Hybrid(Windows/Linux/Mac): Input -> x264/x265/Xvid/VP8/VP9

[ampersand]

Yeah, the audio was the problem. I demuxed the video, converted it with Hybrid and it went fine, 1700kbps. When I remuxed the converted video with the original audio it didn't fit.
I'm guessing it's because of this:
Analyzing video stream (0).
=> Detected stretch factor for track 0 is 4/5.
Prefering container (23.976) over stream frame rate (29.970)
Notice: Video stream raw and container lenght differ, other streched streams might exist.

I've no idea what to do, force 29.970 fps when converting?

[Selur]

Tell Hybrid to prefer the stream instead of the container fps
(Config->Internals->Prefer Original->Frame rate)

Cu Selur

[ampersand]

Thanks!

[locotus]

Just installed last version, but had some problems, attached both Report and debug files.
Seems to my scarce knowledge a kind of sintax problem ordering or finding vapoursynth scripts.

Thankds.

[Selur]

I don't really see the issue.
Debug output seems fine up to the moment when the queue was stopped during audio encoding.

"C:\Program Files\Hybrid\64bit\ffmpeg.exe" -y -threads 4 -loglevel fatal -nostdin -i "D:\AVStoDVD\iId_5_aid_0_lang_en_DELAY_-43ms_2023-03-17@14_38_25_9610_01.aac" -ac 2 -ar 48000 -f sox - | "C:\Program Files\Hybrid\64bit\sox.exe" --multi-threaded --temp "D:\AVStoDVD\2023-03-17@14_38_25_961001" --buffer 524288 -S -t sox - -b 32 -t sox - compand 0.1,0.3 -90,-90,-70,-64,-43,-37,-31,-31,-21,-21,0,-20 0 0 0.1 gain -n | "C:\Program Files\Hybrid\64bit\ffmpeg.exe" -y -threads 4 -f sox -i - -c:a aac -strict -2 -b:a 96k -ar 48000 -channel_layout stereo "D:\AVStoDVD\iId_5_aid_0_lang_en_DELAY_-43ms_2023-03-17@14_38_25_9610_02.aac"

Cu Selur

[locotus]

Still having problems with version 2023.03.17.1,
running win10 updated and Nvidia driver 531.41, uodated last weed,

Here the debug file:

[Selur]

I don't see any direct issue with the Vapoursynth script that is used:

