this is a reference decoder, not an optimized one. AV1's reference decoder couldn't do real-time on consumer hardware either, dav1d is what made it practical, and that took years after the spec. The real question is when something like dav1d-for-AV2 shows up, and whether hardware decode blocks land in the next silicon generation, until then this is a proof of spec correctness, not a deployment signal
This demonstration showed real-time video decoding using ARM Neon vector instructions on an Apple CPU at an unspecified frame rate (presumably lower than that achieved on the x86 CPU) and real-time video decoding at 1080p/24 fps, using AVX2 vector instructions on an x86 CPU.
While this is not an impressive performance (compared to 4k/60fps or better for codecs with HW suport), it is still much, much faster than the early AV1 decoders.
I believe that this was precisely the intention of this demonstration, to show that AV2 is already much more mature performance-wise than AV1 ever was in a similar point of its lifetime, as the main early complaint against AV1 was its slowness in the absence of dedicated hardware support.
Hm. Not a lot of technical details about the bitrate improvement of the streams of the CPU required to decode them. I’m also wondering if all the encoding and decoding was done by software reference implementations (just VLC?) or whether anything had any form of hardware assist? It reads as “We did it” without much other information as to how well it went or what AV2’s benefits are over both AV1 and other codecs and whether those benefits were realized in the demonstration or require downstream work to achieve.
EDIT: When looking first at TFA, I did not notice that only the first demonstration was done on an Apple laptop using Neon instructions, but the second demonstration was done on an unnamed laptop with an x86 CPU, thus using the AVX2 vector instructions.
The x86 demo decoded in real time an 1080p/24 fps video stream. Because for Apple the resolution is not specified, we can assume that it was lower than on the x86 laptop.
I think it’s really cool how good the newest video codecs are, but here I am finally able to (only very recently) safely use HEVC and assume that pretty much every end user device has hardware support. It’ll be a while before AV1 reaches that level of adoption and even longer before AV2 does. See you in a few years.
You, as an individual, might be able to use HEVC "safely", hoping that nobody will come after you.
However, during the last few days there have been many news about the owners of the HEVC patents and about the owners of the older H.264 patents making great efforts to extort much more money from the users of HEVC or of H.264.
This has made recently some big vendors of computers to disable the hardware HEVC codecs in the computers that they are selling, instead of paying increased royalties.
At least the H.264 patents have already expired in all countries, except in less than a handful of countries, including USA, so the use of H.264 is safe wherever the patents are no longer valid (but not in USA, where the H.264 patents will remain valid for less than a couple of years).
On the other hand, for HEVC the patents will remain valid for many years, so using it will not be safe even for those who pay royalties, as the royalties may be increased at any time, as shown by the recent history.
Therefore it is wise to avoid HEVC (and its VVC successor) and prefer alternative codecs.
> AV2 is the next-generation video coding specification from the Alliance for Open Media (AOMedia). Building on the foundation of AV1, AV2 is engineered to provide superior compression efficiency, enabling high-quality video delivery at significantly lower bitrates. It is optimized for the evolving demands of streaming, broadcasting, and real-time video conferencing.
> AV2 provides enhanced support for AR/VR applications, split-screen delivery of multiple programs, improved handling of screen content, and an ability to operate over a wider visual quality range.
>Now that the AV2 specification is publicly available
Draft of AV2 spec. Not final. I think they just tagged the AVM 14 release from their research branch. But personally it feels it is no where near final / finish status.
this is a reference decoder, not an optimized one. AV1's reference decoder couldn't do real-time on consumer hardware either, dav1d is what made it practical, and that took years after the spec. The real question is when something like dav1d-for-AV2 shows up, and whether hardware decode blocks land in the next silicon generation, until then this is a proof of spec correctness, not a deployment signal
This demonstration showed real-time video decoding using ARM Neon vector instructions on an Apple CPU at an unspecified frame rate (presumably lower than that achieved on the x86 CPU) and real-time video decoding at 1080p/24 fps, using AVX2 vector instructions on an x86 CPU.
