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 > New and alternative video codecs

Reply
 
Thread Tools Search this Thread Display Modes
Old 27th August 2019, 20:56   #1801  |  Link
benwaggoner
Moderator
 
Join Date: Jan 2006
Location: Portland, OR
Posts: 2,928
Quote:
Originally Posted by soresu View Post
I found a gitlab repo for the dav1d GPU acceleration GSoC, seems like SGR and CDEF have been implemented in Vulkan, and the same repo even has a GLES branch.

Link here.

It will be interesting to see if they can get weaker non ASIC SoC's running well by taking advantage of the previously untapped GPU.
For all the attention encoding on GPU has had over the years, compression with a modern codec is actually about the worst video-related task to run on a GPU. Preprocessing, compositing, and decoding are all much more determinate and parallelizable processes than optimal encoding in complex modern codecs with so many interrelated mode decisions.
__________________
Ben Waggoner
Principal Video Specialist, Amazon Prime Video

My Compression Book
benwaggoner is offline   Reply With Quote
Old 27th August 2019, 21:31   #1802  |  Link
Nintendo Maniac 64
Registered User
 
Nintendo Maniac 64's Avatar
 
Join Date: Nov 2009
Location: Northeast Ohio
Posts: 407
Quote:
Originally Posted by benwaggoner View Post
For all the attention encoding on GPU has had over the years, compression with a modern codec is actually about the worst video-related task to run on a GPU. Preprocessing, compositing, and decoding are all much more determinate and parallelizable processes than optimal encoding in complex modern codecs with so many interrelated mode decisions.
Considering that the post you quoted is referring to dav1d rather than rav1e, I would presume that they were in fact referring to GPU-accelerated decoding rather than GPU-accelerated encoding.
__________________
____HTPC____  | __Desktop PC__
2.93GHz Xeon x3470 (4c/8t Nehalem) | 4.6GHz Pentium G3258 (2c/2t Haswell)
Radeon HD5870  | Intel iGPU      
2x2GB+2x1GB DDR3-1333 | 4x4GB DDR3-1600       

Win7 x64
Nintendo Maniac 64 is offline   Reply With Quote
Old 27th August 2019, 23:54   #1803  |  Link
soresu
Registered User
 
Join Date: May 2005
Location: Swansea, Wales, UK
Posts: 96
Quote:
Originally Posted by Nintendo Maniac 64 View Post
Considering that the post you quoted is referring to dav1d rather than rav1e, I would presume that they were in fact referring to GPU-accelerated decoding rather than GPU-accelerated encoding.
Correct, I understand there are limitations to what the average ARM SoC can do, but leaving the GPU running idle during decode seems a sad waste.
soresu is offline   Reply With Quote
Old 31st August 2019, 21:31   #1804  |  Link
soresu
Registered User
 
Join Date: May 2005
Location: Swansea, Wales, UK
Posts: 96
Given Qualcomm has still yet to join AOM, I wouldn't expect them to do a Hexagon DSP implementation of an AV1 decoder as they did with HEVC back in the day.
soresu is offline   Reply With Quote
Old 1st September 2019, 08:43   #1805  |  Link
NikosD
Registered User
 
Join Date: Aug 2010
Location: Athens, Greece
Posts: 2,587
@benwaggoner

There is no such thing as "GPU encoding" nowadays.
It's an old term referring to the old days where GPGPU processing used for video encoding.

Nowadays all GPUs from the three major vendors (Intel, nVidia, AMD) contain a fixed-function hardware unit, an ASIC, just for the purpose of video decoding/encoding/pre-post processing.