# Imports
import vapoursynth as vs
import os
import ctypes
# Loading Support Files
Dllref = ctypes.windll.LoadLibrary("C:/Program Files/Hybrid/64bit/vsfilters/Support/libfftw3f-3.dll")
import sys
# getting Vapoursynth core
core = vs.core
# Import scripts folder
scriptPath = 'C:/Program Files/Hybrid/64bit/vsscripts'
sys.path.insert(0, os.path.abspath(scriptPath))
# Loading Plugins
core.std.LoadPlugin(path="C:/Program Files/Hybrid/64bit/vsfilters/SharpenFilter/MSmooth/libmsmoosh.dll")
core.std.LoadPlugin(path="C:/Program Files/Hybrid/64bit/vsfilters/Support/EEDI3m.dll")
core.std.LoadPlugin(path="C:/Program Files/Hybrid/64bit/vsfilters/ResizeFilter/nnedi3/vsznedi3.dll")
core.std.LoadPlugin(path="C:/Program Files/Hybrid/64bit/vsfilters/MiscFilter/MiscFilters/MiscFilters.dll")
core.std.LoadPlugin(path="C:/Program Files/Hybrid/64bit/vsfilters/GrainFilter/RemoveGrain/RemoveGrainVS.dll")
core.std.LoadPlugin(path="C:/Program Files/Hybrid/64bit/vsfilters/GrainFilter/AddGrain/AddGrain.dll")
core.std.LoadPlugin(path="C:/Program Files/Hybrid/64bit/vsfilters/DenoiseFilter/FFT3DFilter/fft3dfilter.dll")
core.std.LoadPlugin(path="C:/Program Files/Hybrid/64bit/vsfilters/DenoiseFilter/DFTTest/DFTTest.dll")
core.std.LoadPlugin(path="C:/Program Files/Hybrid/64bit/vsfilters/Support/libmvtools.dll")
core.std.LoadPlugin(path="C:/Program Files/Hybrid/64bit/vsfilters/Support/scenechange.dll")
core.std.LoadPlugin(path="C:/Program Files/Hybrid/64bit/vsfilters/Support/fmtconv.dll")
core.std.LoadPlugin(path="C:/Program Files/Hybrid/64bit/vsfilters/DeinterlaceFilter/TIVTC/libtivtc.dll")
core.std.LoadPlugin(path="C:/Program Files/Hybrid/64bit/vsfilters/SourceFilter/d2vSource/DGDecode.dll")
# Import scripts
import havsfunc
# source: 'D:\Downloads\Working Files\Sporting Blood 1931\VIDEO_TS'
# current color space: YUV420P8, bit depth: 8, resolution: 720x480, fps: 29.97, color matrix: 709, yuv luminance scale: limited, scanorder: telecine (soft)
# Loading D:\Downloads\Working Files\Sporting Blood 1931\VIDEO_TS\VTS_01_1.VOB using DGDecode
clip = core.dgdecode.MPEG2Source("D:/AVStoDVD/2023-03-30@12_57_19_6410.d2v")# 29.97 fps, scanorder: telecine (soft)
# Setting detected color matrix (709).
clip = core.std.SetFrameProps(clip, _Matrix=1)
# Setting color transfer info (709), when it is not set
clip = clip if not core.text.FrameProps(clip,'_Transfer') else core.std.SetFrameProps(clip, _Transfer=1)
# Setting color primaries info (bt709), when it is not set
clip = clip if not core.text.FrameProps(clip,'_Primaries') else core.std.SetFrameProps(clip, _Primaries=1)
# Setting color range to TV (limited) range.
clip = core.std.SetFrameProp(clip=clip, prop="_ColorRange", intval=1)
# making sure frame rate is set to 29.97
clip = core.std.AssumeFPS(clip=clip, fpsnum=30000, fpsden=1001)
clip = core.std.SetFrameProp(clip=clip, prop="_FieldBased", intval=0) # progressive
# Deinterlacing using TIVTC
clip = core.tivtc.TFM(clip=clip)
clip = core.tivtc.TDecimate(clip=clip)# new fps: 23.976
# Making sure content is preceived as frame based
clip = core.std.SetFrameProp(clip=clip, prop="_FieldBased", intval=0) # progressive
# cropping the video to 718x478
clip = core.std.CropRel(clip=clip, left=0, right=2, top=2, bottom=0)
clip = core.std.AddBorders(clip=clip, left=0, right=2, top=0, bottom=2) # add borders to archive mod 4 (vsQTGMCFilter) - 720x480
# Denoising using QTGMC
clip = havsfunc.QTGMC(Input=clip, Preset="Slow", InputType=1, TR2=1, SourceMatch=0, Lossless=0, EZDenoise=0.25, NoisePreset="Fast")
clip = core.std.CropRel(clip=clip, left=0, right=2, top=0, bottom=2) # removing borders (vsQTGMCFilter) -  718x478
# sharpening using MSharpen
clip = core.msmoosh.MSharpen(clip=clip)
# Resizing using 10 - bicubic spline
clip = core.fmtc.resample(clip=clip, kernel="spline16", w=656, h=492, interlaced=False, interlacedd=False) # resolution 656x492# before YUV420P8 after YUV420P16
# adjusting output color from: YUV420P16 to YUV420P8 for NVEncModel
clip = core.resize.Bicubic(clip=clip, format=vs.YUV420P8, range_s="limited", dither_type="error_diffusion")
# set output frame rate to 23.976fps (progressive)
clip = core.std.AssumeFPS(clip=clip, fpsnum=24000, fpsden=1001)
# Output
clip.set_output()

But since NVEnc fails with "y4m: failed to parse y4m header." the issue is likely somehow with the Vapoursynth script or the used filters.
My guess is that it's an issue with the fft3dfilter version that comes with Hybrid.
Try whether the issue is fixed another denoise in QTGMC than fft3dfilter is used.
If that helps replace the fft3dfilter.dll with the 'R2.AC3-AVX: AVX only R2.AC3 release' from https://github.com/AmusementClub/VapourSynth-FFT3DFilter/releases
The currently included fft3dfilter version in Hybrid is the 'R2.AC3 Even faster....' version, which requires AVX2 (didn't know that at the time).
=> my guess is your cpu doesn't support AVX2 and in combination with that fft3dfilter version you run into issues.
(in case your cpu does not support avx either, try https://github.com/myrsloik/VapourSynth-FFT3DFilter/releases/tag/R2)

Cu Selur

[locotus]

Thanks, will try it later and let you know if works.

[locotus]

Seems to be working fine up to know. Thanks again.

[Compositor]

I'm using 3.17.1 to put a keyframe in a nonstandard location via chapter point. It's not working, Hybrid keeps putting the point at the "normal" location 2 frames away. This reminds me of the same problem from a few versions ago where it was discarding the qp file all the time. Debug output attached, hope you can figure out what happened.

Update: The alternate approach that worked last time didn't work this time. When I instructed Hybrid to encode only the first of 2 chapters on the source, it simply did not write the second chapter point on the encoded video. In previous versions, it would have encoded until the second chapter point and then stopped.

Debug output also attached. I think it might have a bit more than just the encode job. But that was just me seeing if clearing the video source could get Hybrid to encode just that chapter properly.

