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 19th July 2021, 18:17   #441  |  Link
benwaggoner
Moderator
 
Join Date: Jan 2006
Location: Portland, OR
Posts: 4,750
Quote:
Originally Posted by hajj_3 View Post
They are clearly trying to make out as though it only deserves a tiny fee so that if it ever goes to court over FRAND costs/terms that they can use this as one of their arguments.
Fair enough. And on further consideration, their analysis might apply well to the market of licensees, who cover a lot more use cases than premium streaming content.
__________________
Ben Waggoner
Principal Video Specialist, Amazon Prime Video

My Compression Book
benwaggoner is offline   Reply With Quote
Old 20th July 2021, 00:47   #442  |  Link
birdie
Artem S. Tashkinov
 
birdie's Avatar
 
Join Date: Dec 2006
Posts: 337
No idea for whom this could be useful for: Tencent cloud now supports H.266.

Last edited by birdie; 20th July 2021 at 00:50.
birdie is offline   Reply With Quote
Old 21st July 2021, 09:38   #443  |  Link
FranceBB
Broadcast Encoder
 
FranceBB's Avatar
 
Join Date: Nov 2013
Location: Royal Borough of Kensington & Chelsea, UK
Posts: 2,883
Quote:
Originally Posted by birdie View Post
No idea for whom this could be useful for: Tencent cloud now supports H.266.
The more the better.
I'd like to see more companies getting ready to adopt the standard as they'll put the pressure on hardware manufacturers into actually supporting the decoding.
Once we get there, it's done, ehehehehe
In the meantime, I'm still eagerly waiting for a public release of x266, but I guess we'll still have to play with the reference encoder for a bit longer...


Quote:
Originally Posted by benwaggoner View Post
Yeah. Bandwidth savings are just part of the value of newer codecs. Being able to improve quality of service to lower bandwidth customers is also huge. And helping customers watch more content under their capped/metered bandwidth is also important. There are also material costs to higher bitrates in content encoding, storage, and CDN cache consumption (lower bitrates means more minutes of content can be cached at the edge).
Yep, although, sadly, most of the times companies go into the money-saving direction as managers don't know anything about this world and the only thing they want is an immediate cost reduction on the CDN, so they use lower bitrates to achieve more or less the same subjective quality. At least, this was the direction I felt our manager was pushing towards back with H.265: data reduction, not quality improvement. On the other hand, lots of things have changed so maybe just maybe this time it's gonna be about quality improvement at the same bitrate. I'll let you know when we'll get there, however this is now handled by a different department called "Technology", so, whatever they're gonna do, I won't get a say on that. What I'll get a say on, though, is communication-wise for broadcast mezzanines. You see, nowadays we don't really send obi-vans (that fire up a beam to the satellite) for news any longer (unless we really need to) and the cameraman are sent there with backpacks that get the uncompressed FULL HD signal from the camera, encode it live in H.265 and transmit it via internet using 5G (or fallback to 4G, 3G, 2G if the network ain't available). This signal is then routed via a matrix to EVS and an hardware MPEG-2 encoder divides the 50p in fields and encodes it to XDCAM-50 (so MPEG-2 50 Mbit/s 25i). Of course, the live encode can't be that good 'cause it has to work within a relatively short time-frame and the delay can be from 1 second (generally 3 seconds on live events) up to 20 seconds if we're really short on bandwidth. We've seen some remarkable improvements going from H.264 to H.265 back then and I expect a similar improvement going from H.265 to H.266. If that's the case, it's gonna be a game changer for live news coverage around the globe.
FranceBB is offline   Reply With Quote
Old 21st July 2021, 17:05   #444  |  Link
VictorLS
Registered User
 
Join Date: Mar 2009
Location: Russian Crimea
Posts: 335
Quote:
Originally Posted by Jamaika View Post
New player video YUView for VVC
https://ient.github.io/YUView/
Can anybody play H.266 (VVC) with YUView (I can't play with it even more simple files in Win7x64 with YUView) or another players (I haven't tried i.e. VLC yet) test file TransportRecord11973-59130659.ts (94 MB) https://disk.yandex.ru/d/eQseiH5X73uZyQ I've recorded during test translation from 19e2 SAT more than year ago? If yes, how?