The quality of H.265 hardware encoding of Turing cards aka Turing specific ASIC for video encoding is more than good enough and its speed is out of this world compared to software encoders.
__________________
Win 10 x64 (18362.356) - Core i3-9100F - nVidia 1660 (436.15)
HEVC decoding benchmarks
H.264 DXVA Benchmarks for all
NikosD is offline   Reply With Quote
Old 2nd September 2019, 05:56   #1806  |  Link
Nintendo Maniac 64
Registered User
 
Nintendo Maniac 64's Avatar
 
Join Date: Nov 2009
Location: Northeast Ohio
Posts: 407
Quote:
Originally Posted by NikosD View Post
The quality of H.265 hardware encoding of Turing cards aka Turing specific ASIC for video encoding is more than good enough and its speed is out of this world compared to software encoders.
EposVox has similarly praised both the HEVC encoder on Navi as well as Turing's AVC encoder for being very fast with very good quality as well, especially Navi's HEVC encoder (though has also mentioned that actually getting it to work is a bit of a pain).

(Navi's AVC encoder is still no better than previous AMD GPUs however, meaning you really shouldn't use it)
__________________
____HTPC____  | __Desktop PC__
2.93GHz Xeon x3470 (4c/8t Nehalem) | 4.6GHz Pentium G3258 (2c/2t Haswell)
Radeon HD5870  | Intel iGPU      
2x2GB+2x1GB DDR3-1333 | 4x4GB DDR3-1600       

Win7 x64
Nintendo Maniac 64 is offline   Reply With Quote
Old 3rd September 2019, 19:30   #1807  |  Link
benwaggoner
Moderator
 
Join Date: Jan 2006
Location: Portland, OR
Posts: 2,928
Quote:
Originally Posted by NikosD View Post
@benwaggoner

There is no such thing as "GPU encoding" nowadays.
It's an old term referring to the old days where GPGPU processing used for video encoding.
There are still some products and ongoing experimentation for how to leverage GPU in parallel with CPU for improved encoding. But software has certainly pulled ahead in the last five years.

This could be of particular interest for new codecs like AV1, EVC, and VVC where mature fixed-function implementations aren't yet available. The rapid iteration possible with software is a huge benefit in early-stage codec development and deployment.

Quote:
Nowadays all GPUs from the three major vendors (Intel, nVidia, AMD) contain a fixed-function hardware unit, an ASIC, just for the purpose of video decoding/encoding/pre-post processing.
Of course. And they are essential for things like game streaming where "good enough" quality without taxing GPU or CPU primary processing is needed. But the efficiency is going to be a lot lower than with a good software encode (Like 30%+ higher bitrates required).

Quote:
The quality of H.265 hardware encoding of Turing cards aka Turing specific ASIC for video encoding is more than good enough and its speed is out of this world compared to software encoders.
Yes, certainly. The economics might not make sense for content that gets streamed multiple times, but for personal use when file size is less of a concern than encoding time, there's a place for it.

Although we don't have any GPUs with AV1 fixed function units yet, do we?
__________________
Ben Waggoner
Principal Video Specialist, Amazon Prime Video

My Compression Book
benwaggoner is offline   Reply With Quote
Old 3rd September 2019, 19:39   #1808  |  Link
NikosD
Registered User
 
Join Date: Aug 2010
Location: Athens, Greece
Posts: 2,587
Quote:
Originally Posted by benwaggoner View Post
This could be of particular interest for new codecs like AV1, EVC, and VVC where mature fixed-function implementations aren't yet available. The rapid iteration possible with software is a huge benefit in early-stage codec development and deployment.

Although we don't have any GPUs with AV1 fixed function units yet, do we?
I think there are no HW encoders or decoders for AV1 yet.
And I'm starting to believe that due to the complexity of the codec, it could be the first time that we will not see hybrid (GPU+CPU) decoders/ encoders and we will go straight to fixed-function decoders/encoders of AV1.