[Selur]

Looking at the encoding call:

x265 --input - --output-depth 10 --y4m --profile main10 --level-idc 4.1 --limit-tu 4 --me star --subme 7 --merange 44 --max-merge 5 --no-early-skip --rskip 0 --no-strong-intra-smoothing --no-open-gop --opt-ref-list-length-pps --keyint 240 --min-keyint 6 --bframes 8 --weightb --b-intra --ref 5 --rc-lookahead 70 --lookahead-slices 6 --qpmax 32 --crf 16.00 --opt-qp-pps --qpfile GENERATED_QP_FILE --limit-refs 0 --rd 6 --rd-refine --psy-rd 1.20 --aq-mode 3 --vbv-maxrate 50000 --vbv-bufsize 50000 --deblock=-2:-2 --no-repeat-headers --range limited --colormatrix bt709 --output "C:\Users\Jimmy C\AppData\Local\Temp\llsss2ncopv-b116_1.265"

it looks fine, but Hybrid simply removes the GENERATED_QP_FILE part since it couldn't generate that file name,...

2023.04.04 - 13:35:04_Windows 10 Version 22H2 (64bit)_2023.03.17.1 - level 9: createJobQPFile,...
2023.04.04 - 13:35:04_Windows 10 Version 22H2 (64bit)_2023.03.17.1 - level 9: qpFileName is empty,.. -> removing keyfile-option,..
2023.04.04 - 13:35:04_Windows 10 Version 22H2 (64bit)_2023.03.17.1 - level 9: removing --qpfile GENERATED_QP_FILE

I can reproduce the issue => looking into it.

[Selur]

send you a link to a new dev version.

[Compositor]

send you a link to a new dev version.

I'm afraid to report that the dev version resulted in no changes. Same output as previous version using both methods, setting chapter point and setting encode length by chapter. Debug reports attached. Now that I know what to look for, I can see that the "removing --qpfile GENERATED_QP_FILE" line is there in the reports.

[Selur]

Argh,... missed a typo. -> working on it

[Selur]

Argh,... missed a typo. -> will send you a new link once I packaged and uploaded the file in a few minutes.

[Compositor]

Something's gone very wrong. From the job status, I can see Hybrid generates tagging info, then the chapter file. Then it just... ends. Declares the job finished without even trying to encode. Debug output attached.

Update: Additional testing has shown that videos without chapters can be encoded. But videos with chapters will just stop on the generating chapter file step.

[Selur]

Looking at the end of the debug output everything seems fine.
Can't reproduce what you describe at all.

[Compositor]

Then why did it stop without doing any encoding?

[Selur]

Please, write a step-by-step what you do and share a small sample which allows to reproduce the issue.

[Compositor]

Take a video file that has no chapters. Encode it, should have no problems. Add chapters to the video. Encode again. If you do it from the Jobs panel, you can see the state of the job instantly goes to Finished at 22.22%. I've attached the video I've been trying to encode. Try adding and removing chapters from it, using mkvtoolnix's chapter editor, then encoding it. Even a single chapter at the start of the video makes it impossible to encode. Remove the entire edition of chapters, and the video can be encoded again.

[Compositor]

Can't stay up any longer. Catch up with you in a few hours.

[Selur]

Can't really reproduce it since adding chapter in Hybrid doesn't work for me :/ no clue why looking into that.
Have to go to bed now. Will look at it more tomorrow.
Rencoding your file without chapters worked. (1)
Rencoding your file keeping the chapters worked (2)
Reencoding the reencoded file (1) without chapters worked. (3)
Reencoding the reencoded file (2) and adding one chapter worked. (4)

more tomorrow

[Compositor]

I never figured out how to use the chapter editor in Hybrid. I use mkvtoolnix to create and edit chapters before encoding.

Seeing that you can't reproduce the problem made me try reinstalling Hybrid. Still not working.

[Selur]

Just had anaother go at it and I managed to reproduce the issue.
Problem is the qpgile creation call isn't created and thus the calls relying on the qpfile can't start since they miss the file.
-> will try to fix it after work

[Selur]

send you a new link

[Compositor]

That put the chapter point right where I wanted it to be! Thanks!

[Selur]

There were two bugs:
a. chapter editing was broken
b. job creation handling of chapters
Happy both are fixed.

[Compositor]

There were two bugs:
a. chapter editing was broken

About that. Like I said, I never figured out how to use Hybrid's chapter editor. I decided to give it a try. I start with a blank chapter queue. I thought I could just enter a start time and name then click the '+' sign to the right of the name box to add to the list. But it seems I can set the miliseconds up to 999 and the seconds up to 3 and that's it.

Disregard that last. You set the editor to be bounded by the length of the video, right?

[Selur]

You set the editor to be bounded by the length of the video, right?

Yup, doesn't make sense to add chapter points behind the length of the video,...

[sophisticles]

Just a heads up.