StaxRip

[stax76]

I am in the process of transcoding my ripped DVD library from H.264 to H.265 using StaxRip and there is one thing I wonder if it would be possible to automate. The files have a variety of Color primaries, Transfer characteristics and Matrix coefficients set. Is it possible to have StaxRip automatically read this from the input stream and tell x265 to add it to the output stream?

with the test build 1.3.8.5 you can create a C# script, in the main menu go to Tools > Scripts to open the script folder, save following code with csx extension:


using static StaxRip.ShortcutModule;

string GetMediaInfo(string param) {
return MediaInfo.GetVideo(p.FirstOriginalSourceFile, param);
}

void SetParam(string param) {
g.DefaultCommands.ImportVideoEncoderCommandLine(param);
}

if (GetMediaInfo("colour_primaries") == "BT.709") SetParam("--colorprim bt709");
if (GetMediaInfo("colour_primaries") == "BT.2020") SetParam("--colorprim bt2020");

if (GetMediaInfo("transfer_characteristics") == "BT.709") SetParam("--transfer bt709");
if (GetMediaInfo("transfer_characteristics") == "SMPTE ST 2084") SetParam("--transfer smpte-st-2084");

if (GetMediaInfo("matrix_coefficients") == "BT.709") SetParam("--colormatrix bt709");
if (GetMediaInfo("matrix_coefficients") == "BT.2020 non-constant") SetParam("--colormatrix bt2020nc");


The script can then be called from the main menu, to automate it a event command can be created (Tools > Advanced > Event Commands), there is a command called ExecuteScriptFile.

Probably some community work is needed to support more formats...

Latest QSVEncC v2.54 that you include in your latest 1.3.8.1 build has added an interesting option -avsw.

It can decode in SW video files using ffmpeg (so it's like every possible video file out there) instead of Avisynth or HW decoder.

Do you think you could add this option in the decoders section ?

What interests you in this feature?

@robl45

MakeMKV is great for DVD and BD, StaxRip's eac3to demux GUI for BD folders and M2TS is better then what I've seen in other GUIs so it don't matter if you use MakeMKV or something else, what was the error message you got?

What encoder and settings to use isn't a simple topic, the H.265 hardware encoders are not as mature as the H.264 hardware encoders, you can try to find guides, search to learn about particular parameters and you can make test encodes and compare them, StaxRip offers a comparison tool (Tools > Advanced > Video Comparison).

For subtitle height I think you have to use some editor like Aegisub.

[NikosD]

What interests you in this feature?.

I'm interested in a decoder that doesn't use avisynth and is rock solid (or should be) like ffmpeg.

[stax76]

I'm interested in a decoder that doesn't use avisynth and is rock solid (or should be) like ffmpeg.

Next build will support it.

[NikosD]

Next build will support it.

Hello Frank.

It works like a charm in your latest beta, but not in the context of the whole app.

The way it is implemented is like a switch only in QSVEncC.

Meaning, that when I open a video file, the startup menu of preferred decoders that shows up automatically, doesn't allow me to avoid indexing unless I select DSS2.

There must be a way to select SW decoding without indexing from the beginning.

Thanks

[stax76]

You could use DSS2 in your startup template instead of Manual, I use automatic and demux everything automatic so I don't have any dialogs when I open a file.

[JohnLai]

So..stax76....you might wanna be prepared to add new features from rigaya's nvencc soon.....because nvidia just released nvenc sdk7.0.

What’s new in Video Codec SDK 7.0

HEVC 8K (8192 pixels x 8192 pixels) encoding *
HEVC 4:4:4 encoding *
HEVC 10-bit encoding *
HEVC lossless encoding *
HEVC Sample Adaptive Offset (SAO) *
HEVC Motion-Estimation-(ME)-only mode *
HEVC (up to 8K) decoding *
VP9 (up to 8K) decoding *
HEVC long term reference (LTR) frame support
Asynchronous H.264 Motion-Estimation-(ME)-only mode
Look-ahead
Improved H.264 spatial adaptive quantization
H.264 temporal adaptive quantization
Rate control and quality improvements

* These features require Pascal generation GPUs.



And so.......as I mentioned before...pascal still doesn't support hevc b-frames and 64x64 CU size. It is hardware limitation.

[mcjordan]

NVEnc v2.08 update by rigaya
---
Because finally NVENC SDK 7.0 has been published today, the implementation so that it can be used only part somehow in the rush job. So did not have time, please note that not be barely tested in the fullest only implementation.
Basically, should that become such Lookahead and AQ is available. The bottom line of the 10bit is there also implement the likely time-consuming...
---
And responding to NVENC SDK 7.0
Requires NVIDIA graphics driver 368.69 or later

- Added an option of additional features in SDK 7.0.

--vbr-quality <int> (NVEncC only)
It sets the target quality at the time of VBR mode use. (0-51, 0 = automatic)

--lookahead <int>
To enable the lookahead, to specify the target range by the number of frames. (0-32)
With help to improve the image quality, adaptive I, B frame insertion is enabled.

--no-i-apapt (NVEncC only)
lookahead to disable the adaptive I-frame insertion at the time of effectiveness.

--no-b-apapt (NVEncC only)
lookahead to disable the adaptive B-frames insertion at the time of effectiveness.

--strict-gop (NVEncC only)
To force a fixed GOP length.

--enable-ltr (NVEncC only)
LTR the (Long Term Reference pictures) to enable.

--aq
To enable the adaptive quantization of the frame (spatial direction). (Default: off)

--aq-temporal
To enable the adaptive quantization between the frame (time direction). (Default: off)

--aq-strength <int>
To specify the AQ strength. (1 (weak) to 15 (strong), 0 = automatic)

[JohnLai]

NVEnc v2.08 update by rigaya
---
Because finally NVENC SDK 7.0 has been published today, the implementation so that it can be used only part somehow in the rush job. So did not have time, please note that not be barely tested in the fullest only implementation.
Basically, should that become such Lookahead and AQ is available. The bottom line of the 10bit is there also implement the likely time-consuming...
---
And responding to NVENC SDK 7.0
Requires NVIDIA graphics driver 368.69 or later

- Added an option of additional features in SDK 7.0.

