by e.g. ITU-R BT.2100, which addresses colorimetry, luminance and the corresponding transfer functions.
I don't think that's the colloquial meaning. If you asked 100 people on the street to describe HDR, I doubt a single person would bring up ITU-R BT.2100.
HDR has a number of different common meanings, which adds to the confusion.
For example, in video games, "HDR" has been around since the mid '00s, and refers to games that render a wider dynamic range than displays were capable of, and use post-process effects to simulate artifcats like bloom and pupil dilation.
In photography, HDR has almost the opposite meaning of what it does everywhere else. Long and multiple exposures are combined to create an image that has very little contrast, bringing out detail in a shot that would normally be lost in shadows or to overexposure.
Photography’s meaning is also about a hundred years older than video games; high(er) dynamic range was a concern in film processing as far back as Ansel, if not prior. That technology adopted it as a sales keyword is interesting and it’s worth keeping in mind when writing for an audience — but this post is about a photography app, not television content or video games, so one can reasonably expect photography’s definition to be used, even if the audience isn’t necessarily familiar.
It may not be the best explanation, but I think any explanation of HDR beyond a sentence or two of definition that doesn't mention the mess that is tone mapping is entirely remiss.
I have a question: how can I print HDR? Is there any HDR printer + paper + display lighting setup?
My hypothesis are the following:
- Increase display lighting to increase peak white point + use a black ink able to absorb more light (can Vantablack-style pigments be made into ink?) => increase dynamic range of a printable picture
- Alternatively, have the display lighting include visible light + invisible UV light, and have the printed picture include an invisible layer of UV ink that shines white : the pattern printed in invisible UV-ink would be the "gain map" to increase the peak brightness past incident visible light into HDR range.
Use projection mapping, instead of lighting the photo uniformly, use a projector and project a copy of the image onto the photo, thus you get detailed control over how much light the parts of the image receive.
Alternatively, use transparent film and a bright backlight.
You would need a very strong light surrounded by a darkish environent. I don't see a point in "HDR Paper" if you need to place it under a special light.
Having come from professional video/film tooling in the 90's to today, it's interesting to see the evolution of what "HDR" means. I used to be a purist in this space, where SDR meant ~8 stops (powers of two) or less of dynamic range, and HDR meant 10+. Color primaries and transfer function mapping were things I spoke specifically about. At this point, though, folks use "HDR" to refer to combinations of things.
Around this, a bunch of practical tooling surfaced (e.g., hybrid log approaches to luminance mapping) to extend the thinking from 8-bit gamma-mapped content presenting ~8 stops of dynamic range to where we are now. If we get away from just trying to label everyting "HDR", there are some useful things people should familiarize with:
1. Color primaries: examples - SDR: Rec. 601, Rec. 709, sRGB. HDR: Rec. 2020, DCI-P3. The new color primaries expand the chromatic representation capabilities. This is pretty easy to wrap our heads around: https://en.wikipedia.org/wiki/Rec._2020
2. Transfer functions: examples - SDR: sRGB, BT.1886. HDR: Rec. 2100 Perceptual Quantizer (PQ), HLG. The big thing in this space to care about is that SDR transfer functions had reference peak luminance but were otherwise relative to that peak luminance. By contrast, Rec. 2100 PQ code points are absolute, in that each code value has a defined meaning in measurable luminance, per the PQ EOTF transfer function. This is a big departure from our older SDR universe and from Hybrid Log Gamma approaches.
3. Tone mapping: In SDR, we had the comfort of camera and display technologies roughly lining up in the video space, so living in a gamma/inverse-gamma universe was fine. We just controlled the eccentricity of the curve. Now, with HDR, we have formats that can carry tone-mapping information and transports (e.g., HDMI) that can bidirectionally signal display target capabilities, allowing things like source-based tone mapping. Go digging into HDR10+, Dolby Vision, or HDMI SBTM for a deep rabbit hole. https://en.wikipedia.org/wiki/Tone_mapping
So HDR is everything (and nothing), but it's definitely important. If I had to emphasize one thing that is non-obvious to most new entrants into the space, it's that there are elements of description of color and luminance that are absolute in their meaning, rather than relative. That's a substantial shift. Extra points for figuring out that practical adaptation to display targets is built into formats and protocols.
The dad’s photo in the end in SDR looks so much better on a typical desktop IPS panel (Windows 11). The HDR photo looks like the brightness is smushed in the most awful way. On an iPhone, the HDR photo is excellent and the others look muted.
I wonder if there’s an issue in Windows tonemapping or HDR->SDR pipeline, because perceptually the HDR image is really off.
It’s more off than if I took an SDR picture of my iPhone showing the HDR image and showed that SDR picture on the said Windows machine with an IPS panel. Which tells me that the manual HDR->SDR “pipeline” I just described is better.
I think Windows showing HDR content on a non-HDR display should just pick an SDR-sized section of that long dynamic range and show it normally. Without trying to remap the entire large range to a smaller one. Or it should do some other perceptual improvements.
Then again, I know professionally that Windows HDR is complicated and hard to tame. So I’m not really sure the context of remapping as they do, maybe it’s the only way in some contingency/rare scenario.
I'm not entirely convinced that greedy influencers are to blame for people hating on overly bright content. Instead, I think something is different with how displays produce brightness compared to just the nature outside. Light outside is supposed to reach up to tens of thousands of nits, yet even 1000 nits is searing on a display. Is it that displays output polarized light? Is it the spectral distribution of especially the better displays being three really tight peaks? I cannot tell you, but I'm suspecting something isn't quite right.
All this aside, HDR and high brightness are different things - HDR is just a representational thing. You can go full send on your SDR monitor as well, you'll just see more banding. The majority of the article is just content marketing about how they perform automatic tonemapping anyways.
It's all down to the ambient light. That's why bias lighting is now a thing. Try putting a light behind your screen to massively brighten the wall behind it, the 1000 nit peaks will be far less harsh. And if you bring the device out into sunlight I suspect you will wish for everything about its output to be quite a bit brighter.
> Light outside is supposed to reach up to tens of thousands of nits, yet even 1000 nits is searing on a display.
That’s a consequence of https://en.wikipedia.org/wiki/Adaptation_(eye). If you look at 1000 nits on a display in bright sunlight, with your eyes adapted to the bright surroundings, the display would look rather dim.
I find the default HDR (as in gain map) presentation of iPhone photos to look rather garish, rendering highlights too bright and distracting from the content of the images. The solution I came up with for my own camera app was to roll off and lower the highlights in the gain map, which results in final images that I find way more pleasing. This seems to be somewhat similar to what Halide is introducing with their "Standard" option for HDR.
Hopefully HN allows me to share an App Store link... this app works best on Pro iPhones, which support ProRAW, although I do some clever stuff on non-Pro iPhones to get a more natural look.
I see that it says it stores images "in an HDR format by default" but keeps referencing JPEG output. Are you using JPEG-XT? There aren't a lot of "before and after" comparisons so it's hard to know how much it's taking out. I figure those would probably hurt the reputation of the app considering its purpose is to un-pop photos, but I'm in the boat of not really being sure whether I do actually like the pop or not. Is there live-photo support, or is that something that you shouldn't expect from a artist-focused product?
It's a JPEG + gain map format where the gain map is stored in the metadata. Same thing, as far as I can tell, that Halide is now using. It's what the industry is moving towards; it means that images display well on both SDR and HDR displays. I don't know what JPEG-XT is, aside from what I just skimmed on the Wikipedia page.
Not having before-and-after comparisons is mostly down to my being concerned about whether that would pass App Review; the guidelines indicate that the App Store images are supposed to be screenshots of the app, and I'm already pushing that rule with the example images for filters. I'm not sure a hubristic "here's how much better my photos are than Apple's" image would go over well. Maybe in my next update? I should at least have some comparisons on my website, but I've been bad at keeping that updated.
There's no Live Photo support, though I've been thinking about it. The reason is that my current iPhone 14 Pro Max does not support Live Photos while shooting in 48-megapixel mode; the capture process takes too long. I'd have to come up with a compromise such as only having video up to the moment of capture. That doesn't prevent me from implementing it for other iPhones/cameras/resolutions, but I don't like having features unevenly available.
I have similar feeling with HDR to what I have with movie audio mixing. The range is just too big. I think this distaste also amplified in me with age. I appreciate content that makes use of the limited space it gets, and also algorithms that compress the range for me a bit. KMPlayer for example has this on for volume by default, and it makes home movie watching more comfortable, no doubt sacrificing artistic vision in the process. I feel a bit the same with the loudness war - and maybe a lot of other people feel the same too, seeing how compressed audio got. At the very least they don't mind much.
I really appreciate the article. I could feel that they also have a product to present, because of the many references, but it was also very informative besides that.
I do prefer the gain map approach myself. Do a sensible '32 bit' HDR edit, further tweak an SDR version, and export a file that has both.
Creative power is still in your hands versus some tone mapper's guesses at your intent.
Can people go overboard? Sure, but thats something they will do regardless of any hdr or lack thereof.
On an aside its still rough that just about every site that touches gain map (adaptive HDR as this blog calls them) HDR images will lose that metadata if they need to scale, recompress, or transform the images otherwise. Its led me to just make my own site, but also to handle what files a client gets a bit smarter . For instance if a browser doesnt support .jxl or .avif images, im sure it wont want an hdr jpeg either, thats easy to handle on a webserver.
The article starts by saying HDR can mean different things and gives Apple's HDR vs new TV "HDR" advertsing but doesn't explain at all what the TV HDR means and how it is different, unless I missed something. I always assumed they were the same thing.
A lot of these design flaws are fixed by Firefox's picture in picture option but for some reason, with the way you coded it, the prompt to pop it out as PIP doesn't show up
Did I miss something or is the tone mapper not released yet? I admit I'm multitasking here, but just have an exposure slider in the image lab (or whatever it's called).
Sidebar: I kinda miss when Halide's driving purpose was rapid launch and simplicity. I would almost prefer a zoom function to all of this HDR gymnastics (though, to be clear, Halide is my most-used and most-liked camera app).
EDIT: Ah, I see, it's a Mark III feature. That is not REMOTELY clear in the (very long) post.
Is there a consensus definition of what counts as "HDR" in a display? What is the "standard dynamic range" of a typical TV or computer monitor? Is it roughly the same for devices of the same age?
My understanding is most SDR TVs and computer screens have displays about 200-300 nits (aka cd/m²). Is that the correct measure of the range of the display? The brightest white is 300 nits brighter than the darkest black?
this isn't a correct definition. human perception of brightness is roughly logarithmic, so it also matters how deep the blacks get. For a good HDR experience, you need a monitor that gets to 600 nits at a bare minimum, but also which can get very close to 0 nits (e.g. via OLED or less optionally local dimming)
Human vision has around 20 stops of static dynamic range. Modern digital cameras can't match human vision— a $90,000 Arri Alexa boasts 17 stops— but they're way better than SDR screens.
No. When you simply adjust shadow and highlights, you lose local contrast. In an early draft of the post, there was an example, but it was cut for pacing.
As a non-techie I represent the 99.9% of the population who haven't a clue what tone mapping etc. is: NO WAY would we ever touch the various settings possible as opposed to watching the TV/computer screen/etc. as it came out of the box.
As a photographer, one of the things that draws me to the work of artists like Bill Brandt and Daido Moriyama is their use of low dynamic range, and high contrast. I rarely see an HDR image that is aesthetically interesting.
Back in university I implemented a shoddy HDR for my phone camera.
The hardest part of it, by far, was taking hundreds upon hundreds of pictures of a blank piece of paper in different lighting conditions with different settings.
Is there any workflow that can output HDR photos (like the real HDR kind, with metadata to tell the display to go into HDR mode) for photos shot with a mirrorless and not an iPhone?
Yes. For example, Lightroom and Camera Raw support HDR editing and export from RAW images, and Adobe published a good rundown on the feature when they introduced it.
In my opinion, HDR is another marketing gimmick -- the average layman has no idea what it means, but it sounds fancy and is more expensive, so surely it's good.
i am still skeptical about HDR as pretty much all HDR content I see online is awful. But this post makes me believe that Lux/Halide can pull of HDR in a way that I will like. I am looking forward to Halide Mk3.
Well crap, I had written a thank you, which I will gladly write again:
I love when product announcements and ADS in general are high value works. This one was good education for me. Thank you for it!
I had also written about my plasma and CRT displays and how misunderstandings about HDR made things generally worse and how I probably have not seen the best these 10 bit capable displays can do.
And finally, I had written about 3D TV and how fast, at least 60Hz per eye, 3D in my home made for amazing modeling and assembly experiences! I was very sad to see that tech dead end.
3D for technical content create has a lot of legs... if only more people could see it running great...
... something Linux Desktops don't understand and Macs only do well with their own displays with videos. Guess who the winner is on the desktop: Windows oO
Linux does HDR. It's *technically" worked for years but none of the big DEs had it working. Now Ubuntu 25.04 and Fedora 42 (Gnome) have it working out of the box.
No, it isn't. Absolute luminance is a "feature" of PQ specifically used by HDR10(+) and most DolbyVision content (notably the DolbyVision as produced by an iPhone is not PQ, it's not "real" DolbyVision). But this is not the only form of HDR and it's not even the common form for phone cameras. HLG is a lot more popular for cameras and it is not in absolute luminance. The gainmap-based approach that Google, Apple, and Adobe are all using is also very much not absolute luminance, either. In fact that flips it entirely and it's SDR relative instead, which is a much better approach to HDR than what video initially went with.
Ideally in the abstract it could be just that, but in practice it's an umbrella name for many different techniques that provide some aspect of that goal.
Not in the more general sense! It can refer to what its acronym spells out directly: Bigger range between dimmest and brightest capabilities of a display, imaging technique etc.
No. HDR can encode high dynamic range because (typically) it uses floating point encoding.
From a technical point of view, HDR is just a set of standards and formats for encoding absolute-luminance scene-referred images and video, along with a set of standards for reproduction.
No. HDR video (and images) don't use floating point encoding. They generally use a higher bit depth (10 bits or more vs 8 bits) to reduce banding and different transfer characteristics (i.e. PQ or HLG vs sRGB or BT.709), in addition to different YCbCr matrices and mastering metadata.
And no, it's not necessarily absolute luminance. PQ is absolute, HLG is not.
Also DCI-P3 should fit in here somewhere, as it seems to be the most standardized color space for HDR. I would share more insight, if I had it. I thought I understood color profiles well, but I have encountered some challenges when trying to display in one, edit in another, and print “correctly”. And every device seems to treat color profiles a little bit differently.
All transfer functions can generally work on either integer range or floating point. They basically just describe a curve shape, and you can have that curve be over the range of 0.0-1.0 just as easily as you can over 0-255 or 0-1023.
Extended sRGB is about the only thing that basically requires floating point, as it specifically describes 0.0-1.0 as being equivalent to sRGB and then has a valid range larger than that (you end up with something like -.8 to 2.4 or greater). And representing that in integer domain is conceptually possible but practically not really.
> Also DCI-P3 should fit in here somewhere, as it seems to be the most standardized color space for HDR.
BT2020 is the most standardized color space for HDR. DCI-P3 is the most common color gamut of HDR displays that you can actually afford, however, but that's a smaller gamut than what most HDR profiles expect (HDR10, HDR10+, and "professional" DolbyVision are all BT2020 - a wider gamut than P3). Which also means most HDR content specifies a color gamut it doesn't actually benefit from having as all that HDR content is still authored to only use somewhere between the sRGB and DCI-P3 gamut since that's all anyone who views it will actually have.
The math uses real numbers but table 2-4 ("Digital representation") discusses how the signal is quantized to/from analog and digital. The signal is quantized to integers.
This same quantization process is done for sRGB, BT.709, BT.2020, etc. so it's not unique to HLG. It's just how digital images/video are stored.
I think most HDR formats do not typically use 32 bit floating point. The first HDR file format I can remember is Greg Ward’s RGBE format, which is also now more commonly known as .HDR and I think is pretty widely used.
It uses a type of floating point, in a way, but it’s a shared 8 bit exponent across all 3 channels, and the channels are still 8 bits each, so the whole thing fits in 32 bits. Even the .txt file description says it’s not “floating point” per-se since that implies IEEE single precision floats.
Cameras and displays don’t typically use floats, and even CG people working in HDR and using, e.g., OpenEXR, might use half floats more often that float.
Some standards do exist, and it’s improving over time, but the ideas and execution of HDR in various ways preceded any standards, so I think it’s not helpful to define HDR as a set of standards. From my perspective working in CG, HDR began as a way to break away from 8 bits per channel RGB, and it included improving both color range and color resolution, and started the discussion of using physical metrics as opposed to relative [0..1] ranges.
I did my PhD in Atomic, Molecular, and Optical (AMO) physics, and despite "optical" being part of that I realized midway that I didn't know enough about how regular cameras worked!
It didn't take very long to learn, and it turned out to be extremely important in the work I did during the early days at Waymo and later at Motional.
I wanted to pass along this fun video from several years ago that discusses HDR: https://www.youtube.com/watch?v=bkQJdaGGVM8 . It's short and fun, I recommend it to all HN readers.
Separately, if you want a more serious introduction to digital photography, I recommend the lectures by Marc Levoy from his Stanford course: https://www.youtube.com/watch?v=y7HrM-fk_Rc&list=PL8ungNrvUY... . I believe he runs his own group at Adobe now after leading a successful effort at Google making their pixel cameras the best in the industry for a couple of years. (And then everyone more-or-less caught up, just like with most tech improvements in the history of smartphones).
That Kevin Chen video using Excel was amazing and hilarious. Thanks.
Try capturing fire with a non-Sony phone and a Sony phone. At least Samsung doesn't color correct blackbodies right and the flame looks nothing like reality.
Pixel camera hardware or software? Isnt there only one vendor for sensors - Sony?
Samsung also makes sensors for phones. IIRC some Pixels use their sensors.
I think Canon makes at least some of their sensors, and Nikon designs theirs and makes it at a third party I forget the name of that isn't Sony or Samsung but they still do use Sony stuff in a lot of their cameras.
I don't know about Pentax, Panasonic or OMD (formerly Olympus)
I think folks here have some idea how expensive chip fabs are. That's why only Canon is able to make their own sensors.
Sony makes sensors for pretty much everyone else. But it's well known that other folks e.g. Nikon have been able to get better signal-to-noise with Sony-made sensors than Sony themselves. I think Panasonic used to make their own sensors but with some recent re-org, that got spun out.
It's been widely rumored that Leica uses Sony sensors, but this gets repeatedly denied by people claiming inside information. We know that Leica was getting 24MP CMOS sensors from CMOSIS in the 2012 timeframe, but CMOSIS has since been acquired by OSRAM, and there hasn't been any verifiable information since then, whether confirming or denying a continued business relationship.
Is Sony the ASML of the sensor world?
Definitely
He worked mostly on the software side, but of course had important input into what sensors and processors were chosen for the phones.
> Our eyes can see both just fine.
This gets to a gaming rant of mine: Our natural vision can handle these things because our eyes scan sections of the scene with constant adjustment (light-level, focus) while our brain is compositing it together into what feels like a single moment.
