Welcome to Doom9's Forum, THE in-place to be for everyone interested in DVD conversion.

Before you start posting please read the forum rules. By posting to this forum you agree to abide by the rules.

 

Go Back   Doom9's Forum > Video Encoding > VP9 and AV1

Reply
 
Thread Tools Search this Thread Display Modes
Old 17th May 2021, 21:48   #2401  |  Link
benwaggoner
Moderator
 
Join Date: Jan 2006
Location: Portland, OR
Posts: 4,750
Quote:
Originally Posted by soresu View Post
We went through this months ago when I brought up the AOM GPU software decoder branch for XB1 which uses shaders to work.

Both Sony and MS consoles share the base raster/RT/compute µArch of RDNA2, with each having semi custom optimisations of their own + either their own designed video unit, or an iteration of the AMD VideoCoreNext (VCN) block that predates what went into the PC RDNA2 line up of GPU dies.

Only VCN3 forwards has this support.

That means NV2x currently for AMD, and the upcoming Van Gogh and Rembrandt APUs known to have it from early driver code commits.

This is sadly not that surprising as even the AMD Zen3 Cezanne APU launched just this year still lacks HW ASIC support in its VCN block.

I would expect any Slim or Pro type mid cycle console variants using new chips to have AV1 support though, that would seem a serious blunder if not.
That is great info, thanks!

Quote:
That being said dav1d 0.9 was just released with a massive AVX2 SIMD asm dump for 10+ bpc content, so the current consoles should be more than enough to decode it with 8 Zen2 CPU cores - at least for content like Youtube (or Twitch later) not requiring DRM which is still going to be the majority of AV1 content for some time I think.

Facebook and Netflix sponsored most of the big new AVX2 dump - so clearly they have their eyes on using it for their own BW reducing purposes.
DRM requirements for consoles can be more flexible than for Mac/Win/Android due to the strong virtualization and locked-down nature. One can't run WireShark or a debugger against someone else's code! 720p HD content was allowed on both Xbox 360 and PS3 despite not having fixed-function decoders.
__________________
Ben Waggoner
Principal Video Specialist, Amazon Prime Video

My Compression Book
benwaggoner is offline   Reply With Quote
Old 18th May 2021, 21:04   #2402  |  Link
LigH
German doom9/Gleitz SuMo
 
LigH's Avatar
 
Join Date: Oct 2001
Location: Germany, rural Altmark
Posts: 6,750
New uploads: (MSYS2; MinGW32 / MinGW64: GCC 10.3.0)

AOM v3.1.0-220-g5b35124c3

rav1e 0.5.0-alpha (df8b712b / 2021-05-16)

dav1d 0.9.0-0 (g8636b4f / 2021-05-16)

avif 0.9.0_917cc2c
dav1d [dec]:0.9.0-0-g8636b4f, aom [enc/dec]:3.1.0-218-ga862e2058, rav1e [enc]:0.5.0-alpha (p20210511-46-gdf8b712b)
__________________

New German Gleitz board
MediaFire: x264 | x265 | VPx | AOM | Xvid
LigH is offline   Reply With Quote
Old 20th August 2021, 17:41   #2403  |  Link
BlueLane
Registered User
 
Join Date: Jan 2021
Posts: 2
Quote:
Originally Posted by benwaggoner View Post
Coming to think of it, SW AV1 decoding is actually going to have an impact on global CO2 emissions. A CPU can easily draw 20 more watts in SW decode versus HW decode. 500K simultaneous YouTube viewers watching AV1 could be another 5 MWatt more power consumption and emissions than if YouTube used HEVC. Even assuming low-emissions NG plants, that would be around an extra megaton of global CO2 emissions an hour.

Yowza.
This was wildly incorrect. It's disappointing to see people spreading this kind of misinformation, and that not a single person caught it. It's a huge error to not catch.

Natural gas plants emit about 1 pound of CO2 per kWh, so 1,000 lbs per MWh. I'm being generous and overestimating it a bit – it's more like 0.9 and 900 lbs. (Source: https://www.eia.gov/tools/faqs/faq.php?id=74&t=11) But being conservative, this amounts to 5,000 lbs of CO2 from 5 MWh. You said a "megaton". That's a million tons. The actual figure is less than 2.5 tons. Big difference.