--vbr-quality <int> (NVEncC only)
It sets the target quality at the time of VBR mode use. (0-51, 0 = automatic)

--lookahead <int>
To enable the lookahead, to specify the target range by the number of frames. (0-32)
With help to improve the image quality, adaptive I, B frame insertion is enabled.

--no-i-apapt (NVEncC only)
lookahead to disable the adaptive I-frame insertion at the time of effectiveness.

--no-b-apapt (NVEncC only)
lookahead to disable the adaptive B-frames insertion at the time of effectiveness.

--strict-gop (NVEncC only)
To force a fixed GOP length.

--enable-ltr (NVEncC only)
LTR the (Long Term Reference pictures) to enable.

--aq
To enable the adaptive quantization of the frame (spatial direction). (Default: off)

--aq-temporal
To enable the adaptive quantization between the frame (time direction). (Default: off)

--aq-strength <int>
To specify the AQ strength. (1 (weak) to 15 (strong), 0 = automatic)

Well, the look-ahead option works in CQP mode for both HEVC and H264.

I tested HEVC encoding.

With lookahead 32
GOP length
Maximum GOP length 8000 msec (240 frames)
Minimum GOP length 400 msec (12 frames)
Median of GOP length 2800 msec (84 frames)
Standard deviation of GOP length 2491.28 msec (74.74 frames))

Without lookahead option, I-Frame is inserted every 240 frames.
GOP length
Maximum GOP length 8000 msec (240 frames)
Minimum GOP length 8000 msec (240 frames)
Median of GOP length 8000 msec (240 frames)
Standard deviation of GOP length 0.00 msec (0.00 frames)

Those who enables HEVC Long Term Reference pictures might wanna lower --ref by 1. Example, for 1920x1080, max ref for L4.1 is 5 frames. However, with LTF enabled, max ref will drop to 4 frames.

Edit: Enabling LTR and Look-Ahead 32 cause 20fps drop on GTX970....negligible ~.~

[Leo 69]

Hi Stax,

Would you please add finesharp() to Avisynth+ default selection as you did it for two of other great filters (SMDegrain, LSFMod)? That would be fantastic.
Thanks a lot in advance.

[mcjordan]

NVEnc 2.09 (updated by rigaya)
Before the zombies wandering in search of Pokemon and the time being, including the Pokemon GO, updated to allow up to 10bit encoder in the walk through Pascal. ;-)
HEVC 10bit encoding and HEVC YUV4:4:4 additional encoding.
However, 10bit encoding with the following limitations of NVEncC at present.
- Avcuvid leader is non-compliant
- The 10bit read non-compliant
Since doing the direct conversion of the NVEnc.auo of Aviutl from YC48 to 10bit, the accuracy of the loss is not.

[Magik Mark]

Stax. Up to now there's no deband filter in vapoursynth. Is it possible to port f3kdb to vapoursynth? Thanks a LOT!


Sent from my iPhone using Tapatalk

[sneaker_ger]

It already has VapourSynth support.

http://nmm.me/10z
Example:
ret = core.f3kdb.Deband(ret, y=49, output_depth=16, dither_algo=2)

[Magik Mark]

It already has VapourSynth support.

http://nmm.me/10z
Example:
ret = core.f3kdb.Deband(ret, y=49, output_depth=16, dither_algo=2)

Trying to do this:

clip = core.f3kdb.Deband(clip, keep_tv_range=true)

Getting this error:



Python exception: name 'true' is not defined Traceback (most recent call last):

File "src\cython\vapoursynth.pyx", line 1491, in vapoursynth.vpy_evaluateScript (src\cython\vapoursynth.c:26897)
File "C:\Users\Magik Mark\Desktop\StaxRip Temp Files\Mother's Day (2016)_temp\Mother's Day (2016).vpy", line 6, in <module>
clip = core.f3kdb.Deband(clip, keep_tv_range=true)
NameError: name 'true' is not defined

[Yups]

So..stax76....you might wanna be prepared to add new features from rigaya's nvencc soon.....because nvidia just released nvenc sdk7.0.

What’s new in Video Codec SDK 7.0

HEVC 8K (8192 pixels x 8192 pixels) encoding *
HEVC 4:4:4 encoding *
HEVC 10-bit encoding *
HEVC lossless encoding *
HEVC Sample Adaptive Offset (SAO) *
HEVC Motion-Estimation-(ME)-only mode *
HEVC (up to 8K) decoding *
VP9 (up to 8K) decoding *
HEVC long term reference (LTR) frame support
Asynchronous H.264 Motion-Estimation-(ME)-only mode
Look-ahead
Improved H.264 spatial adaptive quantization
H.264 temporal adaptive quantization
Rate control and quality improvements

* These features require Pascal generation GPUs.



And so.......as I mentioned before...pascal still doesn't support hevc b-frames and 64x64 CU size. It is hardware limitation.


That's a huge update, they have easily the best HEVC hardware encoding in the market now in regards to quality. Intel needs Lookahead support for HEVC to counterattack this. And 10 bit support, but as we know it comes with Kabylake.

[sneaker_ger]

NameError: name 'true' is not defined
Not "true", use "True".

[JohnLai]

That's a huge update, they have easily the best HEVC hardware encoding in the market now in regards to quality. Intel needs Lookahead support for HEVC to counterattack this. And 10 bit support, but as we know it comes with Kabylake.

Except the part where........Nvidia nvenc HEVC only use 1 REFERENCE FRAME for encoding. :(

I have been wondering why file size* doesn't change much even though I specified nvenc to use 4, 8, 12 and 16 reference frames.