While this is not an impressive performance (compared to 4k/60fps or better for codecs with HW suport), it is still much, much faster than the early AV1 decoders.
I believe that this was precisely the intention of this demonstration, to show that AV2 is already much more mature performance-wise than AV1 ever was in a similar point of its lifetime, as the main early complaint against AV1 was its slowness in the absence of dedicated hardware support.
Hm. Not a lot of technical details about the bitrate improvement of the streams of the CPU required to decode them. I’m also wondering if all the encoding and decoding was done by software reference implementations (just VLC?) or whether anything had any form of hardware assist? It reads as “We did it” without much other information as to how well it went or what AV2’s benefits are over both AV1 and other codecs and whether those benefits were realized in the demonstration or require downstream work to achieve.
In TFA, there are links to the complete specification of AV2 and to the reference software implementation, which was used in the test.
https://gitlab.com/AOMediaCodec/avm/-/tree/research-v13.0.0/...
TFA says that the test was done on an Apple laptop and the decoding was done on the CPU, so not using any special hardware support.
The reference AV2 implementation uses architecture-specific SIMD instructions on x86-64, Aarch64 and IBM POWER.
So in this test it has used the ARM vector ISA (Neon), written with intrinsics in the C language, as it can be seen in the source files:
https://gitlab.com/AOMediaCodec/avm/-/tree/research-v13.0.0/...
EDIT: When looking first at TFA, I did not notice that only the first demonstration was done on an Apple laptop using Neon instructions, but the second demonstration was done on an unnamed laptop with an x86 CPU, thus using the AVX2 vector instructions.
The x86 demo decoded in real time an 1080p/24 fps video stream. Because for Apple the resolution is not specified, we can assume that it was lower than on the x86 laptop.
I think it’s really cool how good the newest video codecs are, but here I am finally able to (only very recently) safely use HEVC and assume that pretty much every end user device has hardware support. It’ll be a while before AV1 reaches that level of adoption and even longer before AV2 does. See you in a few years.
You, as an individual, might be able to use HEVC "safely", hoping that nobody will come after you.
However, during the last few days there have been many news about the owners of the HEVC patents and about the owners of the older H.264 patents making great efforts to extort much more money from the users of HEVC or of H.264.
This has made recently some big vendors of computers to disable the hardware HEVC codecs in the computers that they are selling, instead of paying increased royalties.
At least the H.264 patents have already expired in all countries, except in less than a handful of countries, including USA, so the use of H.264 is safe wherever the patents are no longer valid (but not in USA, where the H.264 patents will remain valid for less than a couple of years).
On the other hand, for HEVC the patents will remain valid for many years, so using it will not be safe even for those who pay royalties, as the royalties may be increased at any time, as shown by the recent history.
Therefore it is wise to avoid HEVC (and its VVC successor) and prefer alternative codecs.
To note that there is x264 to get around the remaining patent issue.
I'm far away from being able to use HEVC on really anything other than my phone. Some patents will have to expire first.
Huh, I didn't know AV2 was out. What are the new features besides (I assume) incremental compression efficiency?
> AV2 is the next-generation video coding specification from the Alliance for Open Media (AOMedia). Building on the foundation of AV1, AV2 is engineered to provide superior compression efficiency, enabling high-quality video delivery at significantly lower bitrates. It is optimized for the evolving demands of streaming, broadcasting, and real-time video conferencing.
> AV2 provides enhanced support for AR/VR applications, split-screen delivery of multiple programs, improved handling of screen content, and an ability to operate over a wider visual quality range.
More info about AV2 here: https://news.ycombinator.com/item?id=45547537
It’s in the first sentence of the article.
Back of the class you go.
>Now that the AV2 specification is publicly available
Draft of AV2 spec. Not final. I think they just tagged the AVM 14 release from their research branch. But personally it feels it is no where near final / finish status.