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. |
12th October 2018, 12:15 | #1121 | Link | |
Registered User
Join Date: Sep 2002
Location: France
Posts: 432
|
Quote:
As for HEVC, the mentioned WPP+frame threading could be available today. All in all, there's likely 30% speed-up left. Source: said capable devs. Last edited by Kurosu; 12th October 2018 at 12:37. |
|
12th October 2018, 16:30 | #1122 | Link | |
Registered User
Join Date: Jun 2002
Posts: 35
|
Quote:
ffmpeg -benchmark -i Stream2_AV1_4K_22.7mbps.webm -f null - ffmpeg version n4.0.2 [libaom-av1 @ 0x55e28fe6f180] 1.0.0-759-g90a15f4f28 frame= 3604 fps=6.2 q=-0.0 Lsize=N/A time=00:02:24.16 bitrate=N/A speed=0.246x video:1886kB audio:0kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: unknown bench: utime=579.607s bench: maxrss=602748kB ffmpeg -threads 4 -benchmark -i Stream2_AV1_4K_22.7mbps.webm -f null - frame= 3604 fps= 16 q=-0.0 Lsize=N/A time=00:02:24.16 bitrate=N/A speed=0.643x video:1886kB audio:0kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: unknown bench: utime=656.087s bench: maxrss=737080kB |
|
13th October 2018, 00:26 | #1123 | Link | ||
Moderator
Join Date: Jan 2006
Location: Portland, OR
Posts: 4,771
|
Quote:
You are right that lots of ffmpeg and other open-source encoder/tool development is corporate funded, and so a lot of its development is driven by what companies want enough to pay for. And HEVC source is still pretty uncommon outside of some high end contribution streams from live events. Quote:
We'll probably have a good sense of the fundamental real-world decoder perf differences between HEVC and AV1 in 2019. |
||
13th October 2018, 06:30 | #1124 | Link | |||
Registered User
Join Date: Sep 2002
Location: France
Posts: 432
|
Quote:
Quote:
Quote:
|
|||
13th October 2018, 16:29 | #1125 | Link |
Registered User
Join Date: Jun 2018
Posts: 7
|
Bitmovin encoder speedup
New article from Streaming Media:
http://www.streamingmedia.com/Articl...ts-127956.aspx On the second page there is an intriguing slide from Bitmovin that with a pure software encoder version of libaom, they attained a speedup for cpu-used=0 from 1000x of VP9 in the early days down to 40x today with improvements ongoing. The claim from AOM was always that there was lots of room for optimization work once the spec was settled and ... well maybe that's happening. The proof will be when those speedups are in mainline FFmpeg and show up in tests on this very thread! The slide also quotes software decode of 3.4x of VP9 single-threaded. |
13th October 2018, 18:27 | #1126 | Link | |||
Registered User
Join Date: Apr 2018
Posts: 63
|
Quote:
The BBC paper used single pass encode Quote:
Quote:
That is basically three results that are only good for academic publication count and the gutter in the real world On another note, there is a 5 minute 720p clip with super diverse content out there now on youtube where VP9 and AV1 version have basically the same file size: https://www.youtube.com/watch?v=WaWnLiffxJ4 You can do some direct comparison of quality Last edited by utack; 13th October 2018 at 22:35. |
|||
14th October 2018, 08:25 | #1127 | Link | ||
Registered User
Join Date: Jul 2018
Posts: 80
|
Quote:
Another fix was merged Quote:
|
||
14th October 2018, 09:41 | #1128 | Link | |
Registered User
Join Date: Aug 2009
Posts: 201
|
Quote:
I'm intrigued to hear more about the AI strategies they are using and wonder how portable those are once the "trick" has been revealed by the AI. |
|
14th October 2018, 19:55 | #1130 | Link |
Registered User
Join Date: May 2005
Location: Swansea, Wales, UK
Posts: 196
|
Personally, considering that BBC is directly involved in development of VVC, and given that they joined the AV1 party relatively late in the development cycle, I'm inclined to find their so called research analysis somewhat suspect.
The Streaming Media article mentions their credibility on the issue, coming from a member of AOMedia, without taking into account just how late they joined. At the very least it seems a conflict of interest to post such a negative paper so early in the post standard development of AV1, even more so considering that VVC is 2 years away from even being standardised. Does anyone here have any idea what level of patent/proposals BBC have in the VVC working group? Edit: Sorry if this is on the wrong thread, it concerns both AV1 and VVC so I wasnt sure which to drop it in. Last edited by soresu; 14th October 2018 at 19:59. |
15th October 2018, 16:53 | #1131 | Link |
Registered User
Join Date: Apr 2016
Posts: 61
|
AV1 Image File Format (AVIF) https://people.xiph.org/~negge/AVIF2018.pdf
Mile-High Video Workshop videos http://mile-high.video/files/mhv2018/ with topics such as:
|
15th October 2018, 19:32 | #1132 | Link | |
Moderator
Join Date: Jan 2006
Location: Portland, OR
Posts: 4,771
|
Quote:
And things have to get LOTS faster to do the psychovisual and rate control tuning to get AV1 into something that can be practically compared to other encoders/bitstreams. Quality won't matter within an order of magnitude of the current speeds. It's not like anyone is actually delivering in volume 1080p encoded with --preset placebo either! Quality @ Bitrate @ Perf! |
|
16th October 2018, 11:05 | #1133 | Link |
I am maddo saientisto!