Last edited by VictorLS; 21st July 2021 at 17:29.
VictorLS is offline   Reply With Quote
Old 21st July 2021, 17:17   #445  |  Link
quietvoid
Registered User
 
Join Date: Jan 2019
Location: Canada
Posts: 570
It probably has to be demuxed into a raw VVC bitstream to feed to YUView, which then uses vvcdec to decode it into YUV.
I'm not sure what software supports the demuxing part though.
__________________
LG C2 OLED | GitHub Projects
quietvoid is offline   Reply With Quote
Old 29th July 2021, 18:06   #446  |  Link
IHateCMDLine
Registered User
 
Join Date: Jul 2021
Posts: 2
.....

Last edited by IHateCMDLine; 30th July 2021 at 22:29.
IHateCMDLine is offline   Reply With Quote
Old 30th July 2021, 17:53   #447  |  Link
FranceBB
Broadcast Encoder
 
FranceBB's Avatar
 
Join Date: Nov 2013
Location: Royal Borough of Kensington & Chelsea, UK
Posts: 2,883
Question number 1: Why is the Encoder using 01% of my 56c/112th Intel Xeon CPU and can we speed it up?

Question number 2: How can I get the same GOP size I would get with x262/x264/x265 when --ref 4 is set and nothing else is? Keyframes should be 24/250 by default, how can I get that here? I can't seem to get --IntraPeriod=1 --GOPSize=1 to work properly 'cause this way they're all intra, but as I change something, it complains about the GOP and it never works...


Quote:
EncoderApp.exe --SummaryVerboseness --InputFile=test.yuv --BitstreamFile=video_2.vvc --SourceWidth=3840 --SourceHeight=2160 --FrameRate=30000/1001 --InputBitDepth=8 --OutputBitDepth=8 --MSBExtendedBitDepth=8 --InternalBitDepth=0 --InputChromaFormat=420 --TargetBitrate=25000 --IntraPeriod=1 --GOPSize=1 --Profile=auto --Level=5.1 --Tier=high --ChromaFormatIDC=420 --MatrixCoefficients=9 --InputColorPrimaries=1 --LMCSSignalType=2 --ConformanceWindowMode=1 --FramesToBeEncoded=6001 --HashME=1 --IBC=1 --DecodingRefreshType=1 --InputSampleRange=1 --AspectRatioInfoPresent=1 --ChromaLocInfoPresent=1 --MaxCUWidth=32 --MaxCUHeight=32 --CTUSize=64 --MaxBTLumaISlice=64 --MaxBTChromaISlice=64 --MaxBTNonISlice=64 --MaxTTLumaISlice=64 --MaxTTChromaISlice=64 --MaxTTNonISlice=64 --CostMode=lossy --BDPCM=0 --ColorTransform=0 --VideoFullRange=0 --ChromaTS=1 --DepQuant=1 --LMCSEnable=1 --RDOQ=1 --RDOQTS=1 --SBT=1 --ISP=1 --MTS=1 --MTSIntraMaxCand=4 --MTSInterMaxCand=4 --LFNST=1 --JointCbCr=1 --SAO=1 --TransformSkip=1 --TransformSkipFast=1 --TransformSkipLog2MaxSize=5 --SAOLcuBoundary=0 --ALF=1 --CCALF=1 --InputChromaFormat=422 --ChromaFormatIDC=422 --ColorTransform=1 --VideoFullRange=1 --ChromaTS=1 --DepQuant=0 --LMCSEnable=0 --RDOQ=0 --RDOQTS=0 --SBT=0 --ISP=0 --MTS=3 --LFNST=0 --JointCbCr=0 --VerCollocatedChroma=1 --SAO=0 --TransformSkip=1 --TransformSkipFast=1 --TransformSkipLog2MaxSize=5 --SAOLcuBoundary=0 --Log2MaxTbSize=5 --ALF=0 --CCALF=0 --BIO=0 --PROF=0 --SearchRange=64 --QpInValCb="17 22 34 42" --QpOutValCb="17 23 35 39" --BCW=0 --BcwFast=0 --BIO=0 --CIIP=0 --Geo=0 --AffineAmvr=0 --LMCSUpdateCtrl=1 --LMCSOffset=0 --DMVR=0 --SMVD=0 --PROF=0 --ISPFast=1 --FastMIP=1 --FastLFNST=1 --FastLocalDualTreeMode=0 --AffineAmvrEncOpt=0 --MmvdDisNum=8 --OnePictureOnlyConstraintFlag=0 --RateControl=0 --SEIDecodedPictureHash=1 --MaxLayers=1 --CbQpOffset=1 --CrQpOffset=1 --TemporalSubsampleRatio=1 --LCTUFast=1 --TemporalFilter=0 --DualITree=1 --MinQTLumaISlice=8 --MinQTChromaISliceInChromaSamples=4 --MinQTNonISlice=8 --MaxMTTHierarchyDepth=3 --MaxMTTHierarchyDepthISliceL=3 --MaxMTTHierarchyDepthISliceC=3 --MMVD=1 --Affine=1 --MaxNumMergeCand=6 --LMChroma=1 --IMV=1 --MRL=1 --IBC=0 --AllowDisFracMMVD=1 --MIP=1 --PBIntraFast=1 --FastMrg=1 --AMaxBT=1 --HadamardME=1 --FEN=1 --FDM=1