Finally, I use hevc analyser to analyse both FFMPEG (SDK 5.0) and Rigaya-san NvencC HEVC bitstream output only to discover NVENC HEVC encoding actually use 1 reference frame only. Not sure if it is a bug or hardware limitation. Currently awaiting Nvidia reply on this.
If NVENC HEVC encoding only use 1 reference frame, then Long-Term Reference is practically useless.

Not to mention......SDK7.0 doesn't include option to TURN ON SAO in first place. Guess Nvidia forgot to include the SAO code for turning it on and off.


Yups, do you have Skylake GPU? Can you provide 50 frames QSV HEVC encoding video sample (1920x1080) ? I want to verify intel hevc standard capability.

[Yups]

Yups, do you have Skylake GPU? Can you provide 50 frames QSV HEVC encoding video sample (1920x1080) ? I want to verify intel hevc standard capability.


Yes I have. One sample is not enough. Intels HEVC performance is very good except for scene change performance. H264 with Lookahead is much better in this regards. Where scene change matters HEVC is pretty poor especially when using a low bitrate. Intel is working on Lookahead support for HEVC at the moment. I hope it is ready in Q4 just before desktop Kabylake is available.

[JohnLai]

Yes I have. One sample is not enough. Intels HEVC performance is very good except for scene change performance. H264 with Lookahead is much better in this regards. Where scene change matters HEVC is pretty poor especially when using a low bitrate. Intel is working on Lookahead support for HEVC at the moment. I hope it is ready in Q4 just before desktop Kabylake is available.

I know QSV HEVC is fast, I just wanna check the analysis of the bitstream to determine which hevc standard is supported by QSV (Example, type of supported PU, CU size, SAO, deblocking, short + long reference frame usage, IPB frame allocation, GOP length and so on)

[easyfab]

I hope it is ready in Q4 just before desktop Kabylake is available.


Peharps with Intel Media Server Studio 2017 that will be out 3rd quarter of 2016 according to https://software.intel.com/en-us/forums/intel-media-sdk/topic/610546

[IbrahimKh]

Hi Stax,

Would you please add finesharp() to Avisynth+ default selection as you did it for two of other great filters (SMDegrain, LSFMod)? That would be fantastic.
Thanks a lot in advance.

I second this+

[Yups]

I know QSV HEVC is fast, I just wanna check the analysis of the bitstream to determine which hevc standard is supported by QSV (Example, type of supported PU, CU size, SAO, deblocking, short + long reference frame usage, IPB frame allocation, GOP length and so on)


Here is a 50 frames sample.

https://www.sendspace.com/file/51879f

[JohnLai]

Here is a 50 frames sample.

https://www.sendspace.com/file/51879f

Nice......

Initial check....QSV HEVC sure......disappointing.....

In term of supported encoding standard ;

Intra PU sizes
4x4
8x8
16x16

Inter PU sizes
4x8
8x4
8x8
8x16
16x8
16x16

Interesting.....
QSV max PU (Prediction Unit) sizes for both Intra and Inter frame is 16x16 while min size is 4x4

For comparison, Nvenc max PU sizes for Intra and Inter is 32x32. Meanwhile, Nvenc min PU sizes for Intra is 4x4 and Inter is 8x8.

x265 can use up to 64x64 max PU sizes and 4x4 min PU sizes for both Intra and Inter. (If one use 'very slow' preset)

Now, it seems like QSV HEVC only support I-Frame and B-Frame. There is no P-Frame being used at all. QSV HEVC B-Frame doesn't support B-Pyramid (Earlier/preceding and Latter/following reference frames). B-Frame only made use of Earlier reference frames. In the sample above, QSV HEVC intelligently uses up to maximum 3 reference frames as needed. Some B-Frame only use 1 and 2 reference frame.

There is no support for SAO and Long-Term Reference for QSV HEVC.

Both Nvenc and QSV HEVC support the generic deblocking filter.

~.~ Darn, both Nvenc and QSV HEVC doesn't fully support the highest possible HEVC specification features.

[Yups]

The above sample was converted with TU7, can you check if TU1 looks identical?

https://www.sendspace.com/file/sxorj7


Maximum reference frames can be changed manually, I prefer 3.

[JohnLai]

The above sample was converted with TU7, can you check if TU1 looks identical?

https://www.sendspace.com/file/sxorj7


Maximum reference frames can be changed manually, I prefer 3.

Now this is interesting. QSV HEVC TU1 sample shows additional PU sizes ;

Intra PU sizes
4x4
8x8
16x16
32x32

Inter PU sizes
4x8
8x4
8x8
8x16
16x8
16x16
32x32

The remaining result of bitstream analysis remains same.


I noticed a weird pattern on both of Intel QSV HEVC samples.

0th frame = I frame = 62.6 KBytes
1st frame = B frame = 29.4
2nd frame = B frame = 14.1
3rd frame = B frame = 16.6
4th frame = B frame = 14.0
5th frame = B frame = 27.3
6th frame = B frame = 14.1
7th frame = B frame = 15.0
8th frame = B frame = 13.2
9th frame = B frame = 28.9

The weird B-Frame pattern from 1st frame to 4th frame repeats itself until end of 49th frame.
It looks as if Intel QSV specifically treat the 1st B-Frame as 'P-Frame' by allocating extra bit.

If we calculate the ratio;
0th frame = I frame = 62.6 --> 100%
1st frame = B frame = 29.4 --> 46.96% of I-frame size
2nd to 4th B frames = (14.1+16.6+14.0)/3= 14.9 = 23.80% of I-frame size.
Isn't the ratio corresponding to ideal I,P,B frame ratio ?