Join Date: Aug 2018
Posts: 95
|
According to my tests on my F.Y.C clip (1280x720, 480 frames) aomenc right now has got roughly the same encoding quality @ bpp of 1 month ago, with a deviation within the 1%, but is also somewhere between 14-25% faster than 2 months ago for --cpu-used=4
Which means that clip now encodes, in single thread at CQ 20, in 66 minutes rather than 87 on my i7-4770. aomenc 1.0.0-359-g1bc580401: Code:
# aomenc --frame-parallel=0 --tile-columns=0 --auto-alt-ref=2 --cpu-used=4 --passes=2 --threads=1 --lag-in-frames=25 --end-usage=q --cq-level=20 -o test.av1.cq20.webm orig.i420.y4m Pass 1/2 frame 480/481 92352B 1539b/f 36899b/s 21265 ms (22.57 fps) Pass 2/2 frame 480/480 5995490B 99924b/f 2395780b/s 5245119 ms (0.09 fps) Code:
# aomenc --frame-parallel=0 --tile-columns=6 --auto-alt-ref=1 --cpu-used=4 --tune=psnr --passes=2 --threads=1 --end-usage=q --cq-level=20 -o test.av1.cq20.webm orig.i420.y4m Pass 1/2 frame 480/481 92352B 1539b/f 36899b/s 17764 ms (27.02 fps) Pass 2/2 frame 480/480 6092708B 101545b/f 2434645b/s 3969928 ms (0.12 fps) Last edited by SmilingWolf; 16th October 2018 at 11:21. |
16th October 2018, 12:11 | #1134 | Link | |
Registered User
Join Date: Oct 2009
Posts: 930
|
Quote:
|
|
16th October 2018, 16:53 | #1135 | Link |
Moderator
Join Date: Jan 2006
Location: Portland, OR
Posts: 4,771
|
Well, everything about intra coding is relevant to AVIF.
It's interesting to see the use of VMAF for still images. I question its relevance, as VMAF was only calibrated on >=24p moving images. Has anyone published any data suggesting VMAF is useful for measuring still images? Also, the image comparisons versus HEVC don't show the command line. HEVC doesn't have --preset stillimage like x264, but I've tried an adaption of those parameters in x265. I'd expect it would improve the quality of the HEVC (HEIF?) image significantly. I saw some of the new Adobe tools that came out the last few days also include an HEIF exporter, which would be interesting to compare. Looks like fixed QP is being used for both AV1 and HEVC, which is going to be suboptimal for both codecs. Particularly for anything that includes mixed natural/synthetic imagery. It amused me that it ended with a frames-per-minute graph that doesn't reference frame size, nor whether this is for still only or for temporal encoding. The idea of a still image taking 38 seconds to encode is going to horrify anyone using an image server. |
16th October 2018, 17:08 | #1136 | Link | |
Registered User
Join Date: Jul 2018
Posts: 80
|
Quote:
https://www.youtube.com/watch?v=On9VOnIBSEs |
|
16th October 2018, 19:55 | #1137 | Link |
Registered User
Join Date: Jul 2018
Posts: 80
|
Google released today Chrome 70
Chrome 70 adds an AV1 decoder (it's enabled by default) to Chrome Desktop x86-64 based on the official bitstream specification. At this time, support is limited to “Main” profile 0 and does not include encoding capabilities. The supported container is MP4 (ISO-BMFF) (see From raw video to web ready for a brief explanation of containers). To try AV1:
Last edited by marcomsousa; 16th October 2018 at 20:15. |
17th October 2018, 08:19 | #1138 | Link | |
Registered User
Join Date: Aug 2009
Posts: 201
|
Quote:
Bitmovin claims their latest encoder release is twice as fast as stock AV1 and 20x slower than VP9 so things don't seem to be that far off AV1 being delivered on a large scale. They also have the extra bonus of being able to encode their in-house stuff without film grain and add it later, a feature they were keen to have added to AV1. Last edited by dapperdan; 17th October 2018 at 08:55. |
|
17th October 2018, 11:44 | #1139 | Link |
Registered User
Join Date: Oct 2009
Posts: 930
|
Is that still true? My ipression these days from youtube are that the bitrates are quite stable. (Might be wrong)
Stupidest feature ever. They should have implemented an optional "hellyeahIwantnoiseLOL" decoder feature instead. |
17th October 2018, 12:26 | #1140 | Link |
German doom9/Gleitz SuMo
Join Date: Oct 2001
Location: Germany, rural Altmark
Posts: 6,784
|
Don't miss the detail that bitrate optimization for streaming mainly means ABR — and the smaller the assumed decoding buffer and allowed preload time, the closer it gets to CBR, which will cause quality fluctuations. Oh, the joy of the VBV model.
|
|
|