pause

So far it's been 6 minutes and it still hasn't encoded a single frame...
The performance is ridiculous and impossible to use on a daily basis.
FranceBB is offline   Reply With Quote
Old 30th July 2021, 19:41   #448  |  Link
benwaggoner
Moderator
 
Join Date: Jan 2006
Location: Portland, OR
Posts: 4,750
Quote:
Originally Posted by FranceBB View Post
Question number 1: Why is the Encoder using 01% of my 56c/112th Intel Xeon CPU and can we speed it up?
Sounds like it is single-threaded. The encoder needs to be updated to support parallelism. x264/x265 demonstrate that a lot of parallelism is in encoders. VVC is based on those, so I expect equivalent maturity VVC encoders to support at least as much parallelism. Maybe more; x265 supports more than x264.

Quote:
Question number 2: How can I get the same GOP size I would get with x262/x264/x265 when --ref 4 is set and nothing else is? Keyframes should be 24/250 by default, how can I get that here? I can't seem to get --IntraPeriod=1 --GOPSize=1 to work properly 'cause this way they're all intra, but as I change something, it complains about the GOP and it never works...
Ref shouldn't have any impact on GOP size. I'd check the documentation.

Quote:
So far it's been 6 minutes and it still hasn't encoded a single frame...
The performance is ridiculous and impossible to use on a daily basis.
You're using the reference encoder? If so, those are used for developing the codec, and are never suitable for production use. A number of vendors are working on VVC encoders with real-world performance.

There is Fraunhofer's https://github.com/fraunhoferhhi/vvenc. It's explicitly meant to offer practical encoding times.

x266 is also in development, although there haven't been any public releases yet. https://multicorewareinc.com/x266-a-...icoreware-inc/

It generally takes a couple of years from the bitstream standard being complete before stable and reasonably good quality/perf general-use encoders become broadly available.
__________________
Ben Waggoner
Principal Video Specialist, Amazon Prime Video

My Compression Book
benwaggoner is offline   Reply With Quote
Old 30th July 2021, 19:58   #449  |  Link
Funky080900
Registered User
 
Join Date: Aug 2019
Posts: 16
Quote:
Question number 2: How can I get the same GOP size I would get with x262/x264/x265 when --ref 4 is set and nothing else is? Keyframes should be 24/250 by default, how can I get that here? I can't seem to get --IntraPeriod=1 --GOPSize=1 to work properly 'cause this way they're all intra, but as I change something, it complains about the GOP and it never works...
VTM doesn't support GOP size of 250 and neither does VVenC.
https://github.com/fraunhoferhhi/vvenc/issues/28

Quote:
Question number 1: Why is the Encoder using 01% of my 56c/112th Intel Xeon CPU and can we speed it up?
VTM by default is single threaded and you have 112 threads, so that's ~1% utilization.
VVenC has presets and multithreading, so you probably want to use that one instead of VTM.