Nvenc full PU sizes on the other hand ;
Intra PU sizes
4x4
8x8
16x16
32x32

Inter PU sizes
8x8
8x16
8x32
16x8
16x16
16x32
24x32
32x8
32x16
32x24
32x32


x265 only use full PU 4x4 and 64x64 on 'super slow preset'.
Intra PU sizes ----> Same as QSV and Nvenc above

Inter PU sizes
4x8 4x16 8x4 8x8
8x16 8x32 12x16 16x4
16x8 16x12 16x16 16x32
16x64 24x32 32x8 32x16
32x24 32x32 32x64 48x64
64x16 64x32 64x48 64x64

[easyfab]

Can we have some ssim and psnr for both nvenc and qsv compare to x264 or x265 .

If you have ffmpeg you can do something like :

ffmpeg -i original.mp4 -i encoded.mp4 -lavfi ssim="stats_file=ssim.log";[0:v][1:v]psnr="stats_file=psnr.log" -f null -

[Vulpix]

@stax76 , thanks for the VCE update - but somehow it breaks the .bat file creation. The file is created without the executable's path, i.e.

@echo off
--profile High --cqp 22:24:27 --sar 12:11 -i <etc etc etc> ...

and the encoding then (obviously) fails.

[JohnLai]

Can we have some ssim and psnr for both nvenc and qsv compare to x264 or x265 .

If you have ffmpeg you can do something like :

ffmpeg -i original.mp4 -i encoded.mp4 -lavfi ssim="stats_file=ssim.log";[0:v][1:v]psnr="stats_file=psnr.log" -f null -

We gonna need Yups to do it because he has all the hardware for H264 and HEVC encoding XD.

Anyway, I am not certain on how nvidia defines its algorithms usage.
Example:
1) Adaptive Quantization shouldn't work with CQP rate control method, but it does work (judging from the encoding frame rate drop and noticeable picture quality changes on I-Frame and P-frame when AQ is enabled)

2) By right, CQP rate control doesn't need lookahead since it use same QP for all frame. Yet, enabling lookahead in CQP mode for nvidia actually lead to I and B frames being adaptively inserted. (Resulting in better quality)

3) Nvenc SDK enables HEVC Long Term Reference frame support, but nvenc hevc uses 1 reference frame which makes the support pointless. How does LTR works if only 1 reference frame being used?

*Point 1 and 2 were mentioned by Dark Shikari in the past (year 2008 I think) where he wrote "using CQP mode instead of CRF disables adaptive quantization and complexity lookahead." Then again, he was talking about software based encoder....


EDIT: Anyone has HEVC 10bit encode sample from nvidia pascal?

[abahram]

Please someone help me to encode my recorded sat file to x264 MKV by staxrip these are sat program recorded from same provider one in SD MPG and one in TS HD i cut theme bye VideoReDo
I want no resize on SD and 1280x720 resize on HD also i save hd as mkv to staxrip reconize audio in file, is this correct? or there is better way to do that?
Please give me proper setting.

forgive me that i am so noob in encoding and my bad english

http://www.mediafire.com/download/61h5wb91t9ov7o6/ManotoHD23-07-2016.03.25.ts


http://www.mediafire.com/download/715jq2rdvdbca3b/MANOTO.22DEC2015.mpg

[Leo 69]

@abahram

Here is the 720p encode of your first file, I did it just for fun :)
http://www.mediafire.com/download/0k87idx1l39d8tk/Manoto720p.mkv

I have no access to the second file, mediafire says "Permission denied".

[abahram]

Dear Leo 69

I need good setting to encode this kind of file, so upload a sample in 1 minute for anyone that can help me with setting, please if you can give me your setting or send me in StaxRip Templates format also I send another link for SD quality thanks.

https://www.upload.ee/files/6005372/MANOTO.22DEC2015.mpg.html

[mcjordan]

NVEncC v2.10 (updated by rigaya)
And some bug fixes, respond to HEVC YUV444 10bit encoding, further support to the high-bit depth input in NVEncC.
Nearly This new feature of NVENC SDK 7.0 is whether not was pick up.

NVEncC / NVEnc.auo common
· Pascal corresponding to HEVC YUV444 10bit encoding.

NVEncC
· Added an option to set the output bit depth. (--output-Depth <int>)
· Avsw / y4m / vpy corresponding to yuv420 input of high bit depth from the reader.
· Avsw / y4m / vpy from the reader yuv444 (8bit, high bit depth) corresponding to the input of.

Examples of high-bit depth input of y4m
ffmpeg -y -i "<mp4 file>" -an -pix_fmt yuv420p10le -f yuv4mpegpipe -strict -1 - | NVEncC --y4m -i - -o "<outfilename.mp4>"

HEVC 10bit encode - how to use:

HEVC Encoding
NVEncC -c hevc -i "<mp4 file>" -o "<outfilename.mp4>"

HEVC 10bit encoding
NVEncC --avsw -c hevc --output-depth 10 -i "<mp4 file>" -o "<outfilename.mp4>"

HEVC YUV444 Encoding
NVEncC --avsw -c hevc --profile main444 -i "<mp4 file>" -o "<outfilename.mp4>"

HEVC YUV444 10bit encoded
NVEncC --avsw -c hevc --output-depth 10 --profile main444 -i "<mp4 file>" -o "<outfilename.mp4>"

With this update, the avsw / y4m / vpy read receipt in YUV420 / YUV444 (8bit or 10bit), became possible output in YUV420 / YUV444 (8bit / 10bit) in accordance with the encoded output you set this. Interconversion is if there is a lot of pattern of this color space, but I honestly head was hurt, well, was able to somehow implementation. (...mass production by using templates ...)

However, this is what you need to do really all CPU, and say, originally also feel it should do on the GPU side. Especially avcuvid read not been able to correspond to the case of (decoded by GPU) current 10bit encoding, etc., since the data is always on the GPU side, just for color space conversion on the GPU is essential.