Agreed with your post above.
__________________
Win 10 x64 (18362.356) - Core i3-9100F - nVidia 1660 (436.15)
HEVC decoding benchmarks
H.264 DXVA Benchmarks for all
NikosD is offline   Reply With Quote
Old 3rd September 2019, 20:37   #1809  |  Link
NikosD
Registered User
 
Join Date: Aug 2010
Location: Athens, Greece
Posts: 2,587
I think it's already posted, but just as a reminder we could also wait for Vulkan/OpenGL GPU assisted hybrid decoding of dAV1d.
I have my doubts, but OK:
https://www.phoronix.com/scan.php?pa...LES-Experiment
__________________
Win 10 x64 (18362.356) - Core i3-9100F - nVidia 1660 (436.15)
HEVC decoding benchmarks
H.264 DXVA Benchmarks for all
NikosD is offline   Reply With Quote
Old 4th September 2019, 03:44   #1810  |  Link
soresu
Registered User
 
Join Date: May 2005
Location: Swansea, Wales, UK
Posts: 96
Not sure when the GSoC finishes, but he's still posting commits on that branch including CDEF opts.
soresu is offline   Reply With Quote
Old 4th September 2019, 19:06   #1811  |  Link
marcomsousa
Registered User
 
Join Date: Jul 2018
Posts: 51
Quote:
Originally Posted by NikosD View Post
I think there are no HW encoders or decoders for AV1 yet.
Realtek have one HW decoder for av1 - https://www.realtek.com/en/press-roo...-cas-functions

Android 10 is released with the support for Opus audio support and AV1 video codec support- https://www.phoronix.com/scan.php?pa...id-10-Released

Av1 August news - https://www.singhkays.com/blog/av1-e...e-august-2019/
__________________
AV1 win64 VS2017 builds
Last build here | History
I also open source the build scripts at Github: here
marcomsousa is offline   Reply With Quote
Old 4th September 2019, 19:55   #1812  |  Link
NikosD
Registered User
 
Join Date: Aug 2010
Location: Athens, Greece
Posts: 2,587
Quote:
Originally Posted by marcomsousa View Post
Realtek have one HW decoder for av1 - https://www.realtek.com/en/press-roo...-cas-functions
Right!
I had something in mind that I had recently read, so I found it also here, in this thread:
Quote:
Originally Posted by bstrobl View Post
Moreover, I took the chance and found out this recorded video of the chipset in action, decoding in real-time a 4K60fps AV1 clip from YouTube without dropping a frame: https://www.youtube.com/watch?v=IGtjBwMwTtE
Lastly, I even found an announcement of Realtek regarding an integrated circuit RTD2893, capable of 8K AV1 decoding: https://www.realtek.com/en/press-roo...en-awards-copy.

Realtek is first and fast!
__________________
Win 10 x64 (18362.356) - Core i3-9100F - nVidia 1660 (436.15)
HEVC decoding benchmarks
H.264 DXVA Benchmarks for all

Last edited by NikosD; 4th September 2019 at 19:58.
NikosD is offline   Reply With Quote
Old 5th September 2019, 00:20   #1813  |  Link
Blue_MiSfit
Derek Prestegard IRL
 
Blue_MiSfit's Avatar
 
Join Date: Nov 2003
Location: Los Angeles
Posts: 5,518
Quote:
Originally Posted by NikosD View Post
...recorded video of the chipset in action, decoding in real-time a 4K60fps AV1 clip from YouTube without dropping a frame: https://www.youtube.com/watch?v=IGtjBwMwTtE
That's awesome!
Blue_MiSfit is offline   Reply With Quote
Old 5th September 2019, 15:01   #1814  |  Link
soresu
Registered User
 
Join Date: May 2005
Location: Swansea, Wales, UK
Posts: 96
Working AV1 decoder chips are good, I'm guessing they might make it into 2020 TV's.