Example: vvencapp -s 3840x2160 -r 30 -c yuv420 --internal-bitdepth 8 --preset faster --threads 16 --qp 32 -ip 256 -o VVC.266 -i input.yuv

Last edited by Funky080900; 30th July 2021 at 20:06.
Funky080900 is offline   Reply With Quote
Old 30th July 2021, 22:15   #450  |  Link
Jamaika
Registered User
 
Join Date: Jul 2015
Posts: 697
Quote:
Originally Posted by Funky080900 View Post
VTM doesn't support GOP size of 250 and neither does VVenC.
VTM has no frames IPB.
Try EVC
xeve_app.exe -i "113.yuv" -o "output_10bit.evc" -v 2 -d 8 -m 4 -b 15 -I 256 --input-csp 1 --codec-bit-depth 10 --rc-type 1 --bitrate 3Mbps -w 1280 -h 720 -z 29 --frames 1000 --preset medium --level-idc 6.3
Jamaika is offline   Reply With Quote
Old 31st July 2021, 18:28   #451  |  Link
LigH
German doom9/Gleitz SuMo
 
LigH's Avatar
 
Join Date: Oct 2001
Location: Germany, rural Altmark
Posts: 6,753
Inverse Telecine is not a part of the encoding core features. It is rather pre-filtering.
__________________

New German Gleitz board
MediaFire: x264 | x265 | VPx | AOM | Xvid
LigH is offline   Reply With Quote
Old 31st July 2021, 22:13   #452  |  Link
FranceBB
Broadcast Encoder
 
FranceBB's Avatar
 
Join Date: Nov 2013
Location: Royal Borough of Kensington & Chelsea, UK
Posts: 2,883
Thanks for the replies.
Of course I know that reference encoders are what they are, but since we have heard nothing for months and months from multicoreware about x266 (aside for the fact that they're working on it and that they won't make it open source until the right time comes), I was looking for a short time replacement to make some comparisons, but apparently the reference encoder is single threaded and most importantly it's Intra only, with no IPB GOP possible, which makes it unusable.

@Jamaika... I'm gonna try with xeve_app.exe. Do you have a place where I can find the x64 Windows builds?

Quote:
Originally Posted by LigH View Post
Inverse Telecine is not a part of the encoding core features. It is rather pre-filtering.
Yep, it's something that has to be done in a frameserver like Avisynth etc, but this actually leads me to an interesting question: is interlace gone for good in H.266, or is it still gonna be around just to piss Ben? XD

Last edited by FranceBB; 31st July 2021 at 22:19.
FranceBB is offline   Reply With Quote
Old 31st July 2021, 23:53   #453  |  Link
quietvoid
Registered User
 
Join Date: Jan 2019
Location: Canada
Posts: 570
Quote:
Originally Posted by FranceBB View Post
we have heard nothing for months and months from multicoreware about x266 (aside for the fact that they're working on it and that they won't make it open source until the right time comes)
There was this video but it's still months ago: https://www.youtube.com/watch?v=vFefmUUzv-c
Might be interesting to you. At least it's more recent than the announcement itself.
__________________
LG C2 OLED | GitHub Projects
quietvoid is offline   Reply With Quote
Old 1st August 2021, 10:30   #454  |  Link
nevcairiel
Registered Developer
 
Join Date: Mar 2010
Location: Hamburg/Germany
Posts: 10,344
Quote:
Originally Posted by FranceBB View Post
is interlace gone for good in H.266, or is it still gonna be around
Seems like same level as support as HEVC, no specific interlace coding tools, but you can signal that a bitstream contains fields instead of frames - so you can encode interlaced that way.
__________________
LAV Filters - open source ffmpeg based media splitter and decoders
nevcairiel is offline   Reply With Quote
Old 1st August 2021, 11:09   #455  |  Link
Jamaika
Registered User
 