However certain effects in games (i.e. "HDR" and Depth of Field) instead reduce the fidelity of the experience. These features limp along only while our gaze is aimed at the exact spot the software expects. If you glance anywhere else around the scene, you instead percieve an unrealistically wrong coloration or blur that frustratingly persists no matter how much you squint. These problems will remain until gaze-tracking support becomes standard.
So ultimately these features reduce the realism of the experience. They make it less like being there and more like you're watching a second-hand movie recorded on flawed video-cameras. This distinction is even clearer if you consider cases where "film grain" is added.
https://www.realtimerendering.com/blog/thought-for-the-day/
It's crazy that post is 15 years old. Like the OP and this post get at, HDR isn't really a good description of what's happening. HDR often means one or more of at least 3 different things (capture, storage, and presentation). It's just the sticker slapped on advertising.
Things like lens flares, motion blur, film grain, and shallow depth of field are mimicking cameras and not what being there is like--but from a narrative perspective we experience a lot of these things through tv and film. Its visual shorthand. Like Star Wars or Battlestar Galactica copying WWII dogfight footage even though it's less like what it would be like if you were there. High FPS television can feel cheap while 24fps can feel premium and "filmic."
Often those limitations are in place so the experience is consistent for everyone. Games will have you set brightness and contrast--I had friends that would crank everything up to avoid jump scares and to clearly see objects intended to be hidden in shadows. Another reason for consistent presentation is for unfair advantages in multiplayer.
> Things like lens flares, motion blur, film grain, and shallow depth of field are mimicking cameras and not what being there is like
Ignoring film grain, our vision has all these effects all the same.
Look in front of you and only a single plane will be in focus (and only your fovea produces any sort of legibility). Look towards a bright light and you might get flaring from just your eyes. Stare out the side of a car or train when driving at speed and you'll see motion blur, interrupted only by brief clarity if you intentionally try to follow the motion with your eyes.
Without depth of field simulation, the whole scene is just a flat plane with completely unrealistic clarity, and because it's comparatively small, too much of it is smack center on your fovea. The problem is that these are simulations that do not track your eyes, and make the (mostly valid!) assumption that you're looking, nearby or in front of whatever you're controlling.
Maybe motion blur becomes unneccessary given a high enough resolution and refresh rate, but depth of field either requires actual depth or foveal tracking (which only works for one person). Tasteful application of current techniques is probably better.
> High FPS television can feel cheap while 24fps can feel premium and "filmic."
Ugh. I will never understand the obsession this effect. There is no such thing as a "soap opera effect" as people liek to call it, only a slideshow effect.
The history behind this is purely a series of cost-cutting measures entirely unrelated to the user experience or artistic qualities. 24 fps came to be because audio was slapped onto the film, and was the slowest speed where the audio track was acceptable intelligible, saving costly film paper - the sole priority of the time. Before that, we used to record content at variable frame rates but play it back at 30-40 fps.
We're clinging on to a cost-cutting measure that was a significant compromise from the time of hand cranked film recording.
</fist-shaking rant>
> Look in front of you and only a single plane will be in focus (and only your fovea produces any sort of legibility). Look towards a bright light and you might get flaring from just your eyes. Stare out the side of a car or train when driving at speed and you'll see motion blur, interrupted only by brief clarity if you intentionally try to follow the motion with your eyes.
The problem is the mismatch between what you’re looking at on the screen and what the in-game camera is looking at. If these were synchronised perfectly it wouldn’t be a problem.
Indeed - I also mentioned that in the paragraph immediately following.
Derp
> Ugh. I will never understand the obsession this effect.
All of it (lens flares, motion blur, film grain, DoF, tone mapping, and exposure, frame rate) are artistic choices constrained by the equipment we have to collect and present it. I think they'll always follow trends. In my entire career following film, photography, computer graphics, and game dev the only time I've heard anyone talk about how we experience any of those things is when people say humans see roughly equivalent of a 50mm lens (on 35mm film).
Just look at the trend of frame size. Film was roughly 4:3, television copied it. Film started matting/cropping the frame. It got crazy with super wide-screen to where some films used 3 projectors side-by-side and most settled on 16:9. Then television copied it. Widescreen is still seen as more "filmic." I remember being surprised working on a feature that switched to Cinemascope's aspect ratio and seeing that was only 850 pixels tall--a full frame would be about twice that.
To me, high frame rate was always just another style. My only beef was with motion-smoothing muddying up footage shot at different frame rates.
The problem is that it just doesn't work on modern, fast displays. Without motion smoothing on a fast and bright screen, 24fps/30fps goes from "choppy" to "seizure inducing and unwatchable". Older sets would just naturally smooth things out.
Even on my LCD TV, smooth motion like credits at certain speeds are extremely uncomfortable to look at at these frame rates.
I consider it borderline irresponsible to continue using these framerates, forcing users into frame interpolation and horrible artifacts, a decision the manufacturer might even have made for them. 120 Hz is finally becoming the norm for regular content (with monitors going to 500+ nowadays), we should at least be able to get to 60 Hz as the lower bound for regular content delivery.
Going further down for artistic value, e.g. for stop motion or actual slide shows is less of a problem in my opinion. It is not as disturbing, and if regular content was appropriately paced there would be no need for interpolation to mess with it...
> Just look at the trend of frame size.
Frame size is different from the other parameters, as it is solely a physical practicality. Bigger is better in all directions, but a cinema screen needs to fit in the building - making a building much taller is less economical than making it wider, and making it whatever it isn't right now adds novelty.
The content needs to be made for the screen with the appropriate balance of periphery and subject to not be completely wrong, so screen technology and recording technology tends to align. Economy of scale causes standardization on lenses and image circles, and the choice of aspect ratio within that circle on the film, forming a feedback loop that enforces the parameters for almost all content.
If some technology somewhere else in the stack causes a change, some will follow for the novelty but others will simply follow the reducing cost, and soon all content aligns on the format, and the majority of home TV sets will be shaped to fit the majority content it can receive.
> The problem is that it just doesn't work on modern, fast displays.
I'm very confused by this. From what I've seen it's been getting a lot better (since transitioning from CRTs). At least for television, frame-rate matching is becoming more of a thing. Higher frame rates really help. Calling everything fps for simplicity; 120 divides evenly by 24, 30, and 60. Lower values wont match and cause issues.
Similarly, (maybe back in the 90s?) projectors in theaters would double-expose each frame to reduce the flicker in between frames. With digital, they no longer have to advance the film between frames.
> smooth motion like credits at certain speeds are extremely uncomfortable to look at at these frame rates.
I think scrolling credits are the most difficult use case: white on black with hard text and no blur. DLP projectors (common 10+ years ago) drive me nuts displaying R G and B separately.
Outside of credits, cinematographers and other filmmakers do think about these things. I remember hearing a cinematographer talk about working on space documentaries in Imax. If you panned too quickly, the white spaceship over a black star field could jump multiple feet each frame. Sure films shot today are optimized for the theater, but the technology gap between theater and home is nowhere near as crazy as CRT vs acetate.
> Frame size is different from the other parameters, as it is solely a physical practicality.
I'm still struggling to see how its that different. Widescreen meant a lower effective resolution (it didn't have to--it started with Cinerama and Cinemascope), but was adopted for cost and aesthetic reasons.
> If some technology somewhere else in the stack causes a change…and soon all content aligns on the format, and the majority of home TV sets will be shaped to fit the majority content it can receive.
And the industry and audiences are really attached to 24fps. Like you say, home televisions adopted film's aspect ratio and I've also seen them adopt much better support for 24fps.
As kind of an aside, I wonder if the motion blur is what people are attached to more than the actual frame rate. I assume you're talking about frame rates higher than 30? Sure, we have faster films and brighter lights, but exposure time is really short. I saw the Hobbit in theaters in both high frame rate and 24fps and the 24fps one looked weird to me, too--I meant to look it up, but I assume they just dropped frames making the blur odd.
> the poster found it via StumbleUpon.
Such a blast from the past, I used to spend so much time just clicking that button!
I'm with you on depth of field, but I don't understand why you think HDR reduces the fidelity of a game.
If you have a good display (eg an OLED) then the brights are brighter and simultaneously there is more detail in the blacks. Why do you think that is worse than SDR?
Check out this old post: https://www.realtimerendering.com/blog/thought-for-the-day/
HDR in games would frequently mean clipping highlights and adding bloom. Prior the "HDR" exposure looked rather flat.
That's not what it means since 2016 or so when consumer TVs got support for properly displaying brighter whites and colors.
It definitely adds detail now, and for the last 8-9 years.
Though consumer TVs obviously still fall short of being as bright at peak as the real world. (We'll probably never want our TV to burn out our vision like the sun, though, but probably hitting highs at least in the 1-2000nit range vs the 500-700 that a lot peak at right now would be nice for most uses.
OK, so it doesn't mean real HDR but simulated HDR.
Maybe when proper HDR support becomes mainstream in 3D engines, that problem will go away.
Right. Just like the article, HDR is too vague to mean anything specific and a label that's slapped onto products. In gaming, it often meant they were finally simulating light and exposure separately--clipping highlights that would have previously been shown. In their opinion, reducing the fidelity. Same with depth of field blurring things that used to not have blur.
It's HDR at the world data level, but SDR at the rendering level. It's simulating the way film cannot handle real-life high dynamic range and clips it instead of compressing it like "HDR" in photography.
> Instead of compressing it like "HDR" in photography
That's not HDR either, that's tone mapping to SDR. The entire point of HDR is that you don't need to compress it because your display can actually make use of the extra bits of information. Most modern phones take true HDR pictures that look great on an HDR display.
That's revisionist nomenclature. HDR photography has meant tone mapping long before HDR displays existed.
The “HDR” here is in the sense of “tone mapping to SDR”. Should also be said that even “H” DR displays only have a stop or two of more range, still much less than in a real-world high-contrast scenes
It's still better though.
HDR displays are >1000nits while SDR caps out at less than 500nits even on the best displays.
Eg for the Samsung s90c, HDR is 1022nits, SDR is 487nits: https://www.rtings.com/tv/reviews/samsung/s90c-oled#test_608 https://www.rtings.com/tv/reviews/samsung/s90c-oled#test_4
Double the range is undeniably still better.
And also 10bit instead of 8bit, so less posterization as well.
Just because the implementations have been subpar until now doesn't mean it's worthless tech to pursue.
> HDR displays are >1000nits
Displays as low as 400nits have been marketed as "HDR".
But nits are only part of the story. What really matters at the end is the range between the darkest and brightest color the display can show under the lighting conditions you want to use it as. 400 nits in a darkened room where blacks are actually black can have much more actual range than 1000nits with very bright "blacks" due to shitty display tech or excessive external illumination.
The most egregious example is 3D. Only one thing is in focus, even though the scene is stereoscopic. It makes no sense visually.
Hell yeah, this one of many issues I had with the first Avatar movie. The movie was so filled with cool things to look at but none of it was in focus. 10 minutes in I had had enough and was ready for a more traditional movie experience. Impressive yes, for 10 minutes, then exhausting.
this thread is helping me understand why I always thought 3D movies looked _less_ 3D than 2D movies.
That and after seeing Avatar 1 in 3D, then seeing Avatar 2 in 3D over 10 years later and not really noticing any improvement in the 3D made me declare 3D movies officially dead (though I haven’t done side by side comparisons)
Hah - Avatar was exactly what I was thinking of.
I had a similar complaint with the few 3D things I watched when that has been hyped in the past (e.g., when Avatar came out in cinemas, and when 3D home TVs seemed to briefly become a thing 15 years ago). It felt like Hollywood was giving me the freedom to immerse myself, but then simultaneously trying to constrain that freedom and force me to look at specific things in specific ways. I don't know what the specific solution is, but it struck me that we needed to be adopting lessons from live stage productions more than cinema if you really want people to think what they're seeing is real.
Stereo film has its own limitations. Sadly, shooting for stereo was expensive and often corners were cut just to get it to show up in a theater where they can charge a premium for a stereo screening. Home video was always a nightmare--nobody wants to wear glasses (glassesless stereo TVs had a very narrow viewing angle).
It may not be obvious, but film has a visual language. If you look at early film, it wasn't obvious if you cut to something that the audience would understand what was going on. Panning from one object to another implies a connection. It's built on the visual language of still photography (things like rule of thirds, using contrast or color to direct your eye, etc). All directing your eye.
Stereo film has its own limitations that were still being explored. In a regular film, you would do a rack focus to connect something in the foreground to the background. In stereo, when there's a rack focus people don't follow the camera the same way. In regular film, you could show someone's back in the foreground of a shot and cut them off at the waist. In stereo, that looks weird.
When you're presenting something you're always directing where someone is looking--whether its a play, movie, or stereo show. The tools are just adapted for the medium.
I do think it worked way better for movies like Avatar or How to Train Your Dragon and was less impressive for things like rom coms.
HDR, not "HDR", is the biggest leap in gaming visuals made in the last 10 years, I think.
Sure, you need a good HDR-capable display and a native HDR-game (or RTX HDR), but the results are pretty awesome.
These effects are for the artistic intent of the game. Same goes for movies, and has nothing to do with "second hand movies recorded on flawed cameras". or with "realism" in the sense of how we perceive the world.
This is why I always turn off these settings immediately when I turn on any video game for the first time. I could never put my finger on why I didn’t like it, but the camera analogy is perfect
Ok. I like depth of field and prefer it.
It seems like a mistake to lump HDR capture, HDR formats and HDR display together, these are very different things. The claim that Ansel Adams used HDR is super likely to cause confusion, and isn’t particularly accurate.
We’ve had HDR formats and HDR capture and edit workflows since long before HDR displays. The big benefit of HDR capture & formats is that your “negative” doesn’t clip super bright colors and doesn’t lose color resolution in super dark color. As a photographer, with HDR you can re-expose the image when you display/print it, where previously that wasn’t possible. Previously when you took a photo, if you over-exposed it or under-exposed it, you were stuck with what you got. Capturing HDR gives the photographer one degree of extra freedom, allowing them to adjust exposure after the fact. Ansel Adams wasn’t using HDR in the same sense we’re talking about, he was just really good at capturing the right exposure for his medium without needing to adjust it later. There is a very valid argument to be made for doing the work up-front to capture what you’re after, but ignoring that for a moment, it is simply not possible to re-expose Adams’ negatives to reveal color detail he didn’t capture. That’s why he’s not using HDR, and why saying he is will only further muddy the water.
Arguably, even considering HDR a distinct thing is itself weird an inaccurate.
All mediums have a range, and they've never all matched. Sometimes we've tried to calibrate things to match, but anyone watching SDR content for the past many years probably didn't do so on a color-calibrated and brightness calibrated screen - that wouldn't allow you to have a brightness slider.
HDR on monitors is about communicating content brightness and monitor capabilities, but then you have the question of whether to clip the highlights or just map the range when the content is mastered for 4000 nits but your monitor manages 1000-1500 and only in a small window.
This! Yes I think you’re absolutely right. The term “HDR” is in part kind of an artifact of how digital image formats evolved, and it kind of only makes sense relative to a time when the most popular image formats and most common displays were not very sophisticated about colors.
That said, there is one important part that is often lost. One of the ideas behind HDR, sometimes, is to capture absolute values in physical units, rather than relative brightness. This is the distinguishing factor that film and paper and TVs don’t have. Some new displays are getting absolute brightness features, but historically most media display relative color values.
Absolute is also a funny size. From the perspective of human visual perception, an absolute brightness only matters if the entire viewing environment is also controlled to the same absolute values. Visual perception is highly contextual, and we are not only seeing the screen.
It's not fun being unable to watch dark scenes during the day or evening in a living room, nor is vaporizing your retinas if the ambient environment went dark in the meantime. People want good viewing experience in the available environment that is logically similar to what the content intended, but that is not always the same as reproducing the exact same photons as the directors's mastering monitor sent towards their their eyeballs at the time of production.
Indeed. For a movie scene depicting the sky including the Sun, you probably wouldn't want your TV to achieve the same brightness as the Sun. You might want your TV to become significantly brighter than the rest of the scenes, to achieve an effect something like the Sun catching your eye.
Of course, the same thing goes for audio in movies. You probably want a gunshot or explosion to sound loud and even be slightly shocking, but you probably don't want it to be as loud as a real gunshot or explosion would be from the depicted distance.
The difference is that for 3+ decades the dynamic range of ubiquitous audio formats (like 16 bit PCM in audio CDs and DVDs) has provided far more dynamic range than is comfortably usable in normal listening environments. So we're very familiar with audio being mastered with a much smaller dynamic range than the medium supports.
Yep, absolutely! ;)
This brings up a bunch of good points, and it tracks with what I was trying to say about conflating HDR processing with HDR display. But do keep in mind that even when you have absolute value images, that doesn’t imply anything about how you display them. You can experience large benefits with an HDR workflow, even when your output or display is low dynamic range. Assume that there will be some tone mapping process happening and that the way you map tones depends on the display medium and its capabilities, and on the context and environment of the display. Using the term “HDR” shouldn’t imply any mismatch or disconnect in the viewing environment. It only did so in the article because it wasn’t very careful about its terms and definitions.
The term "HDR" arguably makes more sense for the effect achieved by tone mapping multiple exposures of the same subject onto a "normal" (e.g. SRGB) display. In this case, the "high" in "HDR" just means "from a source with higher dynamic range than the display."
Remember "wide gamut" screens ?
This is part of 'HDR' standards too...
And it's quite annoying that 'HDR' (and which specific one ?) is treated as just being 'on' or 'off' even for power users...
> but your monitor manages 1000-1500 and only in a small window.
Owning a display that can do 1300+ nits sustained across a 100% window has been the biggest display upgrade I think I have ever had. It's given me a tolerance for LCD, a technology I've hated since the death of CRTs and turned me away from OLED.
There was a time I would have said i'd never own a non OLED display again. But a capable HDR display changed that logic in a big way.
Too bad the motion resolution on it, especially compared to OLED is meh. Again, at one point, motion was the most important aspect to me (its why I still own CRTs) but this level of HDR...transformative for lack of a better word.
Motion resolution? Do you mean the pixel response time?
CRTs technically have quite a few artifacts in this area, but as content displayed CRTs tend to be built for CRTs this is less of an issue, and in many case even required. The input is expecting specific distortions and effects from scanlines and phosphor, which a "perfect" display wouldn't exhibit...
The aggressive OLED ABL is simply a thermal issue. It can be mitigated with thermal design in smaller devices, and anything that increases efficiency (be it micro lens arrays, stacked "tandem" panels, quantum dots, alternative emitter technology) will lower the thermal load and increase the max full panel brightness.
(LCD with zone dimming would also be able to pull this trick to get even brighter zones, but because the base brightness is high enough it doesn't bother.)
> Motion resolution? Do you mean the pixel response time?
I indeed meant motion resolution, which pixel response time only partially affects. It’s about how clearly a display shows motion, unlike static resolution which only reflects realistically a still image. Even with fast pixels, sample and hold displays blur motion unless framerate and refresh rate is high, or BFI/strobing is used. This blur immediately lowers perceived resolution the moment anything moves on screen.