Not in whether such have to use CUDA is that the implementation ... of the color space conversion on the GPU in the future that ....

NVEnc.auo
· Afs fix did not problems that can HEVC 10bit encoded at the time of use.
· Fixed an issue that is not the progress displayed in the status bar.

Sorry... google translated

[JohnLai]

@stax76
I have 50+ video with 'Undetermined' subtitle that need to be set as 'Default', otherwise MPC-HC and VLC will refuse to autoload the subtitle.
I would like to batch convert those videos, there is only single subtitle in each video.
Is there any option to automatically set the subtitle as 'Default'? (Auto check the 'Default' box)
http://i.imgur.com/OyLlZ8m.jpg

[brumsky]

Does anyone have a tutorial on how to use Vapoursynth with Staxrip? More specifically I'm trying to run this.

import vapoursynth as vs,havsfunc as haf,sys
vid = haf.SMDegrain(vid, tr=3,thSAD=300,RefineMotion=True,contrasharp=True,pel=2) #med grain

I get a script error: Syntax error.



Thanks!

[lordalibaski]

Come across another problem no idea why it's doing this, I have tried 3 different versions of Staxrip

StaxRip_1.2_Technical_Preview
StaxRip-x64-1.3.7.0-stable
StaxRip-x64-1.3.8.7

and still I get the same issue, this issue is with the audio again not like the last issue I had which didn't get resolved but another one which is strange.

If I encode a small part of this video the video properties are below the audio is there but when I encode the whole video there is no audio at all.


Format : MPEG-PS
File size : 1.55 GiB
Duration : 45mn 27s
Overall bit rate mode : Variable
Overall bit rate : 4 892 Kbps

Video
ID : 224 (0xE0)
Format : MPEG Video
Format version : Version 2
Format profile : Main@Main
Format settings, BVOP : Yes
Format settings, Matrix : Custom
Format settings, GOP : Variable
Format settings, picture struc : Frame
Duration : 45mn 27s
Bit rate mode : Variable
Bit rate : 4 539 Kbps
Maximum bit rate : 9 400 Kbps
Width : 720 pixels
Height : 480 pixels
Display aspect ratio : 4:3
Frame rate : 29.970 fps
Standard : NTSC
Color space : YUV
Chroma subsampling : 4:2:0
Bit depth : 8 bits
Scan type : Interlaced
Scan order : Top Field First
Compression mode : Lossy
Bits/(Pixel*Frame) : 0.438
Time code of first frame : 00:00:02:15
Time code source : Group of pictures header
Stream size : 1.44 GiB (93%)
Color primaries : BT.601 NTSC
Transfer characteristics : BT.601
Matrix coefficients : BT.601

Audio
ID : 189 (0xBD)-128 (0x80)
Format : AC-3
Format/Info : Audio Coding 3
Mode extension : CM (complete main)
Format settings, Endianness : Big
Muxing mode : DVD-Video
Duration : 45mn 27s
Bit rate mode : Constant
Bit rate : 256 Kbps
Channel(s) : 2 channels
Channel positions : Front: L R
Sampling rate : 48.0 KHz
Compression mode : Lossy
Delay relative to video : -66ms
Stream size : 83.2 MiB (5%)


Encoded small part of video with audio set at justmux


Format : Matroska
Format version : Version 4 / Version 2
File size : 5.24 MiB
Duration : 58s 496ms
Overall bit rate : 752 Kbps
Encoded date : UTC 2016-07-31 01:24:34
Writing application : mkvmerge v9.3.1 ('Mask Machine') 64bit
Writing library : libebml v1.3.4 + libmatroska v1.4.5
DURATION : 00:00:58.496000000
NUMBER_OF_FRAMES : 1828
NUMBER_OF_BYTES : 1871872
_STATISTICS_WRITING_APP : mkvmerge v9.3.1 ('Mask Machine') 64bit
_STATISTICS_WRITING_DATE_UTC : 2016-07-31 01:24:34
_STATISTICS_TAGS : BPS DURATION NUMBER_OF_FRAMES NUMBER_OF_BYTES

Video
ID : 1
Format : HEVC
Format/Info : High Efficiency Video Coding
Format profile : Main@L3@Main
Codec ID : V_MPEGH/ISO/HEVC
Duration : 58s 492ms
Bit rate : 481 Kbps
Width : 672 pixels
Height : 448 pixels
Display aspect ratio : 4:3
Original display aspect ratio : 4:3
Frame rate mode : Constant
Frame rate : 29.970 fps
Color space : YUV
Chroma subsampling : 4:2:0
Bit depth : 8 bits
Bits/(Pixel*Frame) : 0.053
Stream size : 3.35 MiB (64%)
Writing library : x265 2.0+5-98a948623fdc:[Windows][GCC 5.4.0][64 bit] 8bit+10bit+12bit
Encoding settings : wpp / ctu=64 / min-cu-size=8 / max-tu-size=32 / tu-intra-depth=1 / tu-inter-depth=1 / me=1 / subme=2 / merange=57 / no-rect / no-amp / max-merge=2 / temporal-mvp / no-early-skip / rskip / rdpenalty=0 / no-tskip / no-tskip-fast / strong-intra-smoothing / no-lossless / no-cu-lossless / no-constrained-intra / no-fast-intra / open-gop / no-temporal-layers / interlace=0 / keyint=250 / min-keyint=25 / scenecut=40 / rc-lookahead=20 / lookahead-slices=0 / bframes=4 / bframe-bias=0 / b-adapt=2 / ref=3 / limit-refs=3 / no-limit-modes / weightp / no-weightb / aq-mode=1 / qg-size=32 / aq-strength=1.00 / cbqpoffs=0 / crqpoffs=0 / rd=3 / psy-rd=2.00 / rdoq-level=0 / psy-rdoq=0.00 / no-rd-refine / signhide / deblock=0:0 / sao / no-sao-non-deblock / b-pyramid / cutree / no-intra-refresh / rc=crf / crf=22.0 / qcomp=0.60 / qpmin=0 / qpmax=69 / qpstep=4 / ipratio=1.40 / pbratio=1.30
Default : Yes
Forced : No