Join Date: Jul 2015
Posts: 697
Question number 1: Why is the Encoder using 01% of my 56c/112th Intel Xeon CPU and can we speed it up?
If you used my codec it didn't have SIMD options SSE41, SSE42, AVX, AVX2, AVX512(3). EncoderApp hasn't std::thread function. VTM is only 32bit. No matter if it compiles to x86_64
Are my codecs working properly with multithreading in Windows 10? I don't know, because I don't know what are worth upgrading to gcc 12 in the mingw-std-thread add-on.
https://github.com/Jamaika1/mingw_std_threads
Code:
EncoderApp.exe --SummaryVerboseness --InputFile=test.yuv --BitstreamFile=video_2.vvc --SourceWidth=3840 --SourceHeight=2160 --FrameRate=30000/1001 --InputBitDepth=8 --OutputBitDepth=8 --MSBExtendedBitDep =8 --InternalBitDepth=0 --InputChromaFormat=420 --TargetBitrate=25000 --IntraPeriod=1 --GOPSize=1 --Profile=auto --Level=5.1 --Tier=wysoka --ChromaFormatIDC=420 --MatrixCoefficients=9 --InputColorPrimaries=1 --LMCSSignalType=2 --ConformanceWindowMode=1 --FramesToBeEncoded=6001 --HashME=1 --IBC=1 --DecodingRefreshType= 1 --InputSampleRange=1 --AspectRatioInfoPresent=1 --ChromaLocInfoPresent=1 --MaxCUWidth=32 --MaxCUHeight=32 --CTUSize=64 --MaxBTLumaISlice=64 --MaxBTChromaISlice=64 --MaxBTNonISlice=64 --MaxTTLumaISlice= 64 --MaxTTChromaISlice=64 --MaxTTNonISlice=64 --CostMode=stratna --BDPCM=0 --ColorTransform=0 --VideoFullRange=0 --ChromaTS=1 --DepQuant=1 --LMCSEnable=1 --RDOQ= 1 --RDOQTS=1 --SBT=1 --ISP=1 --MTS=1 --MTSIntraMaxCand=4 --MTSInterMaxCand=4 --LFNST=1 --JointCbCr=1 --SAO=1 --TransformSkip= 1 --TransformSkipFast=1 --TransformSkipLog2MaxSize=5 --SAOLcuBoundary=0 --ALF=1 --CCALF=1 --InputChromaFormat=422 --ChromaFormatIDC=422 --ColorTransform=1 --VideoFullRange=1--ChromaTS=1 --DepQuant=0 --LMCSEnable=0 --RDOQ=0 --RDOQTS=0 --SBT=0 --ISP=0 --MTS=3 --LFNST=0 --JointCbCr=0 --VerCollocatedChroma=1 --SAO=0 --TransformSkip=1 --TransformSkipFast=1 --TransformSkipLog2MaxSize=5 --SAOLcuBoundary=0 --Log2MaxTbSize=5 --ALF=0 --CCALF=0 --BIO=0 --PROF=0 --ZakresWyszukiwania=64 --QpInValCb="17 22 34 42" --QpOutValCb="17 23 35 39"--BCW=0 --BcwFast=0 --BIO=0 --CIIP=0 --Geo=0 --AffineAmvr=0 --LMCSUpdateCtrl=1 --LMCSOffset=0 --DMVR=0 --SMVD=0 --PROF=0 --ISPFast=1 --FastMIP=1 --FastLFNST=1 --FastLocalDualTreeMode=0 --AffineAmvrEncOpt=0 --MmvdDisNum=8 --OnePictureOnlyConstraintFlag=0 --RateControl=0 --SEIDecodedPictureHash=1 --MaxLayers=1 --CbQpOffset=1 --CrQpOffset=1 --TemporalSubsampleRatio=1 --LCTUFast=1 --TemporalFilter=0 --DualITree=1 --MinQTLumaISlice=8 --MinQTChromaISliceInChromaSamples=4 --MinQTNonISlice=8 --MaxMTTHierarchyDepth=3 --MaxMTTHierarchyDepthISliceL=3 --MaxMTTHierarchyDepthISliceC=3 --MMVD=1 --Affine=1 --MaxNumMergeCand=6 --LMChroma=1 --IMV=1 --MRL=1 --IBC=0 --AllowDisFracMMVD=1 --MIP=1 --PBIntraFast=1 --FastMrg=1 --AMaxBT=1 --HadamardME=1 --FEN=1 --FDM=1--Affine=1 --MaxNumMergeCand=6 --LMChroma=1 --IMV=1 --MRL=1 --IBC=0 --AllowDisFracMMVD=1 --MIP=1 --PBIntraFast=1 --FastMrg=1 --AMaxBT=1 --HadamardME=1 --FEN=1 --FDM=1--Affine=1 --MaxNumMergeCand=6 --LMChroma=1 --IMV=1 --MRL=1 --IBC=0 --AllowDisFracMMVD=1 --MIP=1 --PBIntraFast=1 --FastMrg=1 --AMaxBT=1 --HadamardME=1 --FEN=1 --FDM=1
The given functions for encoder are for screenshot losses only. The latest VTM 13.2 codec doesn't work with lossless features. GOP is one. There are no B-frame config given.
Code:
#        Type POC QPoffset QPOffsetModelOff QPOffsetModelScale CbQPoffset CrQPoffset QPfactor tcOffsetDiv2 betaOffsetDiv2 CbTcOffsetDiv2 CbBetaOffsetDiv2 CrTcOffsetDiv2 CrBetaOffsetDiv2 temporal_id #ref_pics_active_L0 #ref_pics_L0   reference_pictures_L0 #ref_pics_active_L1 #ref_pics_L1   reference_pictures_L1
Frame1:   B   16   1        0.0                      0.0            0          0          1.0      0            0                0             0                0               0              0             2                3          16 32 24                    2                2           16 32
Frame2:   B    8   1       -4.8848                   0.2061         0          0          1.0      0            0                0             0                0               0              1             2                2          8 16                        2                2           -8 8