> The input is expecting specific distortions and effects from scanlines and phosphor, which a "perfect" display wouldn't exhibit...
That's true for many CRT purists, but is not a huge deal for me personally. My focus is motion performance. If LCD/OLED matched CRT motion at the same refresh rate, I’d drop CRT in a heartbeat, slap on a CRT shader, and call it a day. Heresy to many CRT enthusiasts.
Ironically, this is an area in which I feel we are getting CLOSE enough with the new higher refresh OLEDs for non HDR retro content in combination with: https://blurbusters.com/crt-simulation-in-a-gpu-shader-looks... (which hopefully will continue to be improved.)
> The aggressive OLED ABL is simply a thermal issue.
Theoretically, yes and there’s been progress, but it’s still unsolved in practice. If someone shipped an OLED twice as thick and full of fans and heatsinks, I’d buy it tomorrow. But that’s not what the market wants, so obviously it's not what they make.
> It can be mitigated with thermal design in smaller devices, and anything that increases efficiency (be it micro lens arrays, stacked "tandem" panels, quantum dots, alternative emitter technology) will lower the thermal load and increase the max full panel brightness.
Sure, in theory. But so far the improvements (like QD-OLED or MLA) haven’t gone far enough. I already own panels using these. Beyond that, much of the tech isn’t in the display types I care about, or isn’t ready yet. Which is a pity, because the tandem based displays I have seen in usage are really decent.
That said, the latest G5 WOLEDs are the first I’d call acceptable for HDR at high APL, for the preferences I hold with very decent real scene brightness, at least in film. Sadly, I doubt we’ll see comparable performance in PC monitors until many years down the track and monitors are my preference.
Hello fellow CRT owner. What is your use case? Retro video games? PC games? Movies?
Hello indeed!
> What is your use case? Retro video games? PC games? Movies?
All of the above! The majority of my interest largely stems from the fact that for whatever reason, I am INCREDIBLY sensitive to sample and hold motion blur. Whilst I tolerate it for modern gaming because I largely have no choice, CRT's mean I do not for my retro gaming, which I very much enjoy. (I was very poor growing up, so most of it for me is not even nostalgia, most of these games are new to me.)
Outside of that, we have a "retro" corner in our home with a 32" trinitron. I collect laserdisc/VHS and we have "retro video" nights where for whatever reason, we watch the worst possible quality copies of movies we could get in significantly higher definition. Much the same as videogames, I was not exposed to a lot of media growing up, my wife has also not seen many things because she was in Russia back then, so there is a ton for us to catch up on very slowly and it just makes for a fun little date night every now and again.
Sadly though, as I get ready to take on a mortgage, it's likely most of my CRT's will be sold, or at least the broadcast monitors. I do not look forward to it haha.
> Outside of that, we have a "retro" corner in our home with a 32" trinitron.
A 32” Trinny. Nice. I have the 32” JVC D-series which I consider my crown jewel. It’s for retro gaming and I have a laserdisc player but a very limited selection of movies. Analog baby.
> Sadly though, as I get ready to take on a mortgage, it's likely most of my CRT's will be sold
Mortgage = space. You won’t believe the nooks and crannies you can fit CRTs into. Attic. Shed. Crawl space. Space under basement stairs. Heck, even the neighbors house. I have no less than 14 CRTs ferreted away in the house. Wife thinks I have only 5. Get creative. Don’t worry about the elements, these puppies were built to survive nuclear blasts. Do I have a sickness? Probably. But analog!!!
Speaking of laser disc it's wild how vivid colors are on that platform. My main example movie is Star Trek First contact and everything is very colorful. DVD is muddy. Even a Blu-ray copy kinda looks like crap. A total side note is the surround sound for that movie is absolutely awesome especially the cube battle scene.
> I have the 32” JVC D-series which
I would love one of these however I have never seen one in my country. Super jealous haha! The tubes they use apparently were an american made tube, with most of the JVCs that were released in my country using different tubes than those released in the US market.
That being said, I do own two JVC "broadcast" monitors that I love. A 17" and a 19". They are no D-series real "TV" but.
Adams adjusted heavily with dodging and burning, even working to invent a new chemical process to provide more control when developing. He was great at determining exposure for his process as well. A key skill was having a vision for what the image would be after adjusting. Adams talked a lot about this as a top priority of his process.
> It's even more incredible that this was done on paper, which has even less dynamic range than computer screens!
I came here to point this out. You have a pretty high dynamic range in the captured medium, and then you can use the tools you have to darken or lighten portions of the photograph when transferring it to paper.
Indeed so. Printing on paper and other substrates is inherently subtractive in nature which limits the gamut of colors and values that can be reproduced. Digital methods make the job of translating additive to subtractive media easier vs. the analog techniques available to film photographers. In any case, the image quality classic photography was able to achieve is truly remarkable.
Notably, the dodging and burning used by photographers aren't obsolete. There's a reason these tools are included in virtually every image-editing program out there. Manipulating dynamic range, particularly in printed images, remains part of the craft of image-making.
Contact printing on azo certainly helped!
> The claim that Ansel Adams used HDR is super likely to cause confusion
That isn't what the article claims. It says:
"Ansel Adams, one of the most revered photographers of the 20th century, was a master at capturing dramatic, high dynamic range scenes."
"Use HDR" (your term) is vague to the point of not meaning much of anything, but the article is clear that Adams was capturing scenes that had a high dynamic range, which is objectively true.
I think about the Ansel Adams zone system
https://www.kimhildebrand.com/how-to-use-the-zone-system/
where my interpretation is colored by the experience of making high quality prints and viewing them under different conditions, particularly poor illumination quality but you could also count "small handheld game console", "halftone screened and printed on newsprint" as other degraded conditions. In those cases you might imagine that the eye can only differentiate between 11 tones so even if an image has finer detail it ought to connect well with people if colors were quantized. (I think about concept art from Pokémon Sun and Moon which looked great printed with a thermal printer because it was designed to look great on a cheap screen.)
In my mind, the ideal image would look good quantized to 11 zones but also has interesting detail in texture in 9 of the zones (extreme white and black don't show texture). That's a bit of an oversimplification (maybe a shot outdoors in the snow is going to trend really bright, maybe for artistic reasons you want things to be really dark, ...) but Ansel Adams manually "tone mapped" his images using dodging, burning and similar techniques to make it so.
Literally the sentence preceding the one you quoted is “What if I told you that analog photographers captured HDR as far back as 1857?”.
And that quote specifically does not "lump HDR capture, HDR formats and HDR display together".
It is directly addressing capture.
Correct. I didn’t say that sentence was the source of the conflation, I said it was the source of the Ansel Adams problem. There are other parts that mix together capture, formats, and display.
Edit: and btw I am objecting to calling film capture “HDR”, I don’t think that helps define HDR nor reflects accurately on the history of the term.
That’s a strange claim because the first digital HDR capture devices were film scanners (for example the Cineon equipment used by the motion picture industry in the 1990s).
Film provided a higher dynamic range than digital sensors, and professionals wanted to capture that for image editing.
Sure, it wasn’t terribly deep HDR by today’s standards. Cineon used 10 bits per channel with the white point at coding value 685 (and a log color space). That’s still a lot more range and superwhite latitude than you got with standard 8-bpc YUV video.
They didn’t call that “HDR” at the time, and it wasn’t based on the idea of recording radiance or other absolute physical units.
I’m certain physicists had high range digital cameras before Cineon, and they were working in absolute physical metrics. That would be a stronger example.
You bring up an important point that is completely lost in the HDR discussion: this is about color resolution at least as much as it’s about range, if not moreso. I can use 10 bits for a [0..1] range just as easily as I can use 4 bits to represent quantized values from 0 to 10^9. Talking about the range of a scene captured is leaving out most of the story, and all of the important parts. We’ve had outdoor photography, high quality films, and the ability to control exposure for a long time, and that doesn’t explain what “HDR” is.
It was called "extended dynamic range" by ILM when they published the OpenEXR spec (2003):
> OpenEXR (www.openexr.net), its previously proprietary extended dynamic range image file format, to the open source community
https://web.archive.org/web/20170721234341/http://www.openex...
And "larger dynamic range" by Rea & Jeffrey (1990):
> With γ = 1 there is equal brightness resolution over the entire unsaturated image at the expense of a larger dynamic range within a given image. Finally, the automatic gain control, AGC, was disabled so that the input/output relation would be constant over the full range of scene luminances.
https://doi.org/10.1080/00994480.1990.10747942
I'm not sure when everyone settled on "high" rather than "large" or "extended", but certainly 'adjective dynamic range' is near-universal.
As I remember it, Paul Debevec had borrowed Greg Ward’s RGBE file format at some point in the late 90s and rebranded it “.hdr” for his image viewer tool (hdrView) and code to convert a stack of LDR exposures into HDR. I can see presentations online from Greg Ward in 2001 that have slides with “HDR” and “HDRI” all over the place. So yeah the term definitely must have started in the late 90s if not earlier. I’m not sure it was as there in the early 90s though.
Oo, interesting! That led me to this pair of sentences:
"Making global illumination user-friendly" (Ward, 1995) https://radsite.lbl.gov/radiance/papers/erw95.1/paper.html
> Variability is a qualitative setting that indicates how much light levels vary in the zone, i.e. the dynamic range of light landing on surfaces.
> By the nature of the situation being modeled, the user knows whether to expect a high degree of variability in the lighting or a low one.
Given those two phrases, 'a high or low degree of variability in the lighting' translates as 'a high or low degree of dynamic range' — or would be likely to, given human abbreviation tendencies, in successive works and conversations.
It certainly was called HDR when those Cineon files were processed in a linear light workflow. And film was the only capture source available that could provide sufficient dynamic range, so IMO that makes it “HDR”.
But I agree that the term is such a wide umbrella that almost anything qualifies. Fifteen years ago you could do a bit of superwhite glows and tone mapping on 8-bpc and people called that look HDR.
Do you have any links from 1990ish that show use of “HDR”? I am interested in when “HDR” became a phrase people used. I believe I remember hearing it first around 1996 or 97, but it may have started earlier. It was certainly common by 2001. I don’t see that used as a term nor an acronym in the Cineon docs from 1995, but it does talk about log and linear spaces and limiting the dynamic range when converting. The Cineon scanner predates sRGB, and used gamma 1.7. https://dotcsw.com/doc/cineon1.pdf
This 10 bit scanner gave you headroom of like 30% above white. So yeah it qualifies as a type of high dynamic range when compared to 8 bit/channel RGB, but on the other hand, a range of [0 .. 1.3] isn’t exactly in the spirit of what “HDR” stands for. The term implicitly means a lot more than 1.0, not just a little. And again people developing HDR like Greg Ward and Paul Debevec were arguing for absolute units such as luminance, which the Cineon scanner does not do.
Yes, Ansel Adams was using a camera to capture a scene that had high dynamic range.
I don't see the confusion here.
HDR is not referring to the scene’s range, and it doesn’t apply to film. It’s referring superficially but specifically to a digital process that improves on 8 bits/channel RGB images. And one of the original intents behind HDR was to capture pixels in absolute physical measurements like radiance, to enable a variety of post-processing workflows that are not available to film.
"High dynamic range" is a phrase that is much older than tone mapping. I see uses of "dynamic range" going back to the 1920s and "high dynamic range" to the 1940s:
https://books.google.com/ngrams/graph?content=dynamic+range%...
You might argue that "HDR" the abbreviation refers to using tone mapping to approximate rendering high dynamic range imagery on lower dynamic range displays. But even then, the sentence in question doesn't use the abbreviation. It is specifically talking about a dynamic range that is high.
Dynamic range is a property of any signal or quantifiable input, including, say sound pressure hitting our ears or photons hitting an eyeball, film, or sensor.
> But even then, the sentence in question doesn’t use the abbreviation
Yes it does. Why are you still looking at a different sentence than the one I quoted??
HDR in this context isn’t referring to just any dynamic range. If it was, then it would be so vague as to be meaningless.
Tone mapping is closely related to HDR and very often used, but is not necessary and does not define HDR. To me it seems like your argument is straw man. Photographers have never broadly used the term “high dynamic range” as a phrase, nor the acronym “HDR” before it showed up in computer apps like hdrView, Photoshop, and iPhone camera.
Oh, sorry, you're right. Mentioning the abbreviation is a red herring. The full quote is:
"But what if we don't need that tradeoff? What if I told you that analog photographers captured HDR as far back as 1857? Ansel Adams, one of the most revered photographers of the 20th century, was a master at capturing dramatic, high dynamic range scenes. It's even more incredible that this was done on paper, which has even less dynamic range than computer screens!"
It seems pretty clear to me that in this context the author is referring to the high dynamic range of the scenes that Adams pointed his camera at. That's why he says "captured HDR" and "high dynamic range scenes".
> It seems pretty clear to me that in this context the author is referring to the high dynamic range of the scenes that Adams pointed his camera at.
Yes, this is the problem I have with the article. “HDR” is not characterized solely by the range of the scene, and never was. It’s a term of art that refers to an increased range (and resolution) on the capture and storage side, and it’s referring to a workflow that involves/enables deferring exposure until display time. The author’s claim here is making the term “HDR” harder to understand, not easier, and it’s leaving out of the most important conceptual aspects. There are some important parallels between film and digital HDR, and there are some important differences. The differences are what make claiming that nineteenth century photographers were capturing HDR problematic and inaccurate.
To further complicate the issue, "high dynamic range" is a phrase that will come up across a few different disciplines, not just related to the capture & reproduction of visual data.
The digital process of tonemapping, aka. 'what Apple calls Smart HDR processing of SDR photos to increase perceptual dynamic range', can be applied to images of any number of channels of any bit depth — though, if you want to tonemap a HyperCard dithered black-and-white image, you'll probably have to decompile the dithering as part of creating the gradient map. Neither RGB nor 8-bit are necessary to make tonemapping a valuable step in image processing.
That’s true, and it’s why tonemapping is distinct from HDR. If you follow the link from @xeonmc’s comment and read the comments, the discussion centers on the conflation of tonemapping and HDR.
https://news.ycombinator.com/item?id=43987923
That said, the entire reason that tonemapping is a thing, and the primary focus of the tonemapping literature, is to solve the problem of squeezing images with very wide ranges into narrow display ranges like print and non-HDR displays, and to achieve a natural look that mirrors human perception of wide ranges. Tonemapping might be technically independent of HDR, but they did co-evolve, and that’s part of the history.
> It seems like a mistake to lump HDR capture, HDR formats and HDR display together, these are very different things.
These are all related things. When you talk about color, you can be talking about color cameras, color image formats, and color screens, but the concept of color transcends the implementation.
> The claim that Ansel Adams used HDR is super likely to cause confusion, and isn’t particularly accurate.
The post never said Adams used HDR. I very carefully chose the words, "capturing dramatic, high dynamic range scenes."
> Previously when you took a photo, if you over-exposed it or under-exposed it, you were stuck with what you got. Capturing HDR gives the photographer one degree of extra freedom, allowing them to adjust exposure after the fact.
This is just factually wrong. Film negatives have 12-stops of useful dynamic range, while photo paper has 8 stops at best. That gave photographers exposure latitude during the print process.
> Ansel Adams wasn’t using HDR in the same sense we’re talking about, he was just really good at capturing the right exposure for his medium without needing to adjust it later.
There's a photo of Ansel Adams in the article, dodging and burning a print. How would you describe that if not adjusting the exposure?
> Film negatives have 12-stops of useful dynamic range
No, that’s not inherently true. AA used 12 zones, that doesn’t mean every negative stock has 12 stops of latitude. Stocks are different, you need to look at the curves.
But yes most modern negatives are very forgiving. FP4 for example has barely any shoulder at all iirc.
I agree capture, format and display are closely related. But HDR capture and processing specifically developed outside of HDR display devices, and use of HDR displays changes how HDR images are used compared to LDR displays.
> The post never said Adams used HDR. I very carefully chose the words
Hey I’m sorry for criticizing, but I honestly feel like you’re being slightly misleading here. The sentence “What if I told you that analog photographers captured HDR as far back as 1857?” is explicitly claiming that analog photographers use “HDR” capture, and the Ansel Adams sentence that follows appears to be merely a specific example of your claim. The result of the juxtaposition is that the article did in fact claim Adams used HDR, even if you didn’t quite intend to.
I think you’re either misunderstanding me a little, or maybe unaware of some of the context of HDR and its development as a term of art in the computer graphics community. Film’s 12 stops is not really “high” range by HDR standards, and a little exposure latitude isn’t where “HDR” came from. The more important part of HDR was the intent to push toward absolute physical units like luminance. That doesn’t just enable deferred exposure, it enables physical and perceptual processing in ways that aren’t possible with film. It enables calibrated integration with CG simulation that isn’t possible with film. And it enables a much wider rage of exposure push/pull than you can do when going from 12 stops to 8. And of course non-destructive digital deferred exposure at display time is quite different from a print exposure.
Perhaps it’s useful to reflect on the fact that HDR has a counterpart called LDR that’s referring to 8 bits/channel RGB. With analog photography, there is no LDR, thus zero reason to invent the notion of a ‘higher’ range. Higher than what? High relative to what? Analog cameras have exposure control and thus can capture any range you want. There is no ‘high’ range in analog photos, there’s just range. HDR was invented to push against and evolve beyond the de-facto digital practices of the 70s-90s, it is not a statement about what range can be captured by a camera.
> The sentence “What if I told you that analog photographers captured HDR as far back as 1857?” is explicitly claiming that analog photographers use “HDR” capture,
No, it isn't. It's saying they captured HDR scenes.
> The result of the juxtaposition is that the article did in fact claim Adams used HDR
You can't "use" HDR. It's an adjective, not a noun.
> Film’s 12 stops is not really “high” range by HDR standards, and a little exposure latitude isn’t where “HDR” came from.
The Reinhard tone mapper, a benchmark that regularly appears in research papers, specifically cites Ansel Adams as inspiration.
"A classic photographic task is the mapping of the potentially high dynamic range of real world luminances to the low dynamic range of the photographic print."
https://www-old.cs.utah.edu/docs/techreports/2002/pdf/UUCS-0...
> Perhaps it’s useful to reflect on the fact that HDR has a counterpart called LDR that’s referring to 8 bits/channel RGB.
8-bits per channel does not describe dynamic range. If I attach an HLG transfer function on an 8-bit signal, I have HDR. Furthermore, assuming you actually meant 8-bit sRGB, nobody calls that "LDR." It's SDR.
> Analog cameras have exposure control and thus can capture any range you want.
This sentence makes no sense.
Sorry man, you seem really defensive, I didn’t mean to put you on edge. Okay, if you are calling the scenes “HDR” then I’m happy to rescind my critique about Ansel Adams and switch instead to pointing out that “HDR” doesn’t refer to the range of the scene, it refers to the range capability of a digital capture process. I think the point ultimately ends up being the same either way. Hey where is HDR defined as an adjective? Last time I checked, “range” could be a noun, I think… no? You must be right, but FWIW, you used HDR as a noun in your 2nd to last point… oh and in the title of your article too.