And your 5 MWh figure is wrong if you're basing it on the 20 watt difference you mentioned. 20 watts × 500,000 people × 1 hour = 10 MWh, not 5. So we're up maybe 5 tons now, compared to the 1,000,000 tons you claimed.

Moreover, no numbers here would be relevant to a rational knower without some context that made them relevant. You committed the Large Number Fallacy, which is just to handwave with large, or large-sounding, numbers like a million, megatons, etc., without providing any information that would make those numbers relevant to the non-gullible. In this case, the fallacy is heightened because a megaton of CO2 is trivial in the context of Earth's climate – there are well over 1,000 gigatons of CO2 in Earth's atmosphere, so a single megaton isn't going to matter much, even if it was hourly. Of course, your megaton claim was off by a factor of 200, so the actual amount definitely doesn't matter.

Ideologies and religions reliably corrode people's grip on reality. Caring about mild, long-term climate change predictions is definitely optional, and probably irrational. False inflated stats only serve to amplify the irrational impulses and arbitrary abstractions that are already doing harm. In any case, don't assume everyone else subscribes to your religion, or that everyone is obligated to conform to its dictates.
BlueLane is offline   Reply With Quote
Old 5th September 2021, 02:44   #2404  |  Link
Mr_Khyron
Member
 
Mr_Khyron's Avatar
 
Join Date: Nov 2002
Posts: 203
AV1: Still The Current Future Of Video
https://blog.xaymar.com/2021/08/19/a...ture-of-video/
Quote:
All the way back in December 2020, I decided it was time to try out how far AV1 had progressed. At the time, SVT AV1 was the only encoder that produced reasonable results with near realtime performance, however that has changed now. A lot of work went into AOM AV1, and it is now capable of encoding in the “frames per second” realm instead of “frames per minute”. So why not take another look at things?

For the tests I used footage I captured myself, as that way I have no problems figuring out who actually owns the distribution rights. As for versions of the encoder, NVIDIA NVENC was run with Driver version 471.41 on a RTX 3090, AOM AV1 was compiled at v3.1.2, and SVT AV1 was on compiled at v0.8.6-76-g44486d23. The tests were run with the following settings:........
Mr_Khyron is offline   Reply With Quote
Old 5th September 2021, 09:12   #2405  |  Link
RanmaCanada
Registered User
 
Join Date: May 2009
Posts: 328
Quote:
Originally Posted by BlueLane View Post
This was wildly incorrect. It's disappointing to see people spreading this kind of misinformation, and that not a single person caught it. It's a huge error to not catch.

Natural gas plants emit about 1 pound of CO2 per kWh, so 1,000 lbs per MWh. I'm being generous and overestimating it a bit – it's more like 0.9 and 900 lbs. (Source: https://www.eia.gov/tools/faqs/faq.php?id=74&t=11) But being conservative, this amounts to 5,000 lbs of CO2 from 5 MWh. You said a "megaton". That's a million tons. The actual figure is less than 2.5 tons. Big difference.

And your 5 MWh figure is wrong if you're basing it on the 20 watt difference you mentioned. 20 watts × 500,000 people × 1 hour = 10 MWh, not 5. So we're up maybe 5 tons now, compared to the 1,000,000 tons you claimed.

Moreover, no numbers here would be relevant to a rational knower without some context that made them relevant. You committed the Large Number Fallacy, which is just to handwave with large, or large-sounding, numbers like a million, megatons, etc., without providing any information that would make those numbers relevant to the non-gullible. In this case, the fallacy is heightened because a megaton of CO2 is trivial in the context of Earth's climate – there are well over 1,000 gigatons of CO2 in Earth's atmosphere, so a single megaton isn't going to matter much, even if it was hourly. Of course, your megaton claim was off by a factor of 200, so the actual amount definitely doesn't matter.