Frame3:   B    4   4       -5.7476                   0.2286         0          0          1.0      0            0                0             0                0               0              2             2                2          4 12                        2                2           -4 -12
Frame4:   B    2   5       -5.90                     0.2333         0          0          1.0      0            0                0             0                0               0              3             2                2          2 10                        2                3           -2 -6 -14
Frame5:   B    1   6       -7.1444                   0.3            0          0          1.0      0            0                0             0                0               0              4             2                2          1 -1                        2                4           -1 -3 -7 -15
Frame6:   B    3   6       -7.1444                   0.3            0          0          1.0      0            0                0             0                0               0              4             2                2          1 3                         2                3           -1 -5 -13
Frame7:   B    6   5       -5.90                     0.2333         0          0          1.0      0            0                0             0                0               0              3             2                2          2 6                         2                2           -2 -10
Frame8:   B    5   6       -7.1444                   0.3            0          0          1.0      0            0                0             0                0               0              4             2                2          1 5                         2                3           -1 -3 -11
Frame9:   B    7   6       -7.1444                   0.3            0          0          1.0      0            0                0             0                0               0              4             2                3          1 3 7                       2                2           -1 -9
Frame10:  B   12   4       -5.7476                   0.2286         0          0          1.0      0            0                0             0                0               0              2             2                2          4 12                        2                2           -4 4
Frame11:  B   10   5       -5.90                     0.2333         0          0          1.0      0            0                0             0                0               0              3             2                2          2 10                        2                2           -2 -6
Frame12:  B    9   6       -7.1444                   0.3            0          0          1.0      0            0                0             0                0               0              4             2                2          1 9                         2                3           -1 -3 -7
Frame13:  B   11   6       -7.1444                   0.3            0          0          1.0      0            0                0             0                0               0              4             2                3          1 3 11                      2                2           -1 -5
Frame14:  B   14   5       -5.90                     0.2333         0          0          1.0      0            0                0             0                0               0              3             2                3          2 6 14                      2                2           -2 2
Frame15:  B   13   6       -7.1444                   0.3            0          0          1.0      0            0                0             0                0               0              4             2                3          1 5 13                      2                2           -1 -3
Frame16:  B   15   6       -7.1444                   0.3            0          0          1.0      0            0                0             0                0               0              4             2                4          1 3 7 15                    2                2           -1 1
The new x266 codec on github doesn't exist. What is invested in vvc?
Openvvc decoder. Everything would be OK, but I can't compile it.
https://github.com/OpenVVC/OpenVVC/tree/dev
Latest untested codecs in my garbage dump.
https://www.sendspace.com/file/4ukanc

