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 13th February 2020, 13:09   #2061  |  Link
sneaker_ger
Registered User
 
Join Date: Dec 2002
Posts: 5,565
Cool that they seem to go 10 bit even for mobile. It's time companies start to see 10 bit as a coding tool increasing compression instead of using it for HDR only. I would hope they start using AV1 10 bit for desktop, too. I'm temporarily on a slow (5~6 Mbps) connection and there can be horrible banding (among other artifacts) when viewing via Chrome (I assume H.264 8 bit).
sneaker_ger is offline   Reply With Quote
Old 13th February 2020, 20:59   #2062  |  Link
utack
Registered User
 
Join Date: Apr 2018
Posts: 63
I have finished three encodes matching up "vmaf_without_preprocessing" against x265 placebo, and I am completely stunned.
The new vmaf tuning has definitely put libaom in first place not only in metrics but clearly in visual quality!
I hope they will manage to speed it up in the future
utack is offline   Reply With Quote
Old 13th February 2020, 21:43   #2063  |  Link
benwaggoner
Moderator
 
Join Date: Jan 2006
Location: Portland, OR
Posts: 4,750
Quote:
Originally Posted by Blue_MiSfit View Post
Surely streaming only low bitrate SD, with just acceptable quality (since it's a low bandwidth opt-in feature).

Still, a great fit for davi1d!
There's no HW DRM for AV1 on any mobile device yet, so premium content above SD would presumably not be allowed in any case.
__________________
Ben Waggoner
Principal Video Specialist, Amazon Prime Video

My Compression Book
benwaggoner is offline   Reply With Quote
Old 14th February 2020, 03:45   #2064  |  Link
Blue_MiSfit
Derek Prestegard IRL
 
Blue_MiSfit's Avatar
 
Join Date: Nov 2003
Location: Los Angeles
Posts: 5,988
Interesting, correct me if I'm wrong but it's not really a matter of DRM "supporting AV1" as much as it is AV1 having a defined mapping into fMP4 with common encryption (CENC), right? Once that's all set then any DRM that supports CENC would work, be it software Widevine L3 or hardware Widevine L1 or PlayReady SL3000 etc?

Is that mapping still really not done for AV1??
Blue_MiSfit is offline   Reply With Quote
Old 14th February 2020, 13:07   #2065  |  Link
Mr_Khyron
Member
 
Mr_Khyron's Avatar
 
Join Date: Nov 2002
Posts: 203
AVIF for Next-Generation Image Coding
https://netflixtechblog.com/avif-for...ng-b1d75675fe4
Quote:
TL; DR
We need an alternative to JPEG that
a) is widely supported
b) has better compression efficiency and
c) has a wider feature set. We believe AV1 Image File Format (AVIF) has the potential.
Using the framework we have open sourced, AVIF compression efficiency can be seen at work and compared against a whole range of image codecs that came before it.
Mr_Khyron is offline   Reply With Quote
Old 14th February 2020, 18:46   #2066  |  Link
Blue_MiSfit
Derek Prestegard IRL
 
Blue_MiSfit's Avatar
 
Join Date: Nov 2003
Location: Los Angeles
Posts: 5,988
That was a great posting, tbh. AVIF shows serious advantages over WebP which I think is the only JPEG alternative format that's gotten any traction on the web from what I can tell.
Blue_MiSfit is offline   Reply With Quote
Old 16th February 2020, 00:31   #2067  |  Link
dapperdan
Registered User
 
Join Date: Aug 2009
Posts: 201
Quote:
Originally Posted by Blue_MiSfit View Post
That was a great posting, tbh. AVIF shows serious advantages over WebP which I think is the only JPEG alternative format that's gotten any traction on the web from what I can tell.
JPEG XL is a strong contender. It seems to have buy in from a similarly diverse set of implementors as AV1, but one key advantage is that it can seamlessly and losslessly upgrade existing JPEG content, which would probably make it attractive to web deployment even if it didn't offer anything else. If it's well integrated with HTTP2 and browsers then progressive loading is a second "killer app" it provides and again that alone could probably justify its rollout as the subjective improvement in partial loading display would be dramatic even if the objective improvements were equal or even negative.
dapperdan is offline   Reply With Quote
Old 16th February 2020, 19:09   #2068  |  Link
benwaggoner
Moderator
 
