StaxRip 2.0 Support Thread

[aegisofrime]

Hello,

I'm getting the following error with certain content, especially wmv files:

vapoursynth.Error: Analyse: failed to retrieve first frame from super clip. Error message: Resize error: Resize error 3074: no path between colorspaces (2/2/2 => 1/2/2). May need to specify additional colorspace parameters.

I noticed that this error is caused by the following line, which appears to be auto-added by StaxRip:

clip = clip.resize.Bicubic(matrix_s = '709', format = vs.YUV420P8)

The strange thing is, the script as a whole works perfectly fine without the above line, so why is it being auto-added and how do I stop StaxRip from auto-adding it?

[NikosD]

Originally Posted by stax76

It's a quality mode and I missed setting the main window to quality mode view, 2.0.4.6 Beta is available.

Hello Frank.

It seems that NVEncC GUI is missing two useful switches: (or I just couldn't find them)

--aq-strength <int> set aq strength (weak 1 - 15 strong) default: 0 = auto
which is used along with --aq

and

--bref-mode <string> set B frame reference mode
- disabled (default)
- each
- middle

Also, I suggest for the GUI to rename these switches at the Rate Control tab:

Adaptive Quantization to Adaptive Quantization (Spatial)
and
AQ Temporal to Adaptive Quantization (Temporal)
to be more descriptive and accurate.

Thanks!

[stax76]

@aegisofrime

When the source is loaded staxrip checks with mediainfo if chroma subsampling is 4:2:0, if not it adds a function to convert it to 4:2:0, this function is faulty for a reason I don't understand. I'm not sure if it's a great idea to automatically insert this function, feedback is welcome, it would not be difficult to add an option so it can be user disabled. Because it's only inserted after the source is loaded you can remove it in the script editor afterwards.

@NikosD

--aq-strength is only visible for h264, does it work with h265?

--bref-mode was marked as ignore for unknown reasons (staxrip analyzes --fullhelp to discover new/changed/removed switches), next build will support it.

Adaptive Quantization to Adaptive Quantization (Spatial)

AQ Temporal to Adaptive Quantization (Temporal)

Available in the next build.

[NikosD]

Originally Posted by stax76

@NikosD

--aq-strength is only visible for h264, does it work with h265?

--bref-mode was marked as ignore for unknown reasons (staxrip analyzes --fullhelp to discover new/changed/removed switches), next build will support it.

Available in the next build.

Thanks.
Both AQ modes work with Turing encoders using both H.264 and H.265.

--aq is spatial and is used along with --aq-strength for both H.264/H.265 for all (almost) nVidia cards

--aq-temporal is temporal obviously and it works for all (almost) nVidia cards using H.264.
But Turing encoders can use it for H.265 too.

A more advanced way to figure out what it works for a specific card in order to offer it to the GUI, could be to run --check-features in real-time in background and read the appropriate fields e.g AQ Temporal in order to decide if you can provide it or not to the GUI/ user.

[Atlantis]

Question. What CRF do you use for your 4K HDR encodes?

With a little detail and noise in the picture, the encoding size could become huge with even CRF 23.

What are your limits for video bitrate? I have an arbitrary limit that my 4K encodes should not go higher than 15 Mbit/s for video.

[Atlantis]

Feature Request: in the demux window, you get buttons to select None for Audio and Subtitle but you do not have buttons for the Attachments. It's better to have the same buttons that you have for Audio and Subtitle also for attachments.

[ZeroCool22]

What does --bref-mode?

[jlw_4049]

Originally Posted by Atlantis

Question. What CRF do you use for your 4K HDR encodes?



With a little detail and noise in the picture, the encoding size could become huge with even CRF 23.



What are your limits for video bitrate? I have an arbitrary limit that my 4K encodes should not go higher than 15 Mbit/s for video.

Varies widely from video to video. I do multiple tests to get the CRF I want for my final encode. Takes me about 10 minutes per video to test for b frames and then test for the proper CRF setting.

Also I shoot for 16 MB's.

Sent from my SM-G965U1 using Tapatalk

[Patman]

Originally Posted by ZeroCool22

What does --bref-mode?

Look here

[JKyle]

2.0.4.8 Beta:

Since my MPC-HC is installed with K-Lite Codec Pack and its exe name is weirdly renamed as mpc-hc64_nvo.exe by K-Lite, I forced StaxRip to find this exe file on the Apps > Manage > MPC-HC menu.
No problem up to here.

But when I try to launch MPC-HC, StaxRip reports the following error message and fails to do it.



I guess it's because StaxRip tries to launch the app by calling the regular mpc-hc64.exe, not the filename altered by K-Lite.

Can you fix this issue please?

[ZeroCool22]

Originally Posted by Patman

Look here

Thank you, i'm really new to Staxrip, this will help a lot, thx.

[ZeroCool22]

Originally Posted by Salem874

I've found StaxRip after trying to find a way to rip HDR content to a smaller size, but KEEPING the HDR metadata.

I currently use Handbrake due to the excellent quality of it's exports, plus it's super easy to use (but still flexible), but as we know it's not yet fully HDR/10-bit compatible.