Audio
ID : 2
Format : AC-3
Format/Info : Audio Coding 3
Mode extension : CM (complete main)
Format settings, Endianness : Big
Codec ID : A_AC3
Duration : 58s 496ms
Bit rate mode : Constant
Bit rate : 256 Kbps
Channel(s) : 2 channels
Channel positions : Front: L R
Sampling rate : 48.0 KHz
Compression mode : Lossy
Stream size : 1.79 MiB (34%)
Language : English
Default : No
Forced : No



Encode of all video with audio set at justmux


Format : Matroska
Format version : Version 4 / Version 2
File size : 325 MiB
Duration : 45mn 27s
Overall bit rate : 1 001 Kbps
Encoded date : UTC 2016-07-31 01:01:20
Writing application : mkvmerge v9.3.1 ('Mask Machine') 64bit
Writing library : libebml v1.3.4 + libmatroska v1.4.5
DURATION : 00:45:27.725000000
NUMBER_OF_FRAMES : 81750
NUMBER_OF_BYTES : 340637024
_STATISTICS_WRITING_APP : mkvmerge v9.3.1 ('Mask Machine') 64bit
_STATISTICS_WRITING_DATE_UTC : 2016-07-31 01:01:20
_STATISTICS_TAGS : BPS DURATION NUMBER_OF_FRAMES NUMBER_OF_BYTES

Video
ID : 1
Format : HEVC
Format/Info : High Efficiency Video Coding
Format profile : Main@L3@Main
Codec ID : V_MPEGH/ISO/HEVC
Duration : 45mn 27s
Bit rate : 981 Kbps
Width : 672 pixels
Height : 448 pixels
Display aspect ratio : 4:3
Original display aspect ratio : 4:3
Frame rate mode : Constant
Frame rate : 29.970 fps
Color space : YUV
Chroma subsampling : 4:2:0
Bit depth : 8 bits
Bits/(Pixel*Frame) : 0.109
Stream size : 319 MiB (98%)
Writing library : x265 2.0+5-98a948623fdc:[Windows][GCC 5.4.0][64 bit] 8bit+10bit+12bit
Encoding settings : wpp / ctu=64 / min-cu-size=8 / max-tu-size=32 / tu-intra-depth=1 / tu-inter-depth=1 / me=1 / subme=2 / merange=57 / no-rect / no-amp / max-merge=2 / temporal-mvp / no-early-skip / rskip / rdpenalty=0 / no-tskip / no-tskip-fast / strong-intra-smoothing / no-lossless / no-cu-lossless / no-constrained-intra / no-fast-intra / open-gop / no-temporal-layers / interlace=0 / keyint=250 / min-keyint=25 / scenecut=40 / rc-lookahead=20 / lookahead-slices=0 / bframes=4 / bframe-bias=0 / b-adapt=2 / ref=3 / limit-refs=3 / no-limit-modes / weightp / no-weightb / aq-mode=1 / qg-size=32 / aq-strength=1.00 / cbqpoffs=0 / crqpoffs=0 / rd=3 / psy-rd=2.00 / rdoq-level=0 / psy-rdoq=0.00 / no-rd-refine / signhide / deblock=0:0 / sao / no-sao-non-deblock / b-pyramid / cutree / no-intra-refresh / rc=crf / crf=20.0 / qcomp=0.60 / qpmin=0 / qpmax=69 / qpstep=4 / ipratio=1.40 / pbratio=1.30
Default : Yes
Forced : No



Why is it not converting the audio when I try the whole video!!!

[lordalibaski]

Sorry for the double post I can't edit my post yet.

Just encoded it to x264/mp4 and the sound is perfectly fine but again no sound when converting to x265/mkv.


Format : MPEG-4
Format profile : Base Media
Codec ID : isom
File size : 455 MiB
Duration : 45mn 27s
Overall bit rate : 1 399 Kbps
Encoded date : UTC 2016-07-31 02:02:08
Tagged date : UTC 2016-07-31 02:02:08

Video
ID : 1
Format : AVC
Format/Info : Advanced Video Codec
Format profile : High@L3
Format settings, CABAC : Yes
Format settings, ReFrames : 4 frames
Codec ID : avc1
Codec ID/Info : Advanced Video Coding
Duration : 45mn 27s
Bit rate : 1 172 Kbps
Maximum bit rate : 2 586 Kbps
Width : 672 pixels
Height : 448 pixels
Display aspect ratio : 4:3
Original display aspect ratio : 4:3
Frame rate mode : Constant
Frame rate : 29.970 fps
Color space : YUV
Chroma subsampling : 4:2:0
Bit depth : 8 bits
Scan type : Progressive
Bits/(Pixel*Frame) : 0.130
Stream size : 381 MiB (84%)
Writing library : x264 core 148 r2694 3b70645
Encoding settings : cabac=1 / ref=3 / deblock=1:0:0 / analyse=0x3:0x113 / me=hex / subme=7 / psy=1 / psy_rd=1.00:0.00 / mixed_ref=1 / me_range=16 / chroma_me=1 / trellis=1 / 8x8dct=1 / cqm=0 / deadzone=21,11 / fast_pskip=1 / chroma_qp_offset=-2 / threads=6 / lookahead_threads=1 / sliced_threads=0 / nr=0 / decimate=1 / interlaced=0 / bluray_compat=0 / constrained_intra=0 / bframes=3 / b_pyramid=2 / b_adapt=1 / b_bias=0 / direct=1 / weightb=1 / open_gop=0 / weightp=2 / keyint=250 / keyint_min=25 / scenecut=40 / intra_refresh=0 / rc_lookahead=40 / rc=crf / mbtree=1 / crf=22.0 / qcomp=0.60 / qpmin=0 / qpmax=69 / qpstep=4 / ip_ratio=1.40 / aq=1:1.00
Encoded date : UTC 2016-07-31 02:02:08
Tagged date : UTC 2016-07-31 02:02:17