Join Date: Jan 2006
Location: Portland, OR
Posts: 4,750
Quote:
Originally Posted by Blue_MiSfit View Post
Interesting, correct me if I'm wrong but it's not really a matter of DRM "supporting AV1" as much as it is AV1 having a defined mapping into fMP4 with common encryption (CENC), right? Once that's all set then any DRM that supports CENC would work, be it software Widevine L3 or hardware Widevine L1 or PlayReady SL3000 etc?

Is that mapping still really not done for AV1??
The mapping is done. But it is really challenging to robustly secure a software-only decoder. It can be done in Trust Zone, but not all devices allow full CPU use for that. And there is still more risk SW being hacked; so much fuzz testing from malformed bitstreams is required, and DRM robustness competes with decode for compute optimization.

Major studios generally require HW DRM integrated with HW decoders for premium HD content. And lots of platforms straight up doing allow their SW codecs to use the HW DRM features. For example, iOS has HEVC software decoders for all device supported by the iOS version that introduced HEVC playback. But FairPlay DRM straight up won't work on the device without the HW decoder.
__________________
Ben Waggoner
Principal Video Specialist, Amazon Prime Video

My Compression Book
benwaggoner is offline   Reply With Quote
Old 16th February 2020, 20:00   #2069  |  Link
benwaggoner
Moderator
 
Join Date: Jan 2006
Location: Portland, OR
Posts: 4,750
Quote:
Originally Posted by dapperdan View Post
JPEG XL is a strong contender. It seems to have buy in from a similarly diverse set of implementors as AV1, but one key advantage is that it can seamlessly and losslessly upgrade existing JPEG content, which would probably make it attractive to web deployment even if it didn't offer anything else. If it's well integrated with HTTP2 and browsers then progressive loading is a second "killer app" it provides and again that alone could probably justify its rollout as the subjective improvement in partial loading display would be dramatic even if the objective improvements were equal or even negative.
HEIF is also a good contender here, and is very well supported in the Apple ecosystem at least. Mostly as HEVC frames. That's the default way to shoot pictures on iPhones now.

But HEVC and patents. AV1 is a strong still image codec, and the patent exposure elminating all the interframe stuff is even smaller. Screen coding tools. And SW decode is a lot more feasible for just images, and DRM is rarely a requirement for them too.

Broad AVIF support in browsers seems like something that could happen quite quickly, and be of use with much less infrastructure than video. A PhotoShop export component and integration into ImageMagick and we'd be ready to go.
__________________
Ben Waggoner
Principal Video Specialist, Amazon Prime Video

My Compression Book
benwaggoner is offline   Reply With Quote
Old 16th February 2020, 21:40   #2070  |  Link
Blue_MiSfit
Derek Prestegard IRL
 
Blue_MiSfit's Avatar
 
Join Date: Nov 2003
Location: Los Angeles
Posts: 5,988
Quote:
Broad AVIF support in browsers seems like something that could happen quite quickly, and be of use with much less infrastructure than video. A PhotoShop export component and integration into ImageMagick and we'd be ready to go.
Seems like a great fit, agreed. Dynamic server-side versioning of images with tools like ImageMagick is a common thing, so this would indeed be an easy win.

Quote:
Major studios generally require HW DRM integrated with HW decoders for premium HD content
Yep! I work for one of them

Quote:
FairPlay DRM straight up won't work on the device without the HW decoder
I actually didn't realize this, but it totally makes sense.