Hey it’s great Reinhard was inspired by Adams. I have been too, like a lot of photographers. And I’ve used the Reinhard tone mapper in research papers, I’m quite familiar with it and personally know all three authors of that paper. I’ve even written a paper or maybe two on color spaces with one of them. Anyway, the inspiration doesn’t change the fact that 12 stops isn’t particularly high dynamic range. It’s barely more than SDR. Even the earliest HDR formats had like 20 or 30 stops, in part because the point was to use physical luminance instead of a relative [0..1] range.
8 bit RGB does sort-of in practice describe a dynamic range, as long as the 1 bit difference is approximately the ‘just noticeable difference’ or JND as some researchers call it. This happens to line up with 8 bits being about 8 stops, which is what RGB images have been doing for like 50 years, give or take. While it’s perfectly valid arithmetic to use 8 bits values to represent an arbitrary amount like 200 stops or 0.003 stops, it’d be pretty weird.
Plenty of people have called and continue to call 8 bit images “LDR”, here’s just three of the thousands of uses of “LDR” [1][2][3], and LDR predates usage of SDR by like 15 years maybe? LDR predates sRGB too, I did not actually mean 8 bit sRGB. LDR and SDR are close but not quite the same thing, so feel free to read up on LDR. It’s disappointing you ducked the actual point I was making, which is still there even if you replace LDR with SDR.
What is confusing about the sentence about analog cameras and exposure control? I’m happy to explain it since you didn’t get it. I was referring to how the aperture can be adjusted on an analog camera to make a scene with any dynamic range fit into the ~12 stops of range the film has, or the ~8 stops of range of paper or an old TV. I was just trying to clarify why HDR is an attribute of digital images, and not of scenes.
[1] https://www.easypano.com/showkb_228.html#:~:text=The%20Dynam...
[2] https://www.researchgate.net/figure/shows-digital-photograph...
[3] https://irisldr.github.io/
You opened this thread arguing that Ansel Adams didn't "use HDR." I linked you to a seminal research paper which argues that he tone mapped HDR content, and goes on to implement a tone mapper based on his approach. This all seems open and shut.
> I’m happy to rescind my critique about Ansel Adams
Great, I'm done.
> and switch instead to pointing out that “HDR” doesn’t refer to the range of the scene
Oh god. Here's the first research paper that popped into my head: https://static.googleusercontent.com/media/hdrplusdata.org/e...
"Surprisingly, daytime shots with high dynamic range may also suffer from lack of light."
"In low light, or in very high dynamic range scenes"
"For high dynamic range scenes we use local tone mapping"
You keep trying to define "HDR" differently than current literature. Not even current— that paper was published in 2016! Hey, maybe HDR meant something different in the 1990s, or maybe it was just ok to use "HDR" as shorthand for when things were less ambiguous. I honestly don't care, and you're only serving to confuse people.
> the aperture can be adjusted on an analog camera to make a scene with any dynamic range fit into the ~12 stops of range the film has, or the ~8 stops of range of paper or an old TV.
You sound nonsensical because you keep using the wrong terms. Going back to your first sentence that made no sense:
> Analog cameras have exposure control and thus can capture any range you want
You keep saying "range" when, from what I can tell, you mean "luminance." Changing a camera's aperture scales the luminance hitting your film or sensor. It does not alter the dynamic range of the scene.
Analog cameras cannot capture any range. By adjusting camera settings or attaching ND filters, you can change the window of luminance values that will fit within the dynamic range of your camera. To say a camera can "capture any range" is like saying, "I can fit that couch through the door, I just have to saw it in half."
> And I’ve used the Reinhard tone mapper in research papers, I’m quite familiar with it and personally know all three authors of that paper. I’ve even written a paper or maybe two on color spaces with one of them.
I'm sorry if correcting you triggers insecurities, but if you're going to make an appeal to authority, please link to your papers instead of hand waving about the people you know.
Hehe outside is “HDR content”? To me that still comes off as confused about what HDR is. I know you aren’t, but that’s what it sounds like. A sunny day has a high dynamic range for sure, but the acronym HDR is a term of art that implies more than that. Your article even explains why.
Tone mapping doesn’t imply HDR. Tone mapping is always present, even in LDR and SDR workflows. The paper you cited explicitly notes the idea is to “extend” Adams’ zone system to very high dynamic range digital images, more than what Adams was working with, by implication.
So how is a “window of luminance values” different from a dynamic range, exactly? Why did you make the incorrect and obviously silly assumption that I was suggesting a camera’s aperture changes the outdoor scene’s dynamic range rather than what I actually said, that it changes the exposure? Your description of what a camera does is functionally identical. I’m kinda baffled as to why you’re arguing this part that we both understand, using hyperbole.
I hope you have a better day tomorrow. Good luck with your app. This convo aside, I am honestly rooting for you.
> Hehe outside is “HDR content”? To me that still comes off as confused about what HDR is.
"Surprisingly, daytime shots with high dynamic range may also suffer from lack of light."
That's from, "Burst photography for high dynamic range and low-light imaging on mobile cameras," written by some of the most respected researchers in computational photography. It has 342 citations according to ACM.
I'm still waiting for a link to your papers.
> Tone mapping doesn’t imply HDR.
https://en.wikipedia.org/wiki/Tone_mapping
First sentence: "Tone mapping is a technique used in image processing and computer graphics to map one set of colors to another to approximate the appearance of high-dynamic-range (HDR) images in a medium that has a more limited dynamic range."
> Why did you make the incorrect and obviously silly assumption that I was suggesting a camera’s aperture changes the outdoor scene’s dynamic range rather than what I actually said, that it changes the exposure?
Because you keep bumbling details like someone with a surface level understanding. Your replies are irrelevant, outdated, or flat out wrong. It all gives me flashbacks to working under engineers-turned-managers who just can't let go, forcing their irrelevant backgrounds into discussions.
It's cool that you studied late 90s 3D rendering. So did I. It doesn't make you an expert in computational photography. Please stop confusing people with your non-sequiturs.
What does the lack of light quote prove? That’s a statement about color resolution, not range, and it uses “high dynamic range” and not “HDR content”. I think you’ve missed my point and are not listening.
Yes tone mapping is used on HDR images. It just doesn’t imply HDR. SDR gamma is tone mapping, for example, which the Wikipedia link you sent explains. Your claim is that Adams use of tone mapping is evidence that he is capturing “HDR content”. The paper you sent doesn’t use that language, it doesn’t ever say Adams was doing tone mapping, it says they develop a tone mapping method inspired by Adams’ zone system that extends the idea into higher dynamic range.
You’re using your own misunderstanding and mis-interpretation of my comments as evidence that they’re wrong. Hey I totally might be wrong about a lot of things, and sure maybe I’m completely non-sensical, but you certainly haven’t convinced me of that. I haven’t had trouble speaking with other people about HDR imaging, people who are HDR experts. All I’m getting out of this so far is that some people react very badly to any hint of critique.
From my perspective, I’m also only hearing bumbling errors, errors like that HDR is an adjective, that LDR doesn’t exist and nobody uses it, that using “range” is incorrect when I say it but not when you do and “window of luminance values” is better, and that Ansel Adams was doing HDR imaging.
Ben, we’re having a bona-fide miscommunication, and I wanted to fix it but I’m failing, and it feels like you’re determined not to fix it or find any common ground. In another environment we’d probably be having a friendly, productive and enlightening conversation. I’m sure there are some things I could learn from you.
But the article even shows Adams dodging/burning a print, which is 'adjusting the exposure' in a localised fashion of the high dynamic range of the film, effectively revealing detail for the LDR of the resulting print that otherwise wouldn't have been visible.
> It seems like a mistake to lump HDR capture, HDR formats and HDR display together
Reminded me of the classic "HDR in games vs HDR in photography" comparison[0]
[0] https://www.realtimerendering.com/blog/thought-for-the-day/
If I look at one of the photography books in my shelf, they are even talking about 18 stops and such for some film material, and how this doesn't translate to paper and all the things that can be done to render it visible in print and how things behave at both extreme ends (towards black and white). Read: Tone-mapping (i.e. trimming down a high DR image to a lower DR output media) is really old.
The good thing about digital is that it can deal with color at decent tonal resolutions (if we assume 16 bits, not the limited 14 bit or even less) and in environments where film has technical limitations.
No, Adams, like everyone who develops their own film (or RAW digital photos) definitely worked in HDR. Film has much more DR than photographic paper, as noted by TFA author (and large digital sensors more than either SDR or HDR displays) especially if you’re such a master of exposure as Adams; trying to preserve the tonalities when developing and printing your photos is the real big issue.
Is there a difference in capturing in HDR vs RAW?
Good question. I think it depends. They are kind of different concepts, but in practice they can overlap considerably. RAW is about using the camera’s full native color resolution, and not having lossy compression. HDR is overloaded, as you can see from the article & comments, but I think HDR capture is conceptually about expressing brightness in physical units like luminance or radiance, and delaying the ‘exposure’ until display time. Both RAW and HDR typically mean using more than 8 bits/channel and capturing high quality images that will withstand more post-processing than ‘exposed’ LDR images can handle.
HDR on displays is actually largely uncomfortable for me. They should reserve the brightest HDR whites for things like the sun itself and caustics, not white walls in indoor photos.
As for tone mapping, I think the examples they show tend way too much towards flat low-local-contrast for my tastes.
HDR is really hard to get right apparently. It seems to get worse in video games too.
I'm a huge fan of Helldivers 2, but playing the game in HDR gives me a headache: the muzzle flash of weapons at high RPMs on a screen that goes to 240hz is basically a continuous flashbang for my eyes.
For a while, No Mans' Sky in HDR mode was basically the color saturation of every planet dialed up to 11.
The only game I've enjoyed at HDR was a port from a console, Returnal. The use of HDR brights was minimalistic and tasteful, often reserved for certain particle effects.
For a year or two after it launched, The Division 2 was a really, really good example of HDR done right. The game had (has?) a day/night cycle, and it had a really good control of the brightness throughout the day. More importantly, it made very good use of the wide color gamut available to it.
I stopped playing that game for several years, and when I went back to it, the color and brightness had been wrecked to all hell. I have heard that it's received wisdom that gamers complain that HDR modes are "too dark", so perhaps that's part of why they ruined their game's renderer.
Some games that I think currently have good HDR:
* Lies of P
* Hunt: Showdown 1896
* Monster Hunter: World (if you increase the game's color saturation a bit from its default settings)
Some games that had decent-to-good HDR the last time I played them, a few years ago:
* Battlefield 1
* Battlefield V
* Battlefield 2042 (If you're looking for a fun game, I do NOT recommend this one. Also, the previous two are probably chock-full of cheaters these days.)
I found Helldivers 2's HDR mode to have blacks that were WAY too bright. In SDR mode, nighttime in forest areas was dark. In HDR mode? It was as if you were standing in the middle of a field during a full moon.
> I have heard that it's received wisdom that gamers complain that HDR modes are "too dark", so perhaps that's part of why they ruined their game's renderer.
A lot of people have cheap panels that claim HDR support (read: can display an HDR signal) but have garbage color space coverage, no local dimming, etc. and to them, HDR ends up looking muted.
For what it's worth, I have a monitor with VA panel that has no local dimming. It has a maximum brightness of a bit shy of 380 nits, and a real contrast ratio of ~3000:1, so HDR stuff ends up looking fine in dim rooms.
I would much rather have this monitor than one the "zoned" backlights that I've seen. They inevitably put nasty, nasty halos around medium-to-high-contrast parts of the picture. IME, any local dimming scheme that's less fine-grained than per-pixel dimming is simply not good enough.
It's like the bloom menace all over again - all new rendering tech gets overused so you can't miss it before developers learn to restrain themselves.
A lot of this is poor QA. When you start to do clever things like HDR you have to test on a bunch of properly calibrated devices of different vendors etc. And if you're targeting Windows you have to accept that HDR is a mess for consumers and even if their display supports, their GPU supports it, they might still have the drivers and color profiles misconfigured. (and many apps are doing it wrong or weird, even when they say they support it)
Also (mostly) on Windows, or on videos for your TV: a lot of cheap displays that say they are HDR are a range of hot garbage.
Most "HDR" monitors are junk that can't display HDR. The HDR formats/signals are designed for brightness levels and viewing conditions that nobody uses.
The end result is a complete chaos. Every piece of the pipeline doing something wrong, and then the software tries to compensate for it by emitting doubly wrong data, without even having reliable information about what it needs to compensate for.
https://docs.google.com/document/d/1A__vvTDKXt4qcuCcSN-vLzcQ...
What we really need is some standards that everybody follows. The reason normal displays work so well is that everyone settled on sRGB, and as long as a display gets close to that, say 95% sRGB, everyone except maybe a few graphics designers will have a n equivalent experience.
But HDR, it's a minefield of different display qualities, color spaces, standards. It's no wonder that nobody gets it right and everyone feels confused.
HDR on a display that has peak brightness of 2000 nits will look completely different than a display with 800 nits, and they both get to claim they are HDR.
We should have a standard equivalent to color spaces. Set, say, 2000 nits as 100% of HDR. Then a 2000 nit display gets to claim it's 100% HDR. A 800 nit display gets to claim 40% HDR, etc. A 2500 nit display could even use 125% HDR in it's marketing.
It's still not perfect - some displays (OLED) can only show peak brightness over a portion of the screen. But it would be an improvement.
DisplayHDR standard is supposed to be it, but they've ruined its reputation by allowing HDR400 to exist when HDR1000 should have been the minimum.
Besides, HDR quality is more complex than just max nits, because it depends on viewing conditions and black levels (and everyone cheats with their contrast metrics).
OLEDs can peak at 600 nits and look awesome — in a pitch black room. LCD monitors could boost to 2000 nits and display white on grey.
We have sRGB kinda working for color primaries and gamma, but it's not the real sRGB at 80 nits. It ended up being relative instead of absolute.
A lot of the mess is caused by the need to adapt content mastered for pitch black cinema at 2000 nits to 800-1000 nits in daylight, which needs very careful processing to preserve highlights and saturation, but software can't rely on the display doing it properly, and doing it in software sends false signal and risks display correcting it twice.
There's a pretty good video on YouTube (more than one, actually) that explains how careless use of HDR in modern cinema is destroying the look and feel of cinema we used to like.
Everything is flattened, contrast is eliminated, lights that should be "burned white" for a cinematic feel are brought back to "reasonable" brightness with HDR, really deep blacks are turned into flat greys, etc. The end result is the flat and washed out look of movies like Wicked. It's often correlated to CGI-heavy movies, but in reality it's starting to affect every movie.
The washed out grey thing was an error that became a style!
Because HDR wasn’t natively supported on most displays and software, for a long time it was just “hacked in there” by squashing the larger dynamic range into a smaller one using a mathematical transform, usually a log function. When viewed without the inverse transform this looks horribly grey and unsaturated.
Directors and editors would see this aesthetic day in, day out, with the final color grade applied only after a long review process.
Some of them got used to it and even liking it, and now here we are: horribly washed out movies made to look like that on purpose.
The transition from Avengers to the later movies is very noticeable, and one of the worst offenders since source material really speaks against the choice.
What you said, it's definitely become a style. But, also, a lot of these movies that look like ass on Joe Public's OOGLAMG $130 85" Black Friday TV in his brightly-lit living room actually look awesome if your entire setup is proper, real HDR devices and software, your screen has proper OLED or local dimming, is calibrated to within an inch of its life etc, and you view them in a dark home theater.
True! But I think that movies adapted for TV must be made for the average "good quality" screen, not the state of the art that almost nobody owns. At least they should look decent enough in a good quality (but not top-notch) setup.
Also, the YouTube video I'm thinking of singles out Wicked as seen in movie theaters. The image "as intended" looks washed out and without contrast.
I've found this especially a problem with those AI systems trying to add HDR to existing images/videos. The worst instance I've seen, was playing one of the recent Spongebob platformer games, and having his eyes glow like giant suns in the menu screen. I have a TV capable of a fairly high maximum brightness, and it was dimming the rest of the image just to make sure Spongebob's eyes lit up my living room like it was midday
It feels like to some photographers/cinematographers/game designers, HDR is a gimmick to make something look more splashy/eye catching. The article touches on this a bit, with some of the 2000s HDR examples in photography. With the rise of HDR TVs, it feels like that trend is just happening again.
HDR is when you’re watching a dark film at night, looking at the subtle nuances between shades of dark and black in the shadows on the screen, making out the faint contours the film director carefully curated, and the subtitles gently deposit 40W of light into your optical nerves with “♪”.
I actually kind of hate the fact that the subtitles on my (SDR) OLED display never get as bright as some parts of the video itself.
As a photographer, I get the appeal of (this new incarnation of) HDR content, but the practical reality is that the photos I see posted in my feeds go from making my display looking normal to having photos searing my retinas, while other content that was uniform white a second prior now looks dull gray.
It's late night here so I was reading this article in dark mode, at a low display brightness - and when I got to the HDR photos I had to turn down my display even more to not strain my eyes, then back up again when I scrolled to the text.
For fullscreen content (games, movies) HDR is alright, but for everyday computing it's a pretty jarring experience as a user.
While it isn't touched on in the post, I think the issue with feeds is that platforms like Instagram have no interest in moderating HDR.
For context: YouTube automatically edits the volume of videos that have an average loudness beyond a certain threshold. I think the solution for HDR is similar penalization based on log luminance or some other reasonable metric.
I don't see this happening on Instagram any time soon, because bad HDR likely makes view counts go up.
As for the HDR photos in the post, well, those are a bit strong to show what HDR can do. That's why the Mark III beta includes a much tamer HDR grade.
> YouTube automatically edits the volume of videos that have an average loudness beyond a certain threshold.
For anyone else who was confused by this, it seems to be a client-side audio compressor feature (not a server-side adjustment) labeled as "Stable Volume". On the web, it's toggleable via the player settings menu.
https://support.google.com/youtube/answer/14106294
I can't find exactly when it appeared but the earliest capture of the help article was from May 2024, so it is a relatively recent feature: https://web.archive.org/web/20240523021242/https://support.g...
I didn't realize this was a thing until just now, but I'm glad they added it because (now that I think about it) it's been awhile since I felt the need to adjust my system volume when a video was too quiet even at 100% player volume. It's a nice little enhancement.
Youtube has been long normalizing videos standard feed, switching to a -14 LUFS target in 2019. But LUFS is a global target, and is meant to allow higher peaks and troughs over the whole video, and the normalization does happen on a global level - if you exceed it by 3dB, then the whole video gets it's volume lowered by 3dB, no matter if the part is quiet or not.
The stable volume thing is meant to essentially level out all of the peaks and troughs, and IIRC it's actually computed server-side, I think yt-dlp can download stable volume streams if asked to.
The client side toggle might be new since 2024 but the volume normalisation has been a thing for a long time.
I know they've automatically boosted brightness in dark scenes for a long time too. It's not rare for people to upload a clip from a video game with a very dark scene and it's way brighter after upload than it was when they played or how it looks in the file they uploaded.