Audio
ID : 2
Format : AC-3
Format/Info : Audio Coding 3
Mode extension : CM (complete main)
Format settings, Endianness : Big
Codec ID : ac-3
Duration : 45mn 27s
Bit rate mode : Constant
Bit rate : 224 Kbps
Maximum bit rate : 237 Kbps
Channel(s) : 2 channels
Channel positions : Front: L R
Sampling rate : 48.0 KHz
Bit depth : 16 bits
Compression mode : Lossy
Stream size : 72.8 MiB (16%)
Encoded date : UTC 2016-07-31 02:02:16
Tagged date : UTC 2016-07-31 02:02:17




Just tried a different file and used just mux again and no sound :( Could this be because the original files audio bitrate is 256 and Staxrip doesn't support 256 bitrare anymore?

http://i67.tinypic.com/20hakpe.png


Format : Matroska
Format version : Version 4 / Version 2
File size : 399 MiB
Duration : 57mn 33s
Overall bit rate : 969 Kbps
Encoded date : UTC 2016-08-01 16:09:54
Writing application : mkvmerge v9.3.1 ('Mask Machine') 64bit
Writing library : libebml v1.3.4 + libmatroska v1.4.5
DURATION : 00:57:33.600000000
NUMBER_OF_FRAMES : 86340
NUMBER_OF_BYTES : 417657825
_STATISTICS_WRITING_APP : mkvmerge v9.3.1 ('Mask Machine') 64bit
_STATISTICS_WRITING_DATE_UTC : 2016-08-01 16:09:54
_STATISTICS_TAGS : BPS DURATION NUMBER_OF_FRAMES NUMBER_OF_BYTES

Video
ID : 1
Format : HEVC
Format/Info : High Efficiency Video Coding
Format profile : Main@L3@Main
Codec ID : V_MPEGH/ISO/HEVC
Duration : 57mn 33s
Bit rate : 950 Kbps
Width : 672 pixels
Height : 544 pixels
Display aspect ratio : 4:3
Original display aspect ratio : 4:3
Frame rate mode : Constant
Frame rate : 25.000 fps
Color space : YUV
Chroma subsampling : 4:2:0
Bit depth : 8 bits
Bits/(Pixel*Frame) : 0.104
Stream size : 391 MiB (98%)
Writing library : x265 2.0+5-98a948623fdc:[Windows][GCC 5.4.0][64 bit] 8bit+10bit+12bit
Encoding settings : wpp / ctu=64 / min-cu-size=8 / max-tu-size=32 / tu-intra-depth=1 / tu-inter-depth=1 / me=1 / subme=2 / merange=57 / no-rect / no-amp / max-merge=2 / temporal-mvp / no-early-skip / rskip / rdpenalty=0 / no-tskip / no-tskip-fast / strong-intra-smoothing / no-lossless / no-cu-lossless / no-constrained-intra / no-fast-intra / open-gop / no-temporal-layers / interlace=0 / keyint=250 / min-keyint=25 / scenecut=40 / rc-lookahead=20 / lookahead-slices=0 / bframes=4 / bframe-bias=0 / b-adapt=2 / ref=3 / limit-refs=3 / no-limit-modes / weightp / no-weightb / aq-mode=1 / qg-size=32 / aq-strength=1.00 / cbqpoffs=0 / crqpoffs=0 / rd=3 / psy-rd=2.00 / rdoq-level=0 / psy-rdoq=0.00 / no-rd-refine / signhide / deblock=0:0 / sao / no-sao-non-deblock / b-pyramid / cutree / no-intra-refresh / rc=crf / crf=22.0 / qcomp=0.60 / qpmin=0 / qpmax=69 / qpstep=4 / ipratio=1.40 / pbratio=1.30
Default : Yes
Forced : No

[JohnLai]

Stax76,
Apparently, staxrip has problem encoding multi chapter / multi video mkv properly. This is prominent in fansub release.

Example, episode 1 video (mkv container) may have op/end linked to another video.
When using staxrip for transcoding such files, the op/end was skipped directly causing out of sync output.

Any suggestion on how to transcode such video?

Menu
00:00:00.000 : en:Prologue / en:Opening / :Ending
00:02:13.008 : en:Episode
00:20:20.029 : :Preview

The problem, the 'Opening' and 'Ending' are linked to two separate independent videos. (Do we call it Multi Chapter MKV or what?)
Meanwhile the audio is continuous.


EDIT:
The one above --->Using DSS2 as source.
Seem like using FFVideoSource in conjunction with avisynth kinda solved the outofsync problem.....by skipping OP and ED completely...well, at least it works.
Question, does FFVideoSource use hardware acceleration for decoding stage?

[MikeTR]

Searching for an 'as much as possible' portable version of Staxrip, I came across the option of making VapourSynth & Python portable by simply merging (http://www.vapoursynth.com/doc/installation.html#windows-portable-instructions) the portable version of VapourSynth into the embeddable Python folder. It seems to me that way both programs could even be distributed in the Staxrip package. Is there a way to make these portable versions work with Staxrip?

EDIT: I see the possibility has already been brought up a few months back, but AFAIK there has been no follow up since (https://github.com/stax76/staxrip/issues/41). Is there anyting I might have missed?

[aegisofrime]

Hi there, just a quick question.

What's the function to get source framerate in Vapoursynth? I googled of course and found some function called Videonode, but I seem to be unable to call it.