Last edited by Blue_MiSfit; 16th February 2020 at 21:42.
Blue_MiSfit is offline   Reply With Quote
Old 20th February 2020, 17:14   #2071  |  Link
marcomsousa
Registered User
 
Join Date: Jul 2018
Posts: 80
rav1e 0.3.1 released

Quote:
25-40% faster speed levels 2 to 5
This is accomplished by disabling costly coding tools
**Fine directional prediction
**Intra block transform splitting in inter frames
Encoding quality is slightly inferior (1-2%), but more in line with the target speed levels
https://github.com/xiph/rav1e/releases
__________________
AV1 win64 VS2019 builds
Last build here
marcomsousa is offline   Reply With Quote
Old 23rd February 2020, 09:53   #2072  |  Link
Nintendo Maniac 64
Registered User
 
Nintendo Maniac 64's Avatar
 
Join Date: Nov 2009
Location: Northeast Ohio
Posts: 447
Quote:
Originally Posted by benwaggoner View Post
Quote:
Originally Posted by Nintendo Maniac 64 View Post
Unfortunately I didn't provide a test file that combined AVC 10bit and FLAC with other known-working codecs, like AVC 8bit + FLAC as well as AVC 10bit + AAC, so it's uncertain whether it's the AVC 10bit or the FLAC that the TV is unable to play back.
That's the thing about consumer electronics devices. The only way to really know if something works is to actually try it. Something can support all the individual components, but not in particular combinations. Something may work via streaming or an app but not as a file in the native player. Everything might work most of the time except if some esoteric parameter exceeds some internal limit even though it is permitted by the spec.

Something that works on a thing is easy. Something interesting that works on EVERYTHING is a nail-biting adventure.

Apologies for the off-topic post, but I just wanted to give an update on the final conclusion - I just confirmed today that FLAC audio is in fact supported (at least up to 192KHz 24bit, same goes for LPCM as 32float LPCM failed to work) and, as expected, it's the 10bit AVC (not a typo) that is completely unsupported on LG's 2019 OLEDs.


...though more surprising to me was that it not only supported multi-audio MKVs/MP4s with an according GUI-based audio track switcher, but even supported freaking SubStation Alpha subtitles in an MKV also with an according GUI-based toggle.


That is all, and I will refrain from going on about this subject farther.




BTW SmilingWolf, if you're reading this, I got to watch our waifu2x-upscaled UBW Vita OP the LG E9 OLED TV (65" model).

10/10 would recommend.
__________________
____HTPC____  | __Desktop PC__
2.93GHz Xeon x3470 (4c/8t Nehalem) | 4.5GHz 1.24v dual-core Haswell G3258
Radeon HD5870  | Intel iGPU      
2x2GB+2x1GB DDR3-1333 | 4x4GB DDR3-1600       
Nintendo Maniac 64 is offline   Reply With Quote
Old 24th February 2020, 00:17   #2073  |  Link
sneaker_ger
Registered User
 
Join Date: Dec 2002
Posts: 5,565
No AV1 with Tiger Lake? I get that right?
sneaker_ger is offline   Reply With Quote
Old 24th February 2020, 07:25   #2074  |  Link
foxyshadis
ангел смерти
 
foxyshadis's Avatar
 
Join Date: Nov 2004
Location: Lost
Posts: 9,558
Quote:
Originally Posted by sneaker_ger View Post
No AV1 with Tiger Lake? I get that right?
Big delays might open an opportunity to add a decoder, but otherwise, it's going to be implemented in gpu, not fixed-function. It'll look the same to software, it just won't perform the same.
foxyshadis is offline   Reply With Quote
Old 24th February 2020, 09:31   #2075  |  Link
Nintendo Maniac 64
Registered User
 
Nintendo Maniac 64's Avatar
 
Join Date: Nov 2009
Location: Northeast Ohio
Posts: 447
Apologies for the delayed response and for another somewhat off-topic post, but it was only just now that I came across a video with the according WebM-in-chunks formatting that I previously mentioned in this thread and I didn't think it'd be wise to make a whole new thread for a discussion that will probably last all of like two posts.