Ideologies and religions reliably corrode people's grip on reality. Caring about mild, long-term climate change predictions is definitely optional, and probably irrational. False inflated stats only serve to amplify the irrational impulses and arbitrary abstractions that are already doing harm. In any case, don't assume everyone else subscribes to your religion, or that everyone is obligated to conform to its dictates.
The problem with your post is believing everyone is using Natural Gas. Coal is still used in the majority of the world. In fact almost 40% of the world's electricity is created with coal powered plants.

Sorry.
RanmaCanada is offline   Reply With Quote
Old 10th October 2021, 10:14   #2406  |  Link
birdie
Artem S. Tashkinov
 
birdie's Avatar
 
Join Date: Dec 2006
Posts: 337
AV2 is already showing serious bitstream savings: https://ottverse.com/av2-video-codec-evaluation/
birdie is offline   Reply With Quote
Old 13th October 2021, 00:34   #2407  |  Link
benwaggoner
Moderator
 
Join Date: Jan 2006
Location: Portland, OR
Posts: 4,750
Quote:
Originally Posted by birdie View Post
AV2 is already showing serious bitstream savings: https://ottverse.com/av2-video-codec-evaluation/
I only saw ~7% BD-rate savings in the proposed intra tools under research.

Which is totally fine at such an early state of development. This is all fixed QP, fixed-GOP, no rate control testing comparing the averages of per-frame scores. It's nearly impossible to extrapolate from this early stage what real world savings might be like in practice. I'm sure AV2's gains over AV1 will be >>7%, but we won't know how it compares to VVC in practice until a couple of years after the AV2 standard is completed.
__________________
Ben Waggoner
Principal Video Specialist, Amazon Prime Video

My Compression Book
benwaggoner is offline   Reply With Quote
Old 15th October 2021, 15:21   #2408  |  Link
hajj_3
Registered User
 
Join Date: Mar 2004
Posts: 1,120
Libaom v3.2.0

This release includes compression efficiency and perceptual quality improvements, speedup and memory optimizations, as well as some new features.
- New Features
* Introduced speeds 7, 8, and 9 for all intra mode.
* Introduced speed 10 for real time mode.
* Introduced an API that allows external partition decisions.
* SVC: added support for compound prediction.
* SVC: added support for fixed SVC modes.
- Compression Efficiency Improvements
* Intra-mode search improvement.
* Improved real time (RT) mode BDrate savings by ~5% (RT speed 5)
and ~12% (RT speed 6). The improvement was measured on the video
conference set.
* Improved real time mode for nonrd path (speed 7, 8, 9): BDrate
gains of ~3-5%.
* Rate control and RD adjustments based on ML research in VP9.
Gains of ~0.5-1.0% for HD.
- Perceptual Quality Improvements
* Added a new mode --deltaq-mode=3 to improve perceptual quality
based on a differential contrast model for still images.
* Added a new mode –deltaq-mode=4 to improve perceptual quality
based on user rated cq_level data set for still images.
* Weighting of some intra mode and partition size choices to better
manage and retain texture.
- Speedup and Memory Optimizations
* Further improved 2-pass good quality encoder speed:
o Speed 2 speedup: 18%
o Speed 3 speedup: 22%
o Speed 4 speedup: 37%
o Speed 5 speedup: 30%
o Speed 6 speedup: 20%
* Optimized the real time encoder (measured on the video conference
set):
o RT speed 5 speedup: 110%
o RT speed 6 speedup: 77%
- Bug Fixes
* Issue 3069: Fix one-pass mode keyframe placement off-by-one error.
* Issue 3156: Fix a bug in av1_quantize_lp AVX2 optimization.
hajj_3 is offline   Reply With Quote
Old 20th October 2021, 19:11   #2409  |  Link
benwaggoner
Moderator
 
Join Date: Jan 2006
Location: Portland, OR
Posts: 4,750
Looks like a good point release, particularly for real-time encoding.