I want to be able to do something like this:

super = core.mv.Super(clip, pel=2)
bv = core.mv.Analyse(super, isb=True, overlap=4)
fv = core.mv.Analyse(super, isb=False, overlap=4)
clip = core.mv.FlowFPS(clip, super, bv, fv, num=srcfps * 2,den=1)

The only thing is I don't know how to get srcfps! Any help would be appreciated!

[Khun_Doug]

Is there any method to select a section of the entire source for test encodes? I know there are similar type controls in myffmpeg, AVIDemux, VirtualDub, and Hybrid. The only way I have found to accomplish this so far is to save an uncompressed sample from another tool and then use that sample file until I have my desired settings.

For example, I would like to have an encode start at an offset of three minutes into a source and run for a length of eight minutes. I don't see anyway to do this.

[JohnLai]

Weird, it had been more than a week (counting 2 weeks) since I reported Nvidia about HEVC encoding only makes use of 1 reference frame bug to Nvidia and didn't get any reply back. (Yeah, both Maxwell V2 and Pascal affected)

Normally they replied within 2 or 3 working days and fixed stuff quickly enough once they confirmed it. (Example, the HEVC CU-Min & Max size bug was fixed before sdk 7.0 came out)

Oh, I got Pascal 10bit HEVC sample from rigaya and it seems SAO support is enabled by default (no on/off option). Seem like nvidia is using a form of 'low complexity sample adaptive offset encoding'. (google it if one wanna read the patent)
Everything else is exactly the same as Maxwell V2 (including bugs).

Speculation section:
I got a bad feeling on both Maxwellv2 and Pascal HEVC's amount of reference frame support and I hope I am wrong in this.
As we know, nvidia HEVC encoding does not support B-Frame.
For software encoder x265, P-Frame can refer multiple preceding P-Frame (and I Frame) as reference frames.
Nvenc cannot do so.
From FFMPEG-with-NVIDIA-Acceleration-on-Ubuntu_UG_v01.pdf, there is a section about B-Frame which nvidia defines "Reference frames are the frames that can be referred to by B frames."
So...based on nvidia definition of reference frame.....and there is no b-frame support for HEVC encoding.....does this mean it is a hardware limitation? :(

[luigizaninoni]

I have these two connected issues:

I am using staxrip 1.3.8.7 test build

The script is the following:

LoadPlugin("C:\Users\luigi\Desktop\video grafica\stax64\Apps\Plugins\avs\L-SMASH-Works\LSMASHSource.dll")
LWLibavVideoSource("C:\Users\luigi\Desktop\ConvertitiInMkv\00000.mkv", format = "YUV420P8")
MCTemporalDenoise()

First issue:
The encode works fine with x264
With x265 it crashes: avs2pipemod64.exe has stopped working

Second issue:
The script can't be previewed in staxrip: Access violation
However the script can be previewed easily with AVSPmod

So, why does it encode with x264 and doesn't with x265 ?
Why does preview work in Avspmod and not in staxrip ?

[burfadel]

I have these two connected issues:

I am using staxrip 1.3.8.7 test build

The script is the following:

LoadPlugin("C:\Users\luigi\Desktop\video grafica\stax64\Apps\Plugins\avs\L-SMASH-Works\LSMASHSource.dll")
LWLibavVideoSource("C:\Users\luigi\Desktop\ConvertitiInMkv\00000.mkv", format = "YUV420P8")
MCTemporalDenoise()

First issue:
The encode works fine with x264
With x265 it crashes: avs2pipemod64.exe has stopped working


Try the latest version of avs2pipemod (now 1.03). Apparently some possible memory leaks were rectified.
https://github.com/chikuzen/avs2pipemod/releases/download/1.0.3/avs2pipemod-1.0.3.zip

[luigizaninoni]

Try the latest version of avs2pipemod (now 1.03). Apparently some possible memory leaks were rectified.
https://github.com/chikuzen/avs2pipemod/releases/download/1.0.3/avs2pipemod-1.0.3.zip

Thanks for the heads-up, but encode with x265 still fails. However the error message has changed, it now gives "output various data from avs to stdout has stopped working"

[luigizaninoni]

The encode works fine with x264
With x265 it crashes: avs2pipemod64.exe has stopped working


Anyway, the x265 encode works fine in staxrip 1.3.7.0, so something must have changed between 1.3.7.0 and 1.3.8.7

[Magik Mark]

Why is it running avsmeter I get around 60fps. When the actual encoding begins it's only 30fps? Is there a variable that it not being considered during the test?

[sneaker_ger]

avsmeter only tests the speed of the script (decoding, filtering). When you encode you run script + x265 encoding. Of course it takes longer.

[Khun_Doug]

Are there any plans to add the TWEAK function/filter to the VapourSynth support?

[ShogoXT]

Hey have you guys seen that AMD released AMF encoding? Not sure if it would be more useful than the already existing one though as h265 is still not available.
https://github.com/GPUOpen-LibrariesAndSDKs/AMF
http://gpuopen.com/gaming-product/advanced-media-framework/

EDIT: I would take a close look at the issues page on the github. There seems to be question about Bframe support and useful links to "experimental" HEVC transcoders.

[JohnLai]

Hey have you guys seen that AMD released AMF encoding? Not sure if it would be more useful than the already existing one though as h265 is still not available.
https://github.com/GPUOpen-LibrariesAndSDKs/AMF
http://gpuopen.com/gaming-product/advanced-media-framework/

EDIT: I would take a close look at the issues page on the github. There seems to be question about Bframe support and useful links to "experimental" HEVC transcoders.

Been checking about AMD HEVC few days ago. Nope, no B-Frame in any of samples provided.
Now, the interesting part.....AMD HEVC sample shows support for 64x64 CU Size.......and minimum 4x4 for intra and 8x8 for inter.

[ismail0100]

can i give setting another video automaticly? (poor english)
So I want to copy all settings (10bit 22crf etc.).