However, i'm extremely confused as to how to re-encode/rip HDR content using StaxRip. Are there any guides/tutorials out there?

But Handbrake doesn't let you encode with H265/NVEnc at 10 Bits, only with x265 and with this last one, you can't make use of the power of the GPU.

[ZeroCool22]

It's possible for StaxRip to encode with H265 at 12 Bits and if yes how do it, what parameter should i use?

Thx in advance?

[NikosD]

Originally Posted by stax76

Available in the next build.

Just downloaded your latest 2.0.4.8 version.

A few comments:

1) AQ Strength should be available only when someone selects Adaptive Quantization (Spatial).
From what I have read, it can't be used as a standalone option or using Adaptive Quantization (Temporal)

2) A small bug regarding "Constant Quality Mode"

Constant Quality Mode can be used only with VBRHQ like this:
"--vbrhq 0 --vbr-quality <float>"

In StaxRip when you select that option, you added the --vbr-quality 0 parameter.
You should add --vbrhq 0 instead and should be available with VBRHQ only.

3) The following HW resize options of --vpp-resize

nn nearest neighbor
npp_linear linear interpolation by NPP library
cubic 4x4 cubic interpolation
super So called "super sampling" by NPP library
lanczos Lanczos interpolation

need the nppi64_10.dll which uses the NPP library and supports x64 version only.

To use those algorithms, you need to download nppi64_10.dll separately and place it in the same folder as NVEncC64.exe.

Thanks!

[ZeroCool22]

Originally Posted by NikosD

Just downloaded your latest 2.0.4.8 version.

A few comments:

1) AQ Strength should be available only when someone selects Adaptive Quantization (Spatial).
From what I have read, it can't be used as a standalone option or using Adaptive Quantization (Temporal)

2) A small bug regarding "Constant Quality Mode"

Constant Quality Mode can be used only with VBRHQ like this:
"--vbrhq 0 --vbr-quality <float>"

In StaxRip when you select that option, you added the --vbr-quality 0 parameter.
You should add --vbrhq 0 instead and should be available with VBRHQ only.

3) The following HW resize options of --vpp-resize

nn nearest neighbor
npp_linear linear interpolation by NPP library
cubic 4x4 cubic interpolation
super So called "super sampling" by NPP library
lanczos Lanczos interpolation

need the nppi64_10.dll which uses the NPP library and supports x64 version only.

To use those algorithms, you need to download nppi64_10.dll separately and place it in the same folder as NVEncC64.exe.

Thanks!

--vbrhq 0 High value will give you better quality or it's always better to leave it at 0?

And also

adaptive quantization (spatial)

adaptive quantization (temporal)

Is always better to have it checked on?

Constant Quality Mode just block the other two, so it have no sense to active one of the other two if we use CQM?

Sorry for make a lot of question but i really want to know what i'm doing and if i'm doing it the correct way.

[Atlantis]

Originally Posted by jlw_4049

Varies widely from video to video. I do multiple tests to get the CRF I want for my final encode. Takes me about 10 minutes per video to test for b frames and then test for the proper CRF setting.

Also I shoot for 16 MB's.

How do you test to see which CRF to use?

And you meant 16Mbit/s not MB, right?

[stax76]

Since my MPC-HC is installed with K-Lite Codec Pack and its exe name is weirdly renamed as mpc-hc64_nvo.exe by K-Lite, I forced StaxRip to find this exe file on the Apps > Manage > MPC-HC menu.
No problem up to here.

But when I try to launch MPC-HC, StaxRip reports the following error message and fails to do it.

Will be fixed in the next build.

Constant Quality Mode can be used only with VBRHQ like this:
"--vbrhq 0 --vbr-quality <float>"

In StaxRip when you select that option, you added the --vbr-quality 0 parameter.
You should add --vbrhq 0 instead and should be available with VBRHQ only.

Please re-check, if you check 'Constant Quality Mode' then you should see --vbrhq 0

3) The following HW resize options of --vpp-resize

nn nearest neighbor
npp_linear linear interpolation by NPP library
cubic 4x4 cubic interpolation
super So called "super sampling" by NPP library
lanczos Lanczos interpolation

need the nppi64_10.dll which uses the NPP library and supports x64 version only.

To use those algorithms, you need to download nppi64_10.dll separately and place it in the same folder as NVEncC64.exe.

How large is the file and where can it be downloaded?

[Taurus]

How large is the file and where can it be downloaded?

Look here at rigayas google drive website:https://drive.google.com/drive/folde...1ZUT1FjTnF3Q0E
Or for the somewhat older npp64_80.dll: https://drive.google.com/drive/folde...0lnQ2tRTF9Zcnc

[JKyle]

Originally Posted by stax76

Please re-check, if you check 'Constant Quality Mode' then you should see --vbrhq 0

Stax is right on this.

[stax76]

Look here at rigayas google drive website:https://drive.google.com/drive/folde...1ZUT1FjTnF3Q0E
Or for the somewhat older npp64_80.dll: https://drive.google.com/drive/folde...0lnQ2tRTF9Zcnc

Thanks, I think it's too large to include by default.