Yes, and I love it! Finally, the volume knob control was pulled from producers all about gimmicks to push their productions.
There are still gimmicks, but at least they do not include music so badly clipped as to be unlistenable... hint: go get the DVD or Blu-Ray release of whatever it is and you are likely to enjoy a not clipped album.
It is all about maximizing the overall sonic impact the music is capable of. Now when levels are sane, song elements well differentiated and equalized such that no or only a minor range of frequencies are crushed due to many sounds all competing for them, it will sound, full, great and not tiring!
Thanks audio industry. Many ears appreciate what was done.
I expected a moderate amount of heat directed at my comment.
No worries. I've friends in various industries doing production who hate the change.
I like it, of course. Losing the volume knob is a direct result of the many abuses.
Instagram has to allow HDR for the same reason that Firefox spent the past twenty years displaying web colors like HN orange at maximum display gamut rather than at sRGB calibrated: because a brighter red than anyone else’s draws people in, and makes the competition seem lifeless by comparison, especially in a mixed-profiles environment. Eventually that is regarded as ‘garishly bright’, so to speak, and people push back against it. I assume Firefox is already fixing this to support the latest CSS color spec (which defines #rrggbb as sRGB and requires it to be presented as such unless stated otherwise in CSS), but I doubt Instagram is willing to literally dim their feed; instead, I would expect them to begin AI-HDR’ing SDR uploads in order that all videos are captivatingly, garishly, bright.
> I think the solution for HDR is similar penalization based on log luminance or some other reasonable metric.
I completely understand the desire to address the issue of content authors misusing or intentionally abusing HDR with some kind of auto-limiting algorithm similar to the way the radio 'loudness wars' were addressed. Unfortunately, I suspect it will be difficult, if not impossible, to achieve without also negatively impacting some content applying HDR correctly for artistically expressive purposes. Static photos may be solvable without excessive false positive over-correction but cinematic video is much more challenging due to the dynamic nature of the content.
As a cinemaphile, I'm starting to wonder if maybe HDR on mobile devices simply isn't a solvable problem in practice. While I think it's solvable technically and certainly addressable from a standards perspective, the reality of having so many stakeholders in the mobile ecosystem (hardware, OS, app, content distributors, original creators) with diverging priorities makes whatever we do from a base technology and standards perspective unlikely to work in practice for most users. Maybe I'm too pessimistic but as a high-end home theater enthusiast I'm continually dismayed how hard it is to correctly display diverse HDR content from different distribution sources in a less complex ecosystem where the stakeholders are more aligned and the leading standards bodies have been around for many decades (SMPTE et al).
I believe everything could be solved the same way we solved high dynamic range in audio, with a volume control.
I find it pretty weird that all tvs and most monitors hide the brightness adjustment under piles and piles of menus when it could be right there in the remote alongside the sound volume buttons. Maybe phones could have hardware brightness buttons too, at least something as easy as it is on adjusting brightness in notebooks that have dedicated brightness fn buttons.
Such brightness slider could also control the amount of tonemapping applied to HDR content. High brightness would mean no to low tonemapping and low brightness would use a very agressive tonemapper producing a similar image to the SDR content along it.
Also note that good audio volume attenuation requires proper loudness contour compensation (as you lower the volume you also increase the bass and treble) for things to sound reasonably good and the "tone" sound well balanced. So, adjusting the tonemapping based on the brightness isn't that far off what we do with audio.
> because bad HDR likely makes view counts go up
Another related parallel trend recently is that bad AI images get very high view and like counts, so much so that I've lost a lot of motivation for doing real photography because the platforms cease to show them to anyone, even my own followers.
Why is nobody talking about the standards development? They (OS, image formats) could just say all stuff by default assumes SDR and if a media file explicitly calls for HDR even then it cannot have sharp transitions except in special cases, and the software just blocks or truncates any non conforming images. The OS should have had something like this for sound, about 25-30 years ago. For example a brightness aware OS/monitor combo could just outright disallow anything about x nits. And disallow certain contrast levels, in the majority of content.
FYI: You wrote Chrome 14 in the post, but I believe you meant Android 14.
Thanks. Updated.
Btw, YouTube doesn't moderate HDR either. I saw one video of a child's violin recital that was insanely bright, and probably just by accident of using a bad HDR recorder.
The effect of HDR increasing views is explicitly mentioned in the article
You are replying to the article's author.
Completely agree. To me, HDR feels like the system is ignoring my screen brightness settings.
I set my screen brightness to a certain level for a reason. Please don’t just arbitrarily turn up the brightness!
There is no good way to disable HDR on photos for iPhone, either. Sure, you can turn off the HDR on photos on your iphone. But then, when you cast to a different display, the TV tries to display the photos in HDR, and it won’t look half as good.
> To me, HDR feels like the system is ignoring my screen brightness settings.
You might be on to something there. Technically, HDR is mostly about profile signaling and therefore about interop. To support it in mpeg dash or hls media you need to make sure certain codec attributes are mentioned in the xml or m3u8 but the actual media payload stays the same.
Any bit or Bob being misconfigured or misinterpreted in the streaming pipeline will result in problems ranging from slightly suboptimal experience to nothing works.
Besides HDR, "spatial audio" formats like Dolby Atmos are notorious for interop isuues
> To me, HDR feels like the system is ignoring my screen brightness settings.
On both Android & iOS/MacOS it's not that HDR is ignoring your screen brightness, but rather the brightness slider is controlling the SDR range and then yes HDR can exceed that, that's the singular purpose of HDR to be honest. All the other purported benefits of HDR are at best just about HDR video profiles and at worst just nonsense bullshit. The only thing HDR actually does is allow for brighter colors vs. SDR. When used selectively this really enhances a scene. But restraint is hard, and most forms of HDR content production are shit. The HDR images that newer iPhones and Pixel phones are capturing are generally quite good because they are actually restrained, but then ironically both of them have horrible HDR video that's just obnoxiously bright.
"On both Android & iOS/MacOS it's not that HDR is ignoring your screen brightness, but rather the brightness slider is controlling the SDR range and then yes HDR can exceed that"
Doesn't this mean HDR is ignoring my brightness setting? Looking at the Mac color profiles, the default HDR has some fixed max brightness regardless of the brightness slider. And it's very bright, 1600 nits vs the SDR max of 600 nits. At least I was able to pick another option capping HDR to 600, but that still allows HDR video to force my screen to its normal full brightness even if I dimmed it.
> Doesn't this mean HDR is ignoring my brightness setting?
Not exactly because it is still being scaled by your brightness setting. As in, start playing an HDR video and then mess with the brightness slider. You will still see the HDR content getting dimmer/brighter.
It's easier to think about in Apple's EDR terms. 0.0-1.0 is the SDR range, and the brightness slider is changing what the nit value is of "1.0" - is it 100 nits? 300 nits? 50 nits? etc... HDR content (in theory) still has that same 0.0-1.0 portion of the range, and it's still being scaled. However it can exceed that 1.0. It's still being scaled, it's still "respecting" that slider. Just the slider wasn't a brightness limit as you're wanting it to be, but a 1.0 alignment point.
The problem comes when HDR content is disrespectful to that. When it just absolutely slams the brightness, pushing all of its content way past that 1.0 value. This is bad content, and unfortunately it's incredibly common in HDR media due in part to the fact that the original HDR specs are very incomplete and in part because it's a new loudness war.
Ah, just tested with https://github.com/dtinth/superwhite and you're correct. I remember it not even scaling with the brightness, but guess I was wrong.
>HDR can exceed that
It's not just the HDR content that gets brighter, but SDR content too. When I test it in Chrome on Android, if an HDR image shows up on screen the phone start overriding the brightness slider completely and making everything brighter, including the phone's system UI.
>The only thing HDR actually does is allow for brighter colors vs. SDR.
Not just brighter, but also darker, so it can preserve detail in dark areas rather than crushing them.
> It's not just the HDR content that gets brighter, but SDR content too. When I test it in Chrome on Android, if an HDR image shows up on screen the phone start overriding the brightness slider completely and making everything brighter, including the phone's system UI.
You have an "old" style handling of HDR on Android. Newer/better devices don't do that (specifically those that support https://source.android.com/docs/core/display/mixed-sdr-hdr )
Similarly MacOS/iOS doesn't do that.
> Not just brighter, but also darker, so it can preserve detail in dark areas rather than crushing them.
It does not get darker, and while PQ allocates more bits to the dark region HLG does not. And, more importantly, neither does the actual display panel which are still typically gamma 2.2-2.4 regardless. So PQ's extra precision in the dark areas is ~never utilized other than as tonemapping input, but the resulting output does not have any increased precision in the darks over SDR.
In fact it actually has less precision in the dark areas as the increased display luminance range means the panels native bit depth need to cover more range.
you are right but at least in my experience it's very easy for a modern iPhone to capture a bad HDR photo, usually because there is some small strong highlight (often a form of specular reflection from a metallic object) that causes everything to be HDR while the photo content wouldn't need it
In beta testing this morning, the Halide “HDR slider” works as intended to solve that. Some of my photos have only needed +0.3 while a couple taken in near-total no-light-pollution darkness have it cranked all the way to max and that still isn’t enough.
It isn't just about the brightness (range).
In practice the 'HDR' standards are also about wider color gamuts (than sRGB), and (as mentioned in parallel) packed into more bits, in a different way, so as to minimise banding while keeping file sizes in check.
I’m under the impression this is caused by the use of “HDR mode”(s) and poor adaptive brightness implementations on devices. Displays such as the iPad Pro w/ OLED are phenomenal and don’t seem to implement an overactive adaptive brightness. HDR content has more depth without causing brightness distortion.
In contrast, my TV will change brightness modes to display HDR content and disables some of the brightness adjustments when displaying HDR content. It can be very uncomfortably bright in a dark room while being excessively dim in a bright room. It requires adjusting settings to a middle ground resulting in a mixed/mediocre experience overall. My wife’s laptop is the worst of all our devices, while reviews seem to praise the display, it has an overreactive adaptive brightness that cannot be disabled (along with decent G2G response but awful B2W/W2B response that causes ghosting).
Apple’s method involves a good deal of what they call “EDR”, wherein the display gamma is ramped down in concert with ramping the brightness up, so that the brighter areas get brighter while the non-bright areas remain dark due to gamma math; that term is helpful for searching their WWDC developer videos for more details.
It still looks like how they say where its way too bright and makes formerly “bright” whites appear even neutral grey surrounding the hdr content. Personally I find it extremely jarring when it is just a frame that is hdr within the overall screen. It is much nicer when the hdr content is full screen. Imo I wish I could just disable partial screen hdr and keep the full screen implementation because it is that distracting.
Out curiosity — I don’t have Instagram to test and this is a perceptual question anyways — if you enable iOS Settings > Accessibility > Display > Reduce White Point (25%) and increase the screen brightness slightly to compensate for the slight dimness; does that reduce or eliminate the ‘jarring’ness?
I’m talking about my macbook and that would be untenable because it needs max brightness on non hdr content in a daylight lit room.
That's not inherent to HDR though. BFV (unless I'm confusing it with something else) has a HDR adjustment routine where you push a slider until the HDR white and the SDR white are identical. Same could be done for desktop environments. In my experience, HDR support is very lacking in PCs atm. You can't even play Dolby Vision on Windows, which is the only widely-used HDR format with dynamic metadata.
If you mean https://i.imgur.com/0LtYuDZ.jpeg that is probably the slider GP wants but it's not about matching HDR white to SDR white, it's just about clamping the peak HDR brightness in its own consideration. The white on the left is the HDR brightness according to a certain value in nits set via the bottom slider. The white on the right is the maximally bright HDR signal. The goal of adjusting the slider is to find how bright of an HDR white your display can actually produce, which is the lowest slider value at which the two whites appear identical to a viewer.
Some games also have a separate slider https://i.imgur.com/wenBfZY.png for adjusting "paper white", which is the HDR white one might normally associate with matching to SDR reference white (100 nits when in a dark room according to the SDR TV color standards, higher in other situations or standards). Extra note: the peak brightness slider in this game (Red Dead Redemption 2) is the same knob as the brightness slider in the above Battlefield V screenshot)
Thanks for clarifying this!
>HDR support is very lacking in PCs atm.
I think it's because no one wants it.
No; Windows 10 barely supports it, and of my hundreds of Steam games, exactly none have any code making use of it; seemingly only AAA mega-budget games have the pipeline bandwidth, but e.g. Dune Imperium and Spaceways have no reason to use it and so don’t bother. Windows 11 focused on improving wide color support which is much more critical, as any game shipping for Mac/Win already has dealt with that aspect of the pipeline and has drabified their game for the pre-Win11 ICC nightmares. Even games like Elite Dangerous, which would be a top candidate for both HDR and wide color, don’t handle this yet; their pipeline is years old and I don’t envy them the work of trying to update it, given how they’re overlaying the in-game UI in the midst of their pipeline.
(It doesn’t help that Windows only allows HDR to be defined in EDID and monitor INF files, and that PC monitors start shutting off calibration features when HDR is enabled because their chipsets can’t keep up — just as most modern Sony televisions can’t do both Dolby Vision and VRR because that requires too much processing power for their design budget.)
It's an absolute mess on Windows. It's one area where I totally bow to Apple's vertical integration which makes this stuff flawless.
Nothing you're written disproves that no one wants it, if anything the fact that nothing really supports it implies that it's not a valuable feature that people are clamoring for.
Given how popular it is in modern console AAA gaming, which many people do play on PCs, one can reasonably expect that any studio desiring to target high contrast gaming — a valuable niche for stealth games, where a bright torch should absolutely stand out brightly against the darkest shadows, as well as star exploration and anything with landscapes or specular highlights — would like to be targeting HDR on PC, if only to simplify their cross-platform build and test pipelines, but cannot due to platform barriers that don’t exist for consoles.
>Given how popular it is in modern console AAA gaming
What are some of the games where it's necessary?
You’ll have to do that research yourself, apologies; I haven’t looked for that data and so I don’t have it available for you.
I want it. And, I'm hardpressed to imagine a multimedia consumer that doesn't care about color and brightness ranges, unless they don't understand what that is. Every movie is recorded and many phones now capture in a range that can't be encoded in 8-bits.
I experience the same thing you do — but my take on it is different. Being hit with HDR images (and videos on YouTube), while unsettling, makes me then realize how just damned dull the SDR world I had been forced to succumb to has been.
Let the whole experience be HDR and perhaps it won't be jarring.
On the browser spec side this is just starting to get implemented as a CSS property https://caniuse.com/mdn-css_properties_dynamic-range-limit so I expect it might start to be a more common thing in web tech based feeds given time.
This happens on Snapchat too with HDR videos. Brightness increases while everything else dims... including the buttons.
This seems more like a "your feeds" problem than an HDR problem. Much in the same way people screencap and convert images willy nilly. I suggest blocking non HDR content
> A big problem is that it costs the TV, Film, and Photography industries billions of dollars (and a bajillion hours of work) to upgrade their infrastructure. For context, it took well over a decade for HDTV to reach critical mass.
This is also true for consumers. I don't own a single 4k or HDR display. I probably won't own an HDR display until my TV dies, and I probably won't own a 4k display until I replace my work screen, at which point I'll also replace one of my home screens so I can remote into it without scaling.
Pretty much any display you can buy today will be HDR capable, though that doesn't mean much.
I think the industry is strangling itself putting "DisplayHDR 400" certification on edgelit/backlit LCD displays. In order for HDR to look "good" you either need high resolution full array local dimming backlighting (which still isn't perfect), or a panel type that doesn't use any kind of backlighting like OLED.
Viewing HDR content on these cheap LCDs often looks worse than SDR content. You still get the wider color gamut, but the contrast just isn't there. Local dimming often loses all detail in shadows whenever there is something bright on the screen.
HDR marketing on monitors almost seems like a scam. Monitors will claim HDR compatibility when what they actually means is they will take the HDR data stream and display it exactly the same as SDR content because they don't actually have the contrast and brightness ability of a proper HDR monitor.
Few things are in absolutes. Yes most consumers wont have every screen hdr nor 4k, but most consumers use a modern smartphone and just about every modern smartphone from the past half decade or more has hdr of some level.
I absolutely loathe consuming content on a mobile screen, but its the reality is the vast majority are using phone and tablets most the time.
Funny enough HDR content works absolutely perfect as long as it stays on device that has both HDR-recording and display tech, aka smartphones.
The problem starts with sending HDR content to SDR-only devices, or even just other HDR-standards. Not even talking about printing here.
This step can inherently only be automated so much, because it's also a stylistic decision on what information to keep or emphasize. This is an editorial process, not something you want to emburden casual users with. What works for some images can't work for others. Even with AI the preference would still need to be aligned.
How would I know if my android phone has an HDR screen?
[edit]
Some googling suggested I check in the Netflix app; at least Netflix thinks my phone does not support HDR. (Unihertz Jelly Max)
I have to think you are are the 1-3% outlier though. Everyone I know has an HDR screen even my friend who never buys anything new, but he did run out and buy an HDR tv to replace his old one that he gave to his son.
I honestly do not know if I have any screen that supports HDR. At least I've never noticed any improved image quality when viewing HDR video content and compare the image on my M3 Macbook Pro screen vs. an old external IPS monitor. Maybe my eyes are just broken?
To demonstrate some contrast (heh) with another data point from someone closer to the other extreme, I’ve owned a very HDR-capable monitor (the Apple Pro Display XDR) since 2020, so that’s 5 years now. Content that takes full advantage of it is still rare, but it’s getting better slowly over time.
I have a screen which is "HDR" but what that means is when you turn the feature on it just makes everything more muted, it doesn't actually have any more dynamic range. When you turn HDR on for a game it basically just makes most things more muddy grey.
I also have a screen which has a huge gamut and blows out colors in a really nice way (a bit like the aftereffects of hallucinogens, it has colors other screens just don't) and you don't have to touch any settings.
My OLED TV has HDR and it actually seems like HDR content makes a difference while regular content is still "correct".
The cheap displays adding broken HDR400 support destroyed so much public opinion on HDR. Not actually providing a wider range but accepting the HDR signal would at least have been a minor improvement if the tone mapping weren't completely broken to the point most people just associate HDR with a washed out picture.
>The cheap displays adding broken HDR400 support destroyed so much public opinion on HDR.
It's funny because the display I have that does this was a relatively premium Odyssey G7 which at $700 isn't at all a cheap monitor. (I like it, it's just not at all HDR, or at least not perceivably so compared to Apple devices and an OLED TV)
The only HDR monitors I have actually enjoyed (which, admittedly, is a slightly higher bar than "most of this price aren't completely miscalibrated") have been $1500+ which is a bit ridiculous considering far cheaper TVs do far better with HDR (though tend to suck as monitors). In the Odyssey line this would be the Neo G9 options, but I've never went home with one because they lack flat variants for the ones with good HDR.
> I don't own a single 4k or HDR display
Don't feel like you have to. I bought a giant fancy TV with it, and even though it's impressive, it's kinda like ultra-hifi-audio. I don't miss it when I watch the same show on one of my older TVs.