I do get nervous about deltaq modes based on still image research, though, as still-image algorithms generally need some tweaking for use with moving images.
__________________
Ben Waggoner
Principal Video Specialist, Amazon Prime Video

My Compression Book
benwaggoner is offline   Reply With Quote
Old 20th October 2021, 20:03   #2410  |  Link
nhw_pulsar
Registered User
 
Join Date: Apr 2017
Posts: 171
Quote:
Originally Posted by benwaggoner View Post
Looks like a good point release, particularly for real-time encoding.

I do get nervous about deltaq modes based on still image research, though, as still-image algorithms generally need some tweaking for use with moving images.
You do get nervous maybe also because inter-frame/motion prediction residuals absolutely don't look like a natural still image.

Still image research is not done for motion prediction residuals, for example my wavelet codec (NHW) puts forward neatness of (natural) images, but I don't know if it'll work as-is with inter-frame residuals which have completely different properties.-It would be challenging (for me) to know if wavelet image codecs could be effectively tuned to work well with a motion prediction scheme and its residuals-... Btw, I'm also waiting eagerly to evaluate this new version of AV1 in AVIF.

Cheers,
Raphael
nhw_pulsar is offline   Reply With Quote
Old 20th October 2021, 20:49   #2411  |  Link
rwill
Registered User
 
Join Date: Dec 2013
Posts: 343
Quote:
Originally Posted by benwaggoner View Post
Looks like a good point release, particularly for real-time encoding.

I do get nervous about deltaq modes based on still image research, though, as still-image algorithms generally need some tweaking for use with moving images.
I am getting nervous when there are quite a couple of deltaq modes to choose from and the encoder developers didn't eliminate all but the 'best overall' for me.
rwill is offline   Reply With Quote
Old 22nd October 2021, 04:24   #2412  |  Link
benwaggoner
Moderator
 
Join Date: Jan 2006
Location: Portland, OR
Posts: 4,750
Quote:
Originally Posted by rwill View Post
I am getting nervous when there are quite a couple of deltaq modes to choose from and the encoder developers didn't eliminate all but the 'best overall' for me.
Well, different modes can be better for different sorts of content. Grainy versus clean film/video versus cel animation/anime versus text/graphics, for example.

The dream is to have those auto-adapt to the content instead of having to pick the least-wrong one for an entire title.
__________________
Ben Waggoner
Principal Video Specialist, Amazon Prime Video

My Compression Book
benwaggoner is offline   Reply With Quote
Old 1st December 2021, 20:53   #2413  |  Link
Yups
Registered User
 
Join Date: Sep 2011
Posts: 362
Quote:
Very concerned about the future of AV1!
No public support on Snapdragon 8 Gen 1 or Apple A15 still.
The two highest volume flagship SoCs lack support.
While Samsung, MediaTek, and Intel have adopted it in latest SoCs, AMD also will not support it in the upcoming Rembrandt APU.
https://twitter.com/dylan522p/status...87898092425220


What is going on with AV1, why is there such a lacklustre support from some major players?
Yups is offline   Reply With Quote
Old 2nd December 2021, 00:59   #2414  |  Link
ksec
Registered User
 
Join Date: Mar 2020
Posts: 117
Quote:
Originally Posted by Yups View Post
https://twitter.com/dylan522p/status...87898092425220


What is going on with AV1, why is there such a lacklustre support from some major players?
It has been explained quite a few times in this thread for the past 2-3 years. But basically die size, cost, ( and profits ) engineering, power usage and politics.

Although I do think Qualcomm would support it next year along side with VVC. The A14 is capable of software decoding 4K AV1 under 1 watts. So if google really want to push it there is no reason why they cant turn on AV1 support on A15 or A16 with software update. But for reference the same 4K on VP9 and HEVC would only require ~150mW on Hardware decoder.
__________________
Previously iwod

Last edited by ksec; 2nd December 2021 at 01:10.
ksec is offline   Reply With Quote
Old 2nd December 2021, 23:47   #2415  |  Link
benwaggoner
Moderator
 
Join Date: Jan 2006
Location: Portland, OR
Posts: 4,750
Quote:
Originally Posted by ksec View Post
It has been explained quite a few times in this thread for the past 2-3 years. But basically die size, cost, ( and profits ) engineering, power usage and politics.