Quote:
Originally Posted by MoSal View Post
Try passing --youtube-skip-dash-manifest to youtube-dl.
This does not seem to work on the following video's 1080p VP9 encode:

https://www.youtube.com/watch?v=K60qpDSSnJU

(assuming that I'm not doing something wrong of course, which is always a possibility)
__________________
____HTPC____  | __Desktop PC__
2.93GHz Xeon x3470 (4c/8t Nehalem) | 4.5GHz 1.24v dual-core Haswell G3258
Radeon HD5870  | Intel iGPU      
2x2GB+2x1GB DDR3-1333 | 4x4GB DDR3-1600       
Nintendo Maniac 64 is offline   Reply With Quote
Old 24th February 2020, 12:22   #2076  |  Link
nevcairiel
Registered Developer
 
Join Date: Mar 2010
Location: Hamburg/Germany
Posts: 10,344
Quote:
Originally Posted by foxyshadis View Post
It'll look the same to software, it just won't perform the same.
Historically, those have been entirely unusable, especially from Intel. So I hope they won't waste their time on it.
__________________
LAV Filters - open source ffmpeg based media splitter and decoders
nevcairiel is offline   Reply With Quote
Old 24th February 2020, 19:33   #2077  |  Link
MoSal
Registered User
 
Join Date: Jun 2013
Posts: 95
Quote:
Originally Posted by Nintendo Maniac 64 View Post
Apologies for the delayed response and for another somewhat off-topic post, but it was only just now that I came across a video with the according WebM-in-chunks formatting that I previously mentioned in this thread and I didn't think it'd be wise to make a whole new thread for a discussion that will probably last all of like two posts.



This does not seem to work on the following video's 1080p VP9 encode:

https://www.youtube.com/watch?v=K60qpDSSnJU

(assuming that I'm not doing something wrong of course, which is always a possibility)
Yep. You get a URL that upon request returns a 404 error. Maybe that's intentional. Maybe not. We would know for sure if their backend stops spitting such URLs.

Or it could be some boring reason, like the video server requiring certain headers, or the server is, for some reason, tied to a specific old quic/http3 draft. Who knows.

Well, at least audio/Opus unfragmented streams still work.
__________________
https://github.com/MoSal
MoSal is offline   Reply With Quote
Old 26th February 2020, 12:19   #2078  |  Link
hajj_3
Registered User
 
Join Date: Mar 2004
Posts: 1,120
AOM newsletter: https://aomedia.org/wp-content/uploa...0-upadated.pdf
hajj_3 is offline   Reply With Quote
Old 27th February 2020, 11:56   #2079  |  Link
mzso
Registered User
 
Join Date: Oct 2009
Posts: 930
Quote:
Originally Posted by Beelzebubu View Post
Yes.



Could you elaborate on what sort of system (CPU chipset etc.), and where you got the content from? In particular, it'd be interesting to know the respective bitrates for the AV1 & VP9 files/streams, but knowing the encoder settings might also be somewhat useful.

Playback speed correlates a lot with bitrate. The 30% numbers that we've shown at conferences and in blogs are for same-quality encodes, where VP9 has a higher bitrate than AV1. If the files are same-bitrate, the performance difference goes up. On easy content, the postfilters also require a higher % of runtime (compared to e.g. inverse transform or predictors), and since AV1 has more postfilters, that means the difference will grow on low-complexity content, and will be smaller on high-complexity content. The 30% was also without film grain (since we assume the GPU will do that for free), but there is currently no browser that does that correctly yet.
(I have a Ryzen 5 1600, with a B350 board)


I just played youtube videos, with AV1 enabled and disabled, nothing scientific.
But roughly that's what I experienced in cpu usage use.

Last edited by mzso; 27th February 2020 at 12:01.
mzso is offline   Reply With Quote
Old 27th February 2020, 14:34   #2080  |  Link
Beelzebubu
Registered User
 
Join Date: Feb 2003
Location: New York, NY (USA)
Posts: 109
Quote:
Originally Posted by mzso View Post
(I have a Ryzen 5 1600, with a B350 board)
What browser/version, and on what platform/OS?

Quote:
I just played youtube videos, with AV1 enabled and disabled, nothing scientific.
But roughly that's what I experienced in cpu usage use.
I think Youtube is known to do significantly higher bitrates for AV1 than for VP9, so that could be part of why...
Beelzebubu 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 21:40.


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