If you ever do get it, I suggest doing for a TV that you watch with your full attention, and watching TV / movies in the dark. It's not very useful on a TV that you might turn on while doing housework; but very useful when you are actively watching TV with your full attention.
I totally love HDR on my OLED TV, and definitely miss it on others.
Like a lot of things, it’s weird how some people are more sensitive to visual changes. For example:
- At this point, I need 120hz displays. I can easily notice when my wife’s phone is in power saver mode at 60hz.
- 4k vs 1080p. This is certainly more subtle, but I definitely miss detail in lower res content.
- High bitrate. This is way more important than 4k vs 1080p or even HDR. But it’s so easy to tell when YouTube lowers the quality setting on me, or when a TV show is streaming at a crappy bitrate.
- HDR is tricky, because it relies completely on the content creator to do a good job producing HDR video. When done well, the image basically sparkles, water looks actually wet, parts of the image basically glow… it looks so good.
I 100% miss this HDR watching equivalent content on other displays. The problem is that a lot of content isn’t produced to take advantage of this very well. The HDR 4k Blu-ray of several Harry Potter movies, for example, has extremely muted colors and dark scenes… so how is the image going to pop? I’m glad we’re seeing more movies rely on bright colors and rich, contrasty color grading. There are so many old film restorations that look excellent in HDR because the original color grade had rich, detailed, contrasty colors.
On top of that, budget HDR implementations, ESPECIALLY in PC monitors, just don’t get very bright. Which means their HDR is basically useless. It’s impossible to replicate the “shiny, wet look” of really good HDR water if the screen can’t get bright enough to make it look shiny. Plus, it needs to be selective about what gets bright, and cheap TVs don’t have a lot of backlighting zones to make that happen very well.
So whereas I can plug in a 4k 120hz monitor and immediately see the benefit in everything I do for normal PC stuff, you can’t get that with HDR unless you have good source material and a decent display.
> At this point, I need 120hz displays. I can easily notice when my wife’s phone is in power saver mode at 60hz.
Yeah, the judder is a lot more noticeable on older TVs now that I have a 120hz TV. IMO, CRTs handled this the best, but I'm not going back.
I don't either see a point of having 4K TV vs 1080p TV. To me is just marketing, I have at my house both a 4K and a 1080p and from a normal viewing distance (that is 3/4 meters) you don't see differences.
Also in my country (Italy) TV transmissions are 1080i at best, a lot are still 570i (PAL resolution). Streaming media can be 4K (if you have enough bandwidth to stream it at that resolution, which I don't have at my house). Sure, if you download pirated movies you find it at 4K, and if you have the bandwidth to afford it... sure.
But even there, sometimes is better a well done 1080p movie than an hyper compressed 4K one, since you see compression artifacts.
To me 1080p, and maybe even 720p, is enough for TV vision. Well, sometimes I miss the CRT TVs, they where low resolution but for example had a much better picture quality than most modern 4K LCD TV where black scenes are gray (I know there is OLED, but is too expensive and has other issues).
I feel the same way. To be honest even the laptop retina screen is excess. I sometimes go back to a 2012 non retina macbook pro and to be honest at normal laptop viewing angles, you can’t really discern pixels. Biggest difference is display scaling but I have my retina scaled at what the old display would be anyhow because otherwise its too small.
Kind of crazy no one thought of this aspect and we just march on to higher resolution and the required hardware for that.
For TVs under ~80" I feel like you'd have to be sitting abnormally close to your TV for it to matter much. At the same time I think the cost difference between producing 1080p and 4k panels is so low it probably doesn't matter. Like you say, things like the backlight technology (or lack thereof) make a much bigger difference in perceived quality but that's also where the actual cost comes in.
I agree about 4k vs non-4k. I will say going OLED was a huge upgrade, even for SDR content. HDR content is hit-or-miss...I find some of it is tastefully done but in many cases is overdone.
My own movie collection is mostly 2-4GB SDR 1080p files and looks wonderful.
You still watch broadcast TV?
Jokes aside, when a 4k TV has a good upscaler, it's hard to tell the difference between 1080 and 4k. Not impossible; I certainly can, but 1080 isn't distracting.
Well, if it was for me I wouldn't even have a TV at home since I wouldn't use it (and thus I would also save the TV tax that we have to pay in our country if we own one). But my girlfriend watches a lot of TV, and I often watch it with her of course, thus I have two TVs.
This is also true for consumers. I don't own a single 4k or HDR display. I probably won't own an HDR display until my TV dies, and I probably won't own a 4k display until I replace my work screen, at which point I'll also replace one of my home screens so I can remote into it without scaling.
People in the HN echo chamber over-estimate hardware adoption rates. For example, there are millions of people who went straight from CDs to streaming, without hitting the iPod era.
A few years ago on HN, there was someone who couldn't wrap their brain around the notion that even though VCRs were invented in the early 1960's that in 1980, not everyone owned one, or if they did, they only had one for the whole family.
Normal people aren't magpies who trash their kit every time something shiny comes along.
> A few years ago on HN, there was someone who couldn't wrap their brain around the notion that even though VCRs were invented in the early 1960's that in 1980, not everyone owned one, or if they did, they only had one for the whole family.
Point of clarification: While the technology behind the VCR was invented in the '50s and matured in the '60s, consumer-grade video tape systems weren't really a thing until Betamax and VHS arrived in 1975 and 1976 respectively.
Early VCRs were also incredibly expensive, with prices ranging from $3,500 to almost $10,000 after adjusting for inflation. Just buying into the VHS ecosystem at the entry level was a similar investment to buying an Apple Vision Pro today.
Exactly my point. But people on HN, especially the person I referenced, don't understand that we didn't just throw stuff away and go into debt to buy the latest gadgets because we were told to.
>there are millions of people who went straight from CDs to streaming, without hitting the iPod era
Who?
There was about a decade there where everyone who had the slightest interest in music had an mp3 player of some kind, at least in the 15-30 age bracket.
I imagine this depends a LOT on your specific age and what you were doing in the 00's when MP3 player usage peaked.
I finished high school in 2001 and didn't immediately go to college, so I just didn't have a need for a personal music player anymore. I was nearly always at home or at work, and I drove a car that actually had an MP3 CD player. I felt no need to get an iPod.
In 2009, I started going to college, but then also got my first smartphone, the Motorola Droid, which acted as my portable MP3 player for when I was studying in the library or taking mass transit.
If you were going to school or taking mass transit in the middle of the '00s, then you were probably more likely to have a dedicated MP3 player.
I don't know if I count, but I never owned a dedicated MP3 player[1], I listened to MP3s on my computer, but used CDs and cassettes while on the move, until I got an android phone that had enough storage to put my music collection on.
1: Well my car would play MP3s burned to CDs in its CD player; not sure if that counts.
I skipped 2 generations for portable music : went straight from cassette to smartphone with MP3 (and radio).
My father, for one. He was entirely happy with radio in the car and CDs at home.
Only my laptop supports HDR. But that's one that I own anyway.
> AI cannot read your mind, so it cannot honor your intent.
This. I can always tell when someone "gets" software development when they either understand (or don't) that computers can't read minds or infer intent like a person can.
Just one other thing. In Analog you also have compensating developers, which will exhaust faster in darker areas (or lighter if you think in negative), and allow for lighter areas more time to develop and show, and hence some more control of the range. Same but to less degree with stand development which uses very low dilutions of the developer, and no agitation. So dodging and burning is not the only way to achieve higher dynamic range in analog photos.
About HDR on phones, I think they are the blight of photography. No more shadows and highlights. I find they are good at capturing family moments, but not as a creative tool.
For analog photos the negative also has more dynamic range than your screen or photopaper without any of that. Contrast is applied in the darkroom by choice of photopaper and enlarger timing and light levels, or after scanning with contrast adjustment applied in post processing. It is really a storage medium of information more than how the final image ought to look.
Slide film has probably a third the dynamic range of negative film and is meant as the final output fit for projection to display.
> About HDR on phones, I think they are the blight of photography. No more shadows and highlights.
HDR is what enables you to capture both the darkest shadow detail and the brightest highlight detail.
With SDR, one or both are often simply just lost. It might come down to preference — if you're an "auto" shooter and like the effect of the visual information at the edges of the available dynamic range being truncated, SDR is for you.
Some people prefer to capture that detail and have the ability to decide whether and how to diminish or remove it, with commensurately more control over the artistic impact. For those folks, HDR is highly desirable.
I wrote a raw processing app Filmulator that simulates stand development/compensating developers to give such an effect to digital photography.
I still use it myself but I need to redo the build system and release it with an updated LibRaw... not looking forward to that.
So, HN, are HDR monitors worth it? I remember ~10 years ago delaying my monitor purchase for the HDR one that was right around the corner, but never (in my purchasing scope) became available. Time for another look?
The utility of HDR (as described in the article) is without question. It's amazing looking at an outdoors (or indoors with windows) scene with your Mk-1 eyeballs, then taking a photo and looking at it on a phone or PC screen. The pic fails to capture what your eyes see for lighting range.
HDR gaming: Yes.
HDR full screen content: Yes.
HDR general desktop usage: No. In fact you'll probably actively dislike it to the point of just turning it off entirely. The ecosystem just isn't ready for this yet, although with things like the "constrained-high" concepts ( https://www.w3.org/TR/css-color-hdr-1/#the-dynamic-range-lim... ) this might, and hopefully does, change & improve to a more pleasing result
Also this is assuming an HDR monitor that's also a good match for your ambient environment. The big thing nobody really talks about wiith HDR is that it's really dominated by how dark you're able to get your surrounding environment such that you can push your display "brightness" (read: SDR whitepoint) lower and lower. OLED HDR monitors, for example, look fantastic in SDR and fantastic in HDR in a dark room, but if you have typical office lighting and so you want an SDR whitepoint of around 200-300 nits? Yeah, they basically don't do HDR at all anymore at that point.
> HDR gaming: Yes.
The difference is absolutely stunning in some games.
In MS Flight Simulator 2024, going from SDR to HDR goes from looking like the computer game it is to looking life-like. Deeper shadows with brighter highlights makes the scene pop in ways that SDR just can't do.
EDIT: You'll almost certainly need an OLED monitor to really appreciate it, though. Local dimming isn't good enough.
I use HDR for general usage, Windows ruins non-HDR content when HDR is enabled due to their choice of sRGB tf. Luckily every Linux DE has chosen to use the gamma 2.2 tf, and looks fine for general usage.
I use a mini-led monitor, and its quite decent, except for starfields, & makes it very usable even in bright conditions, and HDR video still is better in bright conditions than the equivalent SDR video.
https://github.com/dylanraga/win11hdr-srgb-to-gamma2.2-icm
Windows HDR implementation is janky as hell. For months after I got my monitor I couldn't take screenshots because they'd all appear completely blown out, like you cranked the brightness to 300%.
Eventually I did some digging and found there's a setting in Snipping Tool that just... makes screenshots work on HDR displays.
It also seems to add another layer of Your Desktop Trying To Sort Its Shit Out when launching a game that's full screen. Sometimes it's fine, but some games like Balatro will appear fine at first, but then when you quit back to the desktop everything is washed out. Sleeping my PC and waking it back up seems to resolve this.
I recently played through Armored Core VI, and it supports HDR, but whenever I adjust my volume the screen becomes washed out to display the volume slider. Screenshots and recordings also appear washed out in the resulting file.
Agree. Wide gamut and HDR is janky as hell on Windows. I have multi-mon with one SDR and one HDR and that plays havoc with things. But even Microsoft apps aren't updated. I'm pretty certain even Explorer still doesn't support HDR or wide gamut for thumbnails so everything looks either under or oversaturated in the previews. If you open stuff in the default Photos app there is a "flash of content" where it displays it in the wrong profile before it maps it to the correct one, too.
I always thought the "Your Desktop Trying To Sort Its Shit Out" part was a necessary evil, but other platforms don't suffer from this (at least from what I can tell); the state of HDR on Windows is very disappointing, even just adjusting the TF to gamma 2.2 would make it substantially better. Watching all your non hdr content's blacks become gray is terrible. I assume the washed out appearance comes from it giving up on doing SDR->HDR for the desktop.
My brother got an OLED monitor & was telling me how bad his experience was on Windows, & he recently switched to Linux & does not have the issues he was complaining about before. Ofc, downsides to hdr on Linux (no hdr on chromium, hdr on Firefox is unfinished) atm, but the foundation seems better set for it.
Every few years I turn on that HDR toggle for my desktop PC, and it never lasts longer than a day or two.
Top tip: If you have HDR turned on for your display in Windows (at least, MacOS not tested) and then share your screen in Teams, your display will look weirdly dimmed for everyone not using HDR on their display—which is everyone.
HDR on desktop in windows looks straight up broken on some HDR monitors I’ve tried
Like totally washed out
They are working on it. It will be improved with the next release and broken with the sucsequent release. /s
A lot of monitors that advertise HDR support really shouldn't. Many of them can decode the signal but don't have the hardware to accurately reproduce it, so you just end up with a washed out muddy looking mess where you're better off disabling HDR entirely.
As others here have said, OLED monitors are generally excellent at reproducing a HDR signal, especially in a darker space. But they're terrible for productivity work because they'll get burned in for images that don't change a lot. They're fantastic for movies and gaming, though.
There are a few good non-OLED HDR monitors, but not many. I have an AOC Q27G3XMN; its a 27" 1440p 180hz monitor that is good for entry-level HDR, especially in brighter rooms. It has over 1000 nits of brightness, and no major flaws. It only has 336 backlight zones, though, so you might notice some blooming around subtitles or other fine details where there's dark and light content close together. (VA panels are better than IPS at suppressing that, though.) It's also around half the price of a comparable OLED.
Most of the other non-OLED monitors with good HDR support have some other deal-breaking flaws or at least major annoyances, like latency, screwing up SDR content, buggy controls, etc. The Monitors Unboxed channel on YouTube and rtngs.com are both good places to check.
I think an OLED is basically an absolute necessity for HDR content.
My current monitor is an OLED and HDR in games looks absolutely amazing. My previous was an IPS that supported HDR, but turning it on caused the backlight to crank to the max, destroying black levels and basically defeating the entire purpose of HDR. Local dimming only goes so far.
Modern mini-led monitors are very good. The “local” dimming is so local that there isn’t much light bleed even in the worst-case situations (cursor over black background makes it particularly apparent).
The advantage of LEDs is they’re brighter. For example, compare two modern Asus ProArt displays: their mini-LED (PA32UCXR) at 1600 nits and their OLED (PA32DC) at 300ish nits. The OLED is 20% more expensive. These two monitors have otherwise comparable specs. Brightness matters a lot for HDR because if you’re in a bright room, the monitor’s peak brightness needs to overpower the room.
Plus for color managed work, I think LED monitors are supposed to retain their calibration well. OLEDs have to be frequently recalibrated.
And so-called micro-LEDs are coming soon, which promise to make “local” so small that it’s imperceptible. I think the near-term future of displays is really good LEDs.
> My previous was an IPS that supported HDR, but turning it on caused the backlight to crank to the max, destroying black levels and basically defeating the entire purpose of HDR
Yeah, that's kind of what I meant when I said that most monitors that advertise HDR shouldn't.
The AOC monitor is the third or fourth one I've owned that advertised HDR, but the first one that doesn't look like garbage when it's enabled.
I haven't gone oled yet because of both the cost and the risk of burn-in in my use case (lots of coding and other productivity work, occasional gaming).
Avoid cranking up the OLED brightness over 70% for static content and absolutely never drive SDR-Reds into the HDR range by using fake HDR-modes when having brightness high.
I have a LG 2018 OLED that has some burnt in Minecraft hearts because of that, not from Minecraft itself, but just a few hours of minecraft Youtube video in those settings from the built in youtube client, but virtually no other detectable issues after excessive years of use with static content.
You only see them with fairly uniform colors as a background where color banding would usually be my bigger complaint.
So burn-ins definitely happen, but they are far from being a deal breaker over the obvious benefits you get vs other types of displays.
And driving everything possible in dark mode (white text on dark bg) on those displays is even the logical thing to do. Then you dont need much max brightness anyway and even save some energy.
> A lot of monitors that advertise HDR support really shouldn't. Many of them can decode the signal but don't have the hardware to accurately reproduce it,
You can lie as much as you like as an advertiser. At the beginning, HD meant 1080p. Then, 1080p became FullHD, only that 720p to became HD and to make HD digital television possible. (the original SD television is worse than analog)
Yes but with asterisks; Best way I can describe it:
You know the 0-10 brightness slider you have to pick at the start of a game? Imagine setting it to 0 and still being able to spot the faint dark spot. The dynamic range of things you can see is so much expanded.
Early HDR screens were very limited (limited dimming zones, buggy implementation) but if you get one post 2024 (esp the oled ones) they are quite decent. However it needs to be supported at many layers: not just the monitor, but also the operating system, and the content. There are not many games with proper HDR implementation; and even if there is, it may be bad and look worse — the OS can hijack the rendering pipeline and provide HDR map for you (Nvidia RTX HDR) which is a gamble: it may look bleh, but sometimes also better than the native HDR implementation the game has).
But when everything works properly, wow it looks amazing.
> You know the 0-10 brightness slider you have to pick at the start of a game? Imagine setting it to 0 and still being able to spot the faint dark spot. The dynamic range of things you can see is so much expanded.
Note that HDR only actually changes how bright things can get. There's zero difference in the dark regions. This is made confusing because HDR video marketing often claims it does, but it doesn't actually. HDR monitors do not, in general, have any advantage over SDR monitors in terms of the darks. Local dimming zones improve dark contrast. OLED improves dark contrast. Dynamic contrast improves dark contrast. But HDR doesn't.
My understanding is that on the darker scenes (say, 0 to 5 in the brightness slider example), there is difference in luminance value with HDR but not SDR, so there is increased contrast and detail.
This matches my experience; 0 to 5 look identically black if I turn off HDR
You may have a monitor that only enables local dimming zones when fed an HDR signal but not when fed an SDR one, but that would be unusual and certainly not required. And likely something you could change in your monitors controls. On things like an OLED, though, there's no difference in the darks. You'd see a difference between 8bit and 10bit potentially depending on what "0 to 5" means, but 10-bit SDR is absolutely a thing (it predates HDR even)
But like if you can't see a difference between 0 to 5 in a test pattern like this https://images.app.goo.gl/WY3FhCB1okaRANc28 in SDR but you can in HDR then that just means your SDR factory calibration is bad, or you've fiddled with settings that broke it.
> You know the 0-10 brightness slider you have to pick at the start of a game? Imagine setting it to 0 and still being able to spot the faint dark spot
0-10 is so old century. Now it should be 0-2 (0 - default, 1 dark, 2 light,) /s
IF you have a display that can it roughly a 1000 nits, then for movies and games yes definitely the difference with SDR is pretty huge.
If you have say a 400 nits display the HDR may actually look worse than SDR. So it really depends on your screen.