Although I do think Qualcomm would support it next year along side with VVC. The A14 is capable of software decoding 4K AV1 under 1 watts. So if google really want to push it there is no reason why they cant turn on AV1 support on A15 or A16 with software update. But for reference the same 4K on VP9 and HEVC would only require ~150mW on Hardware decoder.
The biggest problem for SW decode is that premium content really drives adoption and usage of new codecs, because quality and efficiency are at a premium and content is of much longer duration. And premium content requires solid HW DRM for HD, and even more secure for 4K. There are platforms where it possible to integrate a SW decoder into DRM, but many can't at all.

And watching 2 hour movie, just one extra watt on a phone can materially reduce battery life.

YouTube, delivering non-DRM'ed, shorter clips with a much lower quality expectation, can use SW decode where streaming service providers like Netflix and Disney+ cannot.
__________________
Ben Waggoner
Principal Video Specialist, Amazon Prime Video

My Compression Book
benwaggoner is offline   Reply With Quote
Old 7th December 2021, 01:54   #2416  |  Link
RanmaCanada
Registered User
 
Join Date: May 2009
Posts: 328
Quote:
Originally Posted by ksec View Post
It has been explained quite a few times in this thread for the past 2-3 years. But basically die size, cost, ( and profits ) engineering, power usage and politics.

Although I do think Qualcomm would support it next year along side with VVC. The A14 is capable of software decoding 4K AV1 under 1 watts. So if google really want to push it there is no reason why they cant turn on AV1 support on A15 or A16 with software update. But for reference the same 4K on VP9 and HEVC would only require ~150mW on Hardware decoder.
There is also the fact that almost no one is paying the programmers to work on it, as most of the development appears to have been done by the community. If Google, MS, Netflix et al were actually serious about this, you figure they would hire a literal army of programmers to solve this. There is also the fact that any encoding to be done with it, is extremely confusing as there is almost no documentation.
RanmaCanada is offline   Reply With Quote
Old 15th December 2021, 17:33   #2417  |  Link
BlueSwordM
Registered User
 
BlueSwordM's Avatar
 
Join Date: Dec 2021
Location: Canada
Posts: 22
What are you talking about?

Most of the people working on AV1 are indeed paid developers by Google, Netflix, Facebook, Visonular, Videolan, Two-Orioles, Intel, etc.

Last edited by BlueSwordM; 15th December 2021 at 19:10. Reason: EVE is a product, not a company :P
BlueSwordM is offline   Reply With Quote
Old 30th December 2021, 11:39   #2418  |  Link
Mr_Khyron
Member
 
Mr_Khyron's Avatar
 
Join Date: Nov 2002
Posts: 203
OBS Studio 27.2 Beta 1

https://github.com/obsproject/obs-st...g/27.2.0-beta1
Quote:
  • Added AOM AV1 and SVT-AV1 encoders (note that these are currently considered experimental, work best with CPUs that have many cores, and are only accessible for recording in advanced output mode)
Mr_Khyron is offline   Reply With Quote
Old 7th January 2022, 01:13   #2419  |  Link
mzso
Registered User
 
Join Date: Oct 2009
Posts: 930
Hmm. I thought AV1 would be further along by now. Did an encode with ffmpeg "-c:v libaom-av1 -crf 35 -cpu-used 3 -tiles 4x3 output.mkv" The output was almost the same size as libx264 -superfast at crf 18. But it took a lot longer and it looks worse.
mzso is offline   Reply With Quote
Old 8th January 2022, 02:32   #2420  |  Link
Blue_MiSfit
Derek Prestegard IRL
 
Blue_MiSfit's Avatar
 
Join Date: Nov 2003
Location: Los Angeles
Posts: 5,988
Something is going wrong then. libaom and svt-av1 absolutely crush x264 at equivalent bitrates as long as you're not using the fastest possible speed presets.
__________________
These are all my personal statements, not those of my employer :)
Blue_MiSfit is offline   Reply With Quote
Reply

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +1. The time now is 09:12.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2024, vBulletin Solutions Inc.