Last edited by Jamaika; 1st August 2021 at 11:49.
Jamaika is offline   Reply With Quote
Old 1st August 2021, 12:29   #456  |  Link
Funky080900
Registered User
 
Join Date: Aug 2019
Posts: 16
Quote:
the reference encoder is single threaded and most importantly it's Intra only
VTM isn't intra only. The only limitation is that the intraperiod (keyframe interval) has to be a multiple of the gopsize (16/32). Intraperiod of 240 is possible when gopsize=16. Since 250 isn't a multiple of 32 the closest possible intraperiod is 256. The same applies to VVenC currently.
Funky080900 is offline   Reply With Quote
Old 1st August 2021, 12:59   #457  |  Link
Jamaika
Registered User
 
Join Date: Jul 2015
Posts: 697
Quote:
Originally Posted by Funky080900 View Post
VTM isn't intra only. The only limitation is that the intraperiod (keyframe interval) has to be a multiple of the gopsize (16/32). Intraperiod of 240 is possible when gopsize=16. Since 250 isn't a multiple of 32 the closest possible intraperiod is 256. The same applies to VVenC currently.
This is not true for today.
Code:
Parameter Check Error: GOP Size must be between 1 and 64
Parameter Check Error: GOP list auto config only supported GOP sizes: 1, 16, 32
cannot init the VVenC encoder
VVEncoderApp.exe [error]: cannot create encoder, error -2
Jamaika is offline   Reply With Quote
Old 1st August 2021, 13:10   #458  |  Link
Jamaika
Registered User
 
Join Date: Jul 2015
Posts: 697
Quote:
Originally Posted by IHateCMDLine View Post
Can VVC do 29.97 to 23.97 lower frame rate encoding?
If it is to be divisible by 1001 then you have to add it manually in the codec.
Code:
  switch( vvencappCfg.m_FrameRate )
  {
  case 23: temporalRate = 24000; temporalScale = 1001; break;
  case 29: temporalRate = 30000; temporalScale = 1001; break;
  case 59: temporalRate = 60000; temporalScale = 1001; break;
  default: break;
  }
Jamaika is offline   Reply With Quote
Old 1st August 2021, 13:22   #459  |  Link
Funky080900
Registered User
 
Join Date: Aug 2019
Posts: 16
Well it does say that supported GOP sizes are 1, 16 and 32. GOP size 1 is probably for low delay and all intra. Just leave GOP size at default.
It appears that FranceBB had a problem because he set --IntraPeriod=1 --GOPSize=1. Try --IntraPeriod=256 --GOPSize=32 instead.
Funky080900 is offline   Reply With Quote
Old 1st August 2021, 19:10   #460  |  Link
Jamaika
Registered User
 
Join Date: Jul 2015
Posts: 697
Test codecs:
VVEncoderApp.exe -i "113.yuv" -o "output_10bit.vvc" -v 5 -t 4 -s 1280x720 -r 29 -c yuv420 -g 32 -ip 256 --internal-bitdepth 10 --bitrate 3000000 --passes 1 -f 1000 --preset medium --level 6.3 --tier high
xeve_app.exe -i "113.yuv" -o "output_10abit.evc" -v 3 -d 8 -m 4 -w 1280 -h 720 -z 29 -b 15 -I 256 --input-csp 1 --codec-bit-depth 10 --rc-type 1 --bitrate 3Mbps --frames 1000 --preset medium --level-idc 6.3 --profile main
EncoderApp.exe --SummaryVerboseness -c "encoder_randomaccess_vtm.cfg" --InputFile=113.yuv --BitstreamFile=video_2.vvc --SourceWidth=1280 --SourceHeight=720 --FrameRate=29 --InputBitDepth=8 --InternalBitDepth=12 --OutputBitDepth=10 --MSBExtendedBitDepth=8 --InputChromaFormat=420 --ChromaFormatIDC=420 --ConformanceWindowMode=1 --FramesToBeEncoded=1000 --LMCSSignalType=0 --Level=6.3 --BDPCM=1 --Tier=high --IBC=1 --OverscanInfoPresent=1 --AspectRatioInfoPresent=1 --ChromaLocInfoPresent=1 --Log2MaxTbSize=5 --VerCollocatedChroma=1 --DecodingRefreshType=1

VVEncoderApp is now faster than xeve, but incompatible with newer EncoderApp.
Support for players is currently impossible because decoders don't have Y4M, i.e. they don't contain framerate, colormatrix, output bitdepth information.
Standard xeve baseline profile is fast but has much poorer quality at low bitrates.

Last edited by Jamaika; 1st August 2021 at 19:13.
Jamaika 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 23:07.


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