Honestly, I find the extended brightness FAR less important than the extended color gamut. I have a ~300 nit VA monitor that I'm generally quite happy with and that looks fantastic with well-built HDR renderers.
Given that monitors report information about their HDR minimum and maximum panel brightness capabilities to the machine they are connected to, any competently-built HDR renderer (whether that be for games or movies or whatever) will be able to take that information and adjust the picture appropriately.
I've been thinking of moving out of the Apple ecosystem but after seeing Severance on my iPhone Pro screen I feel like I want the keep the option to have the same HDR experience for movies specifically. With HDR support landing in Linux just a month ago I'm inclined to spend on a good monitor. However, I have IPS HDR 600 monitor but I never felt that the screen was as glorious as the iPhone screen.
I'd also be interested in hearing whether it makes sense to look into OLED HDR 400 screens (Samsung, LG) or is it really necessary to get an Asus ProArt which can push the same 1000 nits average as the Apple XDR display (which, mind you, is IPS).
I think it depends on the screen and also what you use it for. My OLED is unusable for normal work in HDR because it's designed around only a small portion of the screen being at max brightness - reasonable for a game or movie, but the result is that a small window with white background will look really bright, but if I maximize it, it'll look washed out, grey not white.
Also the maximum brightness isn't even that bright at 800 nits, so no HDR content really looks that different. I think newer OLEDs are brighter though. I'm still happy with the screen in general, even in SDR the OLED really shines. But it made me aware not all HDR screens are equal.
Also, in my very short experiment using HDR for daily work I ran into several problems, the most serious of which was the discovery that you can no longer just screenshot something and expect it to look the same on someone else's computer.
> ...the most serious of which was the discovery that you can no longer just screenshot something and expect it to look the same on someone else's computer.
To be pedantic, this has always been the case... Who the hell knows what bonkers "color enhancement" your recipient has going on on their end?
But (more seriously) it's very, very stupid that most systems out there will ignore color profile data embedded in pictures (and many video players ignore the same in videos [0]). It's quite possible to tone-map HDR stuff so it looks reasonable on SDR displays, but color management is like accessibility in that nearly noone who's in charge of paying for software development appears to give any shits about it.
[0] A notable exception to this is MPV. I can't recommend this video player highly enough.
HDR when it works properly is nice, but nearly all HDR LCD monitors are so bad, they're basically a scam.
The high-end LCD monitors (with full-array local dimming) barely make any difference, while you'll get a lot of downsides from bad HDR software implementations that struggle to get the correct brightness/gamma and saturation.
IMHO HDR is only worth viewing on OLED screens, and requires a dimly lit environment. Otherwise either the hardware is not capable enough, or the content is mastered for wrong brightness levels, and the software trying to fix that makes it look even worse.
Apple's Displays yes. But I got a Phillips 4k OLED recently, and I'm already regretting that decision. I need to turn it off every 4 hours to refresh the pixels. Sometimes an entire line of pixels is brighter than the rest. I wiped it with a cloth while pixel refresh was running, and then saw burned in streaks in the direction of the wipe.
And thats now that all the LEDs are still fresh. I can't imagine how bad it will be in a few years.
Also, a lot of Software doesn't expect the subpixel arrangement, so text will often look terrible.
For gaming, definitely. An HDR Oled monitor is so immersive.
I've had an OLED TV since 2017 and the answer is a resounding yes... if you get an OLED and use it for movies or full screen gaming. Anything else is basically pointless.
For desktop work, don't bother unless your work involves HDR content.
Dunno if it's just my screen or setup but on windows I have a Dell U4025QW and HDR on the desktop just looks strange, overly dull. Looks good in games but I have to manually turn it on and off each time on the screen.
On my Macbook Pro only activates when it needs to but honestly I've only seen one video [1] that impressed me with it, the rest was completely meh. Not sure if its because it's mostly iPhone photography you see in HDR which is overall pretty meh looking anyway.
[1] https://www.youtube.com/watch?v=UwCFY6pmaYY I understand this isn't a true HDR process but someone messing with it in post, but it's the only video I've seen that noticeably shows you colors you can't see on a screen otherwise.
For the love of gods, read the reviews though. Most HDR displays will make the picture worse not better because they have only implemented enough to put a sticker on the box.
You have to spend really good money to get a display which does HDR properly.
for movies yes. for vim/vscode nope.
Note for Firefox users - view the page in Chrome to see more of what they are talking about. I was very confused by some of the images, and it was a world of difference when I tried again in Chrome. Things began to make a lot more sense - is there a flag I am missing in Firefox on the Mac?
https://bugzilla.mozilla.org/show_bug.cgi?id=hdr there's the tracking issue for HDR support.
Tbh, I'm glad this isn't supported in Firefox as of right now
HDR support in Chrome (Android) looks still broken for me. For one, some of the images on the blog have a posterization effect, which is clearly wrong.
Second, the HDR effect seems to be implemented in a very crude way, which causes the whole Android UI (including the Android status bar at the top) to become brighter when HDR content is on screen. That's clearly not right. Though, of course, this might also be some issue of Android rather than Chrome, or perhaps of the Qualcomm graphics driver for my Adreno GPU, etc.
Yeah, the HDR videos on my Asus Zenfone 9 (on Android 14) look like really terrible.
Which Android phone are you using?
This one: https://gsmarena.com/motorola_edge_30_ultra-11206.php
Can confirm on Windows 11 with HDR enabled on my display— I see the photos in the article correctly on Chrome and they're a grey mess on Firefox.
On macOS even without HDR enabled on my display there's a striking difference due to better tone mapping between Safari and Firefox.
If I enable HDR the Firefox ones become a gray mess vs the lights feeling like actual lights in Safari.
For what it's worth, your comment has me convinced I just "can't see" HDR properly because I have the same page side-by-side on Firefox and Chrome on my M4 MBP and honestly? Can't see the difference.
edit: Ah, nevermind. It seems Firefox is doing some sort of post-processing (maybe bad tonemapping?) on-the-fly as the pictures start out similar but degrade to washed out after some time. In particular, the "OVERTHROW BOXING CLUB" photo makes this quite apparent.
That's a damn shame Firefox. C'mon, HDR support feels like table stakes at this point.
edit2: Apparently it's not table stakes.
> Browser support is halfway there. Google beat Apple to the punch with their own version of Adaptive HDR they call Ultra HDR, which Chrome 14 now supports. Safari has added HDR support into its developer preview, then it disabled it, due to bugs within iOS.
at which point I would just say to `lux.camera` authors - why not put a big fat warning at the top for users with a Firefox or Safari (stable) browser? With all the emphasis on supposedly simplifying a difficult standard, the article has fallen for one of its most famous pitfalls.
"It's not you. HDR confuses tons of people."
Yep, and you've made it even worse for a huge chunk of people. :shrug: Great article n' all just saying.
I wasn't using Firefox, but I had the page open on an old monitor. I dragged the page to an HDR display and the images pop.
Does anyone else find the hubris in the first paragraph writing as off-putting as I do?
"we finally explain what HDR actually means"
Then spends 2/3rds of the article on a tone mapping expedition, only to not address the elephant in the room, that is the almost complete absence of predictable color management in consumer-grade digital environments.
UIs are hardly ever tested in HDR: I don't want my subtitles to burn out my eyes in actual HDR display.
It is here, where you, the consumer, are as vulnerable to light in a proper dark environment for movie watching, as when raising the window curtains on a bright summer morning. (That brightness abuse by content is actually discussed here)
Dolby Vision and Apple have the lead here as a closed platforms, on the web it's simply not predictably possible yet.
Best hope is the efforts of the Color on the Web Community Group from my impression.
https://github.com/w3c/ColorWeb-CG
> Does anyone else find the hubris in the first paragraph writing as off-putting as I do? > "we finally explain what HDR actually means"
No. Because it's written for the many casual photographers we've spoken with who are confused and asked for an explainer.
> Then spends 2/3rds of the article on a tone mapping expedition, only to not address the elephant in the room, that is the almost complete absence of predictable color management in consumer-grade digital environments.
That's because this post is about HDR and not color management, which is different topic.
>No. Because it's written for the many casual photographers we've spoken with who are confused and asked for an explainer.
To be fair, it would be pretty weird if you found your own post off-putting :P
Me, routinely, reading things I wrote a while ago: what is this dreck
> That's because this post is about HDR
It's about HDR from the perspective of still photography, in your app, on iOS, in the context of hand-held mobile devices. The post's title ("What Is HDR, Anyway?"), content level and focus would be appropriate in the context of your company's social media feeds for users of your app - which is probably the audience and context it was written for. However in the much broader context of HN, a highly technical community whose interests in imaging are diverse, the article's content level and narrow focus aren't consistent with the headline title. It seems written at a level appropriate for novice users.
If this post was titled "How does Halide handle HDR, anyway?" or even "How should iOS photo apps handle HDR, anyway?" I'd have no objection about the title's promise not matching the content for the HN audience. When I saw the post's headline I thought "Cool! We really need a good technical deep dive into the mess that is HDR - including tech, specs, standards, formats, content acquisition, distribution and display across content types including stills, video clips and cinematic story-telling and diverse viewing contexts from phones to TVs to cinemas to VR." When I started reading and the article only used photos to illustrate concepts best conveyed with color gradient graphs PLUS photos, I started to feel duped by the title.
(Note: I don't use iOS or your app but the photo comparison of the elderly man near the end of the article confused me. From my perspective (video, cinematography and color grading), the "before" photo looks like a raw capture with flat LUT (or no LUT) applied. Yet the text seemed to imply Halide's feature was 'fixing' some problem with the image. Perhaps I'm misunderstanding since I don't know the tool(s) or workflow but I don't see anything wrong with the original image. It's what you want in a flat capture for later grading.)
> It's about HDR from the perspective of still photography, in your app, on iOS, in the context of hand-held mobile devices.
It's from the perspective of still photography, video, film, desktop computing, decades of research papers, and hundreds of years of analog photography, condensed into something approachable.
> However in the much broader context of HN, a highly technical community whose interests in imaging are diverse, the article's content level and narrow focus aren't consistent with the headline title. It seems written at a level appropriate for novice users.
"On-Topic: Anything that good hackers would find interesting. That includes more than hacking and startups. If you had to reduce it to a sentence, the answer might be: anything that gratifies one's intellectual curiosity"
To be clear, I didn't submit the post, and I never submit my posts. I don't care if my posts make a splash here, and kind of dread when they do because anything involving photography or video attracts the most annoying "well actually" guys on the Internet.
> When I saw the post's headline I thought "Cool! We really need a good technical deep dive into the mess that is HDR - including tech, specs, standards, formats, content acquisition, distribution and display across content types including stills, video clips and cinematic story-telling and diverse viewing contexts from phones to TVs to cinemas to VR."
The post is called, "What is HDR," and the introduction explains the intended audience. That audience is much larger than "people who want to read about ITU-R Recommendation BT.2100." But if you think people are interested in a post like that, by all means write it.
> However in the much broader context of HN, a highly technical community whose interests in imaging are diverse, the article's content level and narrow focus aren't consistent with the headline title. It seems written at a level appropriate for novice users.
That is hardly the fault of the authors though. The article seems entirely appropriate for its intended audience, and they can’t control who posts it on a site like HN.
Maybe my response was part of the broader HDR symptom—that the acronym is overloaded with different meanings depending on where you're coming from.
On the HN frontpage, people are likely thinking of one of at least three things:
HDR as display tech (hardware)
HDR as wide gamut data format (content)
HDR as tone mapping (processing)
...
So when the first paragraph says we finally explain what HDR actually means, it set me off on the wrong foot—it comes across pretty strongly for a term that’s notoriously context-dependent. Especially in a blog post that reads like a general explainer rather than a direct Q&A response when not coming through your apps channels.
Then followed up by The first HDR is the "HDR mode" introduced to the iPhone camera in 2010. caused me to write the comment.
For people over 35 with even the faintest interest in photography, the first exposure to the HDR acronym probably didn’t arrive with the iPhone in 2010, but HDR IS equivalent to Photomatix style tone mapping starting in 2005 as even mentioned later. The ambiguity of the term is a given now. I think it's futile to insist or police one meaning other the other in non-scientific informal communication, just use more specific terminology.
So the correlation of what HDR means or what sentiment it evokes in people by age group and self-assesed photography skill might be something worthwhile to explore.
The post get's a lot better after that. That said, I really did enjoy the depth. The dive into the classic dodge and burn and the linked YouTube piece. One explainer at a time makes sense—and tone mapping is a good place to start. Even tone mapping is fine in moderation :)
I took the post about the same way. Thought it excellent because of depth.
Often, we don't get that and this topic, plus my relative ignorance on it, welcomed the post as written.
Just out of curiosity since your profile suggests your from an older cohort. Do you actively remember the Pixelmatix tone mapping era, or where you already old enough to see this as a passing fad, or was this a more niche thing than I remember?
Now I even remember the 2005 HDR HL2 Lost Coast Demo was a thing 20 years ago: https://bit-tech.net/previews/gaming/pc/hl2_hdr_overview/1/
I was old enough to see it as the passing fad it was.
Niche, style points first kind of thing for sure.
Meta: old enough that getting either a new color not intended, or an additional one visible on screen and having the machine remain able to perform was a big deal.
I missed the MDA/EGA/CGA/Hercules era and jumped right into glorious VGA. Only start options for some DOS-games informed you about that drama in the mid 90s not having any idea what that meant otherwise.
It is a fun era NOW. I love the pre VGA PC and earlier systems, like Apple 2, Atari. Am building a CGA system with a K2 CPU to go hacking on to see what was possible. I have, as do many, unfinished business :)
Back then, it was fun at times, bit was also limiting in ways sometimes hard to fathom ways.
Things are crazy good now, BTW. Almost anything is a few clicks away. The CRT is old, panels so damn good..
> "The first HDR is the "HDR mode" introduced to the iPhone camera in 2010."
Yeah, I had a full halt and process exception on that line too. I guess all the research, technical papers and standards development work done by SMPTE, Kodak, et al in the 1990s and early 2000s just didn't happen? Turns out Apple invented it all in 2010 (pack up those Oscars and Emmys awarded for technical achievement and send'em back boys!)
This post is written for people who have heard "HDR" and feel confused. That introduction lists two types of HDR people might think about. "The first" means "the first of two types we're going to explain," not "the first research in the chronological history of HDR."
They also make no mention of transfer functions, which is the main mechanism which explains why the images “burn your eyes” – content creators should use HLG (which has relative luminance) and not PQ (which has absolute luminance) when they create HDR content for the web.
In theory PQ specifies absolute values, but in practice it's treated as relative. Go load some PQ encoded content on an iPhone, adjust your screen brightness, and watch the HDR brightness also change. Beyond the iPhone, it would be ridiculous to render absolute values as-is, given SDR white is supposedly 100-nits; that would be unwatchable in most living rooms.
Bad HDR boils down to poor taste and the failure of platforms to rein it in. You can't fix bad HDR by switching encodings any more than you can fix global warming by switching from Fahrenheit to Celsius.
> That brightness abuse by content
I predict HDR content on the web will eventually be disabled or mitigated on popular browsers similarly to how auto-playing audio content is no longer allowed [1]
Spammers and advertisers haven't caught on yet to how abusively attention grabbing eye-searingly bright HDR content can be, but any day now they will and it'll be everywhere.
1. https://hacks.mozilla.org/2019/02/firefox-66-to-block-automa...
This seems like a fairly easy problem to solve from a UX standpoint, even moreso than auto-playing audio/video. Present all pages in SDR by default, let the user click a button on the toolbar or url bar to toggle HDR rendering when HDR content is detected.
They haven't, but influencers certainly have, I get regular still images which are rendered as a video to get the HDR brightness boost in Instagram, etc.
It seems fine to me. Made sense on the first read and matches my experiences with OpenEXR and ProPhoto RGB and pre-Apple monitors.
High dynamic resolution has always been about tone mapping. Post-sRGB color profile support is called “Wide color” these days, has been available for twenty years or more on all DSLR cameras (such as Nikon ProPhoto RGB supported in-camera on my old D70), and has nothing to do with the dynamic range and tone mapping of the photo. It’s convenient that we don’t have to use EXR files anymore, though!
An HDR photo in sRGB will have the same defects beyond peak saturation at any given hue point, as an SDR photo in sRGB would, relative to either in DCI-P3 or ProPhoto. Even a two-bit black-or-white “what’s color? on or off pixels only” HyperCard dithered image file can still be HDR or SDR. In OKLCH, the selected luminosity will also impact the available chroma range; at some point you start spending your new post-sRGB peak chroma on luminosity instead; but the exact characteristic of that tradeoff at any given hue point is defined by the color profile algorithm, not by whether the photo is SDR or HDR, and the highest peak saturation possible for each hue is fixed, whatever luminosity it happens to be at.
It's a blog for a (fancy) iPhone camera app.
Color management and handling HDR in UIs is probably a bit out of scope.
I on the other hand never thought or cared about HDR much before but I remember seeing it everywhere. But I feel the article explains well and clearly with examples, for someone like me who isn't much camera literate.
Isn't that the point of the article? That the colloquial meaning of HDR is quite overloaded, and when people complain about HDR, they mean bad tone-mapping? I say this as someone as close to totally ignorant about photography as you can get; I personally thought the article was pretty spectacular.
When I complain about HDR it's because I've intentionally set the brightness of pure white to a comfortable level, and then suddenly parts of my screen are brighter than that. You fundamentally can't solve that problem with just better tone mapping, can you?
You can for some definition of "solve", by tone-mapping all HDR content back down into an SDR range for display.
Well yeah. I considered adding that caveat but elected not to because it's obvious and doesn't add anything to the conversation, since that's obviously not what's meant when the industry talks about "HDR". Should've remembered this is HN.
The bit about "confused" turns me off right away. The kind of high-pressure stereo salesman who hopes I am the kind of 'audiophile' who prevents me from calling myself an 'audiophile' (wants mercury-filled cables for a more 'fluid' sound) always presupposes the reader/listener is "confused".
But it's not the colloquial meaning, HDR is fairly well defined by e.g. ITU-R BT.2100, which addresses colorimetry, luminance and the corresponding transfer functions.
I don't think that's the colloquial meaning. If you asked 100 people on the street to describe HDR, I doubt a single person would bring up ITU-R BT.2100.
HDR has a number of different common meanings, which adds to the confusion.
For example, in video games, "HDR" has been around since the mid '00s, and refers to games that render a wider dynamic range than displays were capable of, and use post-process effects to simulate artifcats like bloom and pupil dilation.
In photography, HDR has almost the opposite meaning of what it does everywhere else. Long and multiple exposures are combined to create an image that has very little contrast, bringing out detail in a shot that would normally be lost in shadows or to overexposure.
Photography’s meaning is also about a hundred years older than video games; high(er) dynamic range was a concern in film processing as far back as Ansel, if not prior. That technology adopted it as a sales keyword is interesting and it’s worth keeping in mind when writing for an audience — but this post is about a photography app, not television content or video games, so one can reasonably expect photography’s definition to be used, even if the audience isn’t necessarily familiar.