Of course there are also the forthcoming Rockchip SoC's coming next year - I'm pretty sure one of them (RK3530 I think) was aimed at HDTV's too, with the other (RK3588) likely to make it into cheap Android TV boxes with a bit more oomph than previously seen in that segment, given most non smartphone ARM based kit tops out at low clocked A72/A73 CPU cores.
soresu is offline   Reply With Quote
Old 5th September 2019, 16:13   #1815  |  Link
soresu
Registered User
 
Join Date: May 2005
Location: Swansea, Wales, UK
Posts: 96
On a side note, the experimental branch of AOM has a patch for 'NN based entropy coding' which presumably is a bit more involved than mere rate decision or motion estimation optimisations.

I wonder how it compares to the AV1 entropy coding for quality/complexity, isn't that a derivative of the Daala technique?
soresu is offline   Reply With Quote
Old 5th September 2019, 20:48   #1816  |  Link
marcomsousa
Registered User
 
Join Date: Jul 2018
Posts: 51
Paint.NET adding AV1 (*.avif) decoding in v4.2.2
__________________
AV1 win64 VS2017 builds
Last build here | History
I also open source the build scripts at Github: here
marcomsousa is offline   Reply With Quote
Old 11th September 2019, 03:48   #1817  |  Link
vidschlub
Registered User
 
Join Date: May 2016
Posts: 18
Quote:
Originally Posted by birdie View Post

It's kinda strange we have projects like x264/x265 for patent encumbered H.264/H.265 codecs, yet nothing like that for VP9/AV1.
I am unsure if I previously responded to this but this is more info I wasn't aware of.

Isn't there 3x AV1 encoders?
Are all 3 of them closed source or only designed for mass scale installs on cloud hardware or something?

Why are the 3 AV1 encoders inferior (in some ways?) to the x264 and x265 projects?
vidschlub is offline   Reply With Quote
Old 11th September 2019, 08:05   #1818  |  Link
LigH
German doom9/Gleitz SuMo
 
LigH's Avatar
 
Join Date: Oct 2001
Location: Germany, rural Altmark
Posts: 5,903
How much time has already been spent to develop x264 and x265? And since when are "final" AV1 specifications available? Compare and ask again.
__________________

New German Gleitz board
MediaFire: x264 | x265 | VPx | AOM | Xvid
LigH is offline   Reply With Quote
Old 11th September 2019, 12:06   #1819  |  Link
hajj_3
Registered User
 
Join Date: Mar 2004
Posts: 897
Possible updates about AV1 adoption over the next week: https://aomedia.org/aomedia-members-...-capabilities/
hajj_3 is offline   Reply With Quote
Old 11th September 2019, 18:43   #1820  |  Link
soresu
Registered User
 
Join Date: May 2005
Location: Swansea, Wales, UK
Posts: 96
Quote:
Originally Posted by vidschlub View Post
I am unsure if I previously responded to this but this is more info I wasn't aware of.

Isn't there 3x AV1 encoders?
Are all 3 of them closed source or only designed for mass scale installs on cloud hardware or something?

Why are the 3 AV1 encoders inferior (in some ways?) to the x264 and x265 projects?
There are 3 open source AV1 encoders that I know of - LIBAOM (AOM reference), SVT-AV1 (Intel) and RAV1E (Xiph/Mozilla), in various stages of speed and quality optimisation.

There are other closed source, proprietary encoders like Aurora (ML focused optimisations galore), EVE-AV1 (Two Orioles/Ronald Bultje), and likely the usual customers like Ateme, Cisco and such have their own encoders in various stages at the moment.

The recent Big Apple Video conference gave a run down on most of them save RAV1E, likely due to Vimeo's conflict of interest there as a conference sponsor and a RAV1E contributor/sponsor.

The main point to take away is that x265 has 6-7 years of development, and x264 has closer to 15 years of development. They are both fairly mature, if not necessarilly the fastest options today (SVT-HEVC may change the game on x265 on speed).

Meanwhile AV1 was only standardised last year, it takes time to get these new codecs ship shape for production purposes, let alone significant maturity.
soresu 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 14:34.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2019, vBulletin Solutions Inc.