I think you may be operating with an idiosyncratic definition of "colloquial"
Colloquial meaning and the well defined meaning are two different things in most cases, right?
It may not be the best explanation, but I think any explanation of HDR beyond a sentence or two of definition that doesn't mention the mess that is tone mapping is entirely remiss.
It's nonsense that an image/video gets to override my screen brightness, end of story. I want that removed, not even a setting, just gone.
The photo capture HDR is good. That's a totally different thing and shouldn't have had its name stolen.
I have a question: how can I print HDR? Is there any HDR printer + paper + display lighting setup?
My hypothesis are the following:
- Increase display lighting to increase peak white point + use a black ink able to absorb more light (can Vantablack-style pigments be made into ink?) => increase dynamic range of a printable picture
- Alternatively, have the display lighting include visible light + invisible UV light, and have the printed picture include an invisible layer of UV ink that shines white : the pattern printed in invisible UV-ink would be the "gain map" to increase the peak brightness past incident visible light into HDR range.
What do you folks think?
Use projection mapping, instead of lighting the photo uniformly, use a projector and project a copy of the image onto the photo, thus you get detailed control over how much light the parts of the image receive.
Alternatively, use transparent film and a bright backlight.
You would need a very strong light surrounded by a darkish environent. I don't see a point in "HDR Paper" if you need to place it under a special light.
usecase: a photo exhibition that wants to show pictures in HDR = An HDR print setup (i.e. paper + appropriately controlled lighting).
Haven't you ever been to a photo exhibition ?
Having come from professional video/film tooling in the 90's to today, it's interesting to see the evolution of what "HDR" means. I used to be a purist in this space, where SDR meant ~8 stops (powers of two) or less of dynamic range, and HDR meant 10+. Color primaries and transfer function mapping were things I spoke specifically about. At this point, though, folks use "HDR" to refer to combinations of things.
Around this, a bunch of practical tooling surfaced (e.g., hybrid log approaches to luminance mapping) to extend the thinking from 8-bit gamma-mapped content presenting ~8 stops of dynamic range to where we are now. If we get away from just trying to label everyting "HDR", there are some useful things people should familiarize with:
1. Color primaries: examples - SDR: Rec. 601, Rec. 709, sRGB. HDR: Rec. 2020, DCI-P3. The new color primaries expand the chromatic representation capabilities. This is pretty easy to wrap our heads around: https://en.wikipedia.org/wiki/Rec._2020
2. Transfer functions: examples - SDR: sRGB, BT.1886. HDR: Rec. 2100 Perceptual Quantizer (PQ), HLG. The big thing in this space to care about is that SDR transfer functions had reference peak luminance but were otherwise relative to that peak luminance. By contrast, Rec. 2100 PQ code points are absolute, in that each code value has a defined meaning in measurable luminance, per the PQ EOTF transfer function. This is a big departure from our older SDR universe and from Hybrid Log Gamma approaches.
3. Tone mapping: In SDR, we had the comfort of camera and display technologies roughly lining up in the video space, so living in a gamma/inverse-gamma universe was fine. We just controlled the eccentricity of the curve. Now, with HDR, we have formats that can carry tone-mapping information and transports (e.g., HDMI) that can bidirectionally signal display target capabilities, allowing things like source-based tone mapping. Go digging into HDR10+, Dolby Vision, or HDMI SBTM for a deep rabbit hole. https://en.wikipedia.org/wiki/Tone_mapping
So HDR is everything (and nothing), but it's definitely important. If I had to emphasize one thing that is non-obvious to most new entrants into the space, it's that there are elements of description of color and luminance that are absolute in their meaning, rather than relative. That's a substantial shift. Extra points for figuring out that practical adaptation to display targets is built into formats and protocols.
A related article on dpreview about Sigmas hdr map in jpegs has some tasteful hdr photos. You have to click/tap the photo to see the effect.
https://www.dpreview.com/news/7452255382/sigma-brings-hdr-br...
The dad’s photo in the end in SDR looks so much better on a typical desktop IPS panel (Windows 11). The HDR photo looks like the brightness is smushed in the most awful way. On an iPhone, the HDR photo is excellent and the others look muted.
I wonder if there’s an issue in Windows tonemapping or HDR->SDR pipeline, because perceptually the HDR image is really off.
It’s more off than if I took an SDR picture of my iPhone showing the HDR image and showed that SDR picture on the said Windows machine with an IPS panel. Which tells me that the manual HDR->SDR “pipeline” I just described is better.
I think Windows showing HDR content on a non-HDR display should just pick an SDR-sized section of that long dynamic range and show it normally. Without trying to remap the entire large range to a smaller one. Or it should do some other perceptual improvements.
Then again, I know professionally that Windows HDR is complicated and hard to tame. So I’m not really sure the context of remapping as they do, maybe it’s the only way in some contingency/rare scenario.
Question to the author: Why are the images in the "Solution 2: Genuine HDR Displays" section actually videos? E.g.
Your question is answered in the article.
Look for the word video.
Yes but I'm interested in the technical reasons.
Safari doesn't support HDR images and there's no particularly good reason for it.
I'm not entirely convinced that greedy influencers are to blame for people hating on overly bright content. Instead, I think something is different with how displays produce brightness compared to just the nature outside. Light outside is supposed to reach up to tens of thousands of nits, yet even 1000 nits is searing on a display. Is it that displays output polarized light? Is it the spectral distribution of especially the better displays being three really tight peaks? I cannot tell you, but I'm suspecting something isn't quite right.
All this aside, HDR and high brightness are different things - HDR is just a representational thing. You can go full send on your SDR monitor as well, you'll just see more banding. The majority of the article is just content marketing about how they perform automatic tonemapping anyways.
It's all down to the ambient light. That's why bias lighting is now a thing. Try putting a light behind your screen to massively brighten the wall behind it, the 1000 nit peaks will be far less harsh. And if you bring the device out into sunlight I suspect you will wish for everything about its output to be quite a bit brighter.
> Light outside is supposed to reach up to tens of thousands of nits, yet even 1000 nits is searing on a display.
That’s a consequence of https://en.wikipedia.org/wiki/Adaptation_(eye). If you look at 1000 nits on a display in bright sunlight, with your eyes adapted to the bright surroundings, the display would look rather dim.
I find the default HDR (as in gain map) presentation of iPhone photos to look rather garish, rendering highlights too bright and distracting from the content of the images. The solution I came up with for my own camera app was to roll off and lower the highlights in the gain map, which results in final images that I find way more pleasing. This seems to be somewhat similar to what Halide is introducing with their "Standard" option for HDR.
Hopefully HN allows me to share an App Store link... this app works best on Pro iPhones, which support ProRAW, although I do some clever stuff on non-Pro iPhones to get a more natural look.
https://apps.apple.com/us/app/unpro-camera/id6535677796
I see that it says it stores images "in an HDR format by default" but keeps referencing JPEG output. Are you using JPEG-XT? There aren't a lot of "before and after" comparisons so it's hard to know how much it's taking out. I figure those would probably hurt the reputation of the app considering its purpose is to un-pop photos, but I'm in the boat of not really being sure whether I do actually like the pop or not. Is there live-photo support, or is that something that you shouldn't expect from a artist-focused product?
It's a JPEG + gain map format where the gain map is stored in the metadata. Same thing, as far as I can tell, that Halide is now using. It's what the industry is moving towards; it means that images display well on both SDR and HDR displays. I don't know what JPEG-XT is, aside from what I just skimmed on the Wikipedia page.
Not having before-and-after comparisons is mostly down to my being concerned about whether that would pass App Review; the guidelines indicate that the App Store images are supposed to be screenshots of the app, and I'm already pushing that rule with the example images for filters. I'm not sure a hubristic "here's how much better my photos are than Apple's" image would go over well. Maybe in my next update? I should at least have some comparisons on my website, but I've been bad at keeping that updated.
There's no Live Photo support, though I've been thinking about it. The reason is that my current iPhone 14 Pro Max does not support Live Photos while shooting in 48-megapixel mode; the capture process takes too long. I'd have to come up with a compromise such as only having video up to the moment of capture. That doesn't prevent me from implementing it for other iPhones/cameras/resolutions, but I don't like having features unevenly available.
I have similar feeling with HDR to what I have with movie audio mixing. The range is just too big. I think this distaste also amplified in me with age. I appreciate content that makes use of the limited space it gets, and also algorithms that compress the range for me a bit. KMPlayer for example has this on for volume by default, and it makes home movie watching more comfortable, no doubt sacrificing artistic vision in the process. I feel a bit the same with the loudness war - and maybe a lot of other people feel the same too, seeing how compressed audio got. At the very least they don't mind much.
I really appreciate the article. I could feel that they also have a product to present, because of the many references, but it was also very informative besides that.
I do prefer the gain map approach myself. Do a sensible '32 bit' HDR edit, further tweak an SDR version, and export a file that has both.
Creative power is still in your hands versus some tone mapper's guesses at your intent.
Can people go overboard? Sure, but thats something they will do regardless of any hdr or lack thereof.
On an aside its still rough that just about every site that touches gain map (adaptive HDR as this blog calls them) HDR images will lose that metadata if they need to scale, recompress, or transform the images otherwise. Its led me to just make my own site, but also to handle what files a client gets a bit smarter . For instance if a browser doesnt support .jxl or .avif images, im sure it wont want an hdr jpeg either, thats easy to handle on a webserver.
If anyone was hoping for a more technical explanation, I find these pages do a good job explaining the inner workings behind the format
https://docs.krita.org/en/general_concepts/colors/bit_depth....
https://docs.krita.org/en/general_concepts/colors/color_spac...
https://docs.krita.org/en/general_concepts/colors/scene_line...
The article starts by saying HDR can mean different things and gives Apple's HDR vs new TV "HDR" advertsing but doesn't explain at all what the TV HDR means and how it is different, unless I missed something. I always assumed they were the same thing.
For the Halide's updated Image Lab demo about 2/3rd of the way down the page (https://www.lux.camera/content/media/2025/05/skyline-edit-tr...), you made the demo so tall desktop users can't both see the sky & the controls at the same time
A lot of these design flaws are fixed by Firefox's picture in picture option but for some reason, with the way you coded it, the prompt to pop it out as PIP doesn't show up
Did I miss something or is the tone mapper not released yet? I admit I'm multitasking here, but just have an exposure slider in the image lab (or whatever it's called).
Sidebar: I kinda miss when Halide's driving purpose was rapid launch and simplicity. I would almost prefer a zoom function to all of this HDR gymnastics (though, to be clear, Halide is my most-used and most-liked camera app).
EDIT: Ah, I see, it's a Mark III feature. That is not REMOTELY clear in the (very long) post.
Is there a consensus definition of what counts as "HDR" in a display? What is the "standard dynamic range" of a typical TV or computer monitor? Is it roughly the same for devices of the same age?
My understanding is most SDR TVs and computer screens have displays about 200-300 nits (aka cd/m²). Is that the correct measure of the range of the display? The brightest white is 300 nits brighter than the darkest black?
Yes, and not just static but dynamic properties: https://displayhdr.org/performance-criteria/
this isn't a correct definition. human perception of brightness is roughly logarithmic, so it also matters how deep the blacks get. For a good HDR experience, you need a monitor that gets to 600 nits at a bare minimum, but also which can get very close to 0 nits (e.g. via OLED or less optionally local dimming)
So, if cameras have poor dynamic range, how are they getting away with a single exposure? They didn't explain that at all...
Human vision has around 20 stops of static dynamic range. Modern digital cameras can't match human vision— a $90,000 Arri Alexa boasts 17 stops— but they're way better than SDR screens.
Isn't the result of their tone mapping algo similar to adjusting shadow and highlight sliders in other software?
No. When you simply adjust shadow and highlights, you lose local contrast. In an early draft of the post, there was an example, but it was cut for pacing.
As a non-techie I represent the 99.9% of the population who haven't a clue what tone mapping etc. is: NO WAY would we ever touch the various settings possible as opposed to watching the TV/computer screen/etc. as it came out of the box.
This was super awesome. Thanks for this! Especially the HDR photo reveal felt really awesome.
As a photographer, one of the things that draws me to the work of artists like Bill Brandt and Daido Moriyama is their use of low dynamic range, and high contrast. I rarely see an HDR image that is aesthetically interesting.
Interesting these HDR controls of their app Halide. Does Android have similar apps?
Back in university I implemented a shoddy HDR for my phone camera.
The hardest part of it, by far, was taking hundreds upon hundreds of pictures of a blank piece of paper in different lighting conditions with different settings.
Is there any workflow that can output HDR photos (like the real HDR kind, with metadata to tell the display to go into HDR mode) for photos shot with a mirrorless and not an iPhone?
Yes. For example, Lightroom and Camera Raw support HDR editing and export from RAW images, and Adobe published a good rundown on the feature when they introduced it.
https://blog.adobe.com/en/publish/2023/10/10/hdr-explained
Greg Benz Photography maintains a list of software here:
https://gregbenzphotography.com/hdr-display-photo-software/
I'm not sure what FOSS options there are; it's difficult to search for given that "HDR" can mean three or four different things in common usage.
This page crashed Brave on Android three times before I gave up.
For me this crashed in android webview, android chrome, and android firefox. Impressive.
Same for me. Pixel 8, chrome
In my opinion, HDR is another marketing gimmick -- the average layman has no idea what it means, but it sounds fancy and is more expensive, so surely it's good.
Half life 2 lost coast was exciting
Glad all this "Instagram influences searing eyeballs with bright whites" is news to me. All I know about is QR code mode doing that.
"The Ed Hardy T-Shirt of Photography"
Literal snort.
I chuckled at "The Ed Hardy t-shirt of photography" for the early, overdone "HDR-mode" images.
i am still skeptical about HDR as pretty much all HDR content I see online is awful. But this post makes me believe that Lux/Halide can pull of HDR in a way that I will like. I am looking forward to Halide Mk3.
Digital camera technology is so fascinating.
Well crap, I had written a thank you, which I will gladly write again:
I love when product announcements and ADS in general are high value works. This one was good education for me. Thank you for it!
I had also written about my plasma and CRT displays and how misunderstandings about HDR made things generally worse and how I probably have not seen the best these 10 bit capable displays can do.
And finally, I had written about 3D TV and how fast, at least 60Hz per eye, 3D in my home made for amazing modeling and assembly experiences! I was very sad to see that tech dead end.
3D for technical content create has a lot of legs... if only more people could see it running great...
Thanks again. I appreciate the education.
... something Linux Desktops don't understand and Macs only do well with their own displays with videos. Guess who the winner is on the desktop: Windows oO
Linux does HDR. It's *technically" worked for years but none of the big DEs had it working. Now Ubuntu 25.04 and Fedora 42 (Gnome) have it working out of the box.
ha, interesting. thanks!
HDR on my MacBook means play the youtube video at full brightness no matter how dim I’ve made the screen.
It’s infuriating.
e.g. Open this in macOS Chrome: https://www.youtube.com/watch?v=Gq7H6PI4JF8
Tldr; lots of vivid colors.
[flagged]
HDR is just a scene-referred image using absolute luminance.
No, it isn't. Absolute luminance is a "feature" of PQ specifically used by HDR10(+) and most DolbyVision content (notably the DolbyVision as produced by an iPhone is not PQ, it's not "real" DolbyVision). But this is not the only form of HDR and it's not even the common form for phone cameras. HLG is a lot more popular for cameras and it is not in absolute luminance. The gainmap-based approach that Google, Apple, and Adobe are all using is also very much not absolute luminance, either. In fact that flips it entirely and it's SDR relative instead, which is a much better approach to HDR than what video initially went with.
Ideally in the abstract it could be just that, but in practice it's an umbrella name for many different techniques that provide some aspect of that goal.
Not in the more general sense! It can refer to what its acronym spells out directly: Bigger range between dimmest and brightest capabilities of a display, imaging technique etc.
No. HDR can encode high dynamic range because (typically) it uses floating point encoding.
From a technical point of view, HDR is just a set of standards and formats for encoding absolute-luminance scene-referred images and video, along with a set of standards for reproduction.
No. HDR video (and images) don't use floating point encoding. They generally use a higher bit depth (10 bits or more vs 8 bits) to reduce banding and different transfer characteristics (i.e. PQ or HLG vs sRGB or BT.709), in addition to different YCbCr matrices and mastering metadata.
And no, it's not necessarily absolute luminance. PQ is absolute, HLG is not.
Isn’t HLG using floating point(s)?
Also DCI-P3 should fit in here somewhere, as it seems to be the most standardized color space for HDR. I would share more insight, if I had it. I thought I understood color profiles well, but I have encountered some challenges when trying to display in one, edit in another, and print “correctly”. And every device seems to treat color profiles a little bit differently.
> Isn’t HLG using floating point(s)?
All transfer functions can generally work on either integer range or floating point. They basically just describe a curve shape, and you can have that curve be over the range of 0.0-1.0 just as easily as you can over 0-255 or 0-1023.
Extended sRGB is about the only thing that basically requires floating point, as it specifically describes 0.0-1.0 as being equivalent to sRGB and then has a valid range larger than that (you end up with something like -.8 to 2.4 or greater). And representing that in integer domain is conceptually possible but practically not really.
> Also DCI-P3 should fit in here somewhere, as it seems to be the most standardized color space for HDR.
BT2020 is the most standardized color space for HDR. DCI-P3 is the most common color gamut of HDR displays that you can actually afford, however, but that's a smaller gamut than what most HDR profiles expect (HDR10, HDR10+, and "professional" DolbyVision are all BT2020 - a wider gamut than P3). Which also means most HDR content specifies a color gamut it doesn't actually benefit from having as all that HDR content is still authored to only use somewhere between the sRGB and DCI-P3 gamut since that's all anyone who views it will actually have.
You can read the actual HLG spec here: https://www.arib.or.jp/english/html/overview/doc/2-STD-B67v2...
The math uses real numbers but table 2-4 ("Digital representation") discusses how the signal is quantized to/from analog and digital. The signal is quantized to integers.
This same quantization process is done for sRGB, BT.709, BT.2020, etc. so it's not unique to HLG. It's just how digital images/video are stored.
I think most HDR formats do not typically use 32 bit floating point. The first HDR file format I can remember is Greg Ward’s RGBE format, which is also now more commonly known as .HDR and I think is pretty widely used.
https://www.graphics.cornell.edu/~bjw/rgbe.html
It uses a type of floating point, in a way, but it’s a shared 8 bit exponent across all 3 channels, and the channels are still 8 bits each, so the whole thing fits in 32 bits. Even the .txt file description says it’s not “floating point” per-se since that implies IEEE single precision floats.
Cameras and displays don’t typically use floats, and even CG people working in HDR and using, e.g., OpenEXR, might use half floats more often that float.
Some standards do exist, and it’s improving over time, but the ideas and execution of HDR in various ways preceded any standards, so I think it’s not helpful to define HDR as a set of standards. From my perspective working in CG, HDR began as a way to break away from 8 bits per channel RGB, and it included improving both color range and color resolution, and started the discussion of using physical metrics as opposed to relative [0..1] ranges.