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 > MPEG-4 Encoder GUIs
Register FAQ Calendar Today's Posts Search

Reply
 
Thread Tools Search this Thread Display Modes
Old 16th September 2017, 15:13   #2521  |  Link
stax76
Registered User
 
stax76's Avatar
 
Join Date: Jun 2002
Location: On thin ice
Posts: 6,837
You can report it here:

http://rationalqm.us/board/index.php
stax76 is offline   Reply With Quote
Old 16th September 2017, 19:42   #2522  |  Link
mparade
Registered User
 
Join Date: Nov 2013
Posts: 577
Quote:
Originally Posted by stax76 View Post
You can report it here:

http://rationalqm.us/board/index.php
I will be reporting the problem to Donald Graft urgently because I am quite sure it is a CUVID bug.
mparade is offline   Reply With Quote
Old 17th September 2017, 16:06   #2523  |  Link
mparade
Registered User
 
Join Date: Nov 2013
Posts: 577
Quote:
First I wanted to support the mastering display parameters based on recent posts in the x265 thread but it was a little bit tricky. If you post your source media info and your resulting command line it will help me to code it.
If media info tells:

Quote:
Color range : Limited
Color primaries : BT.2020
Transfer characteristics : PQ
Matrix coefficients : BT.2020 non-constant
Mastering display color primaries : R: x=0.680000 y=0.320000, G: x=0.265000 y=0.690000, B: x=0.150000 y=0.060000, White point: x=0.312700 y=0.329000
Mastering display luminance : min: 0.0050 cd/m2, max: 4000.0000 cd/m2
Maximum Content Light Level : 4000 cd/m2
Maximum Frame-Average Light Level : 290 cd/m2
a sample command line would be according to x265 specs (16bit input is assumed via DGDecNV and if I understand the specs correctly):

Quote:
--crf 24 --preset slow --tune grain --profile main10 --output-depth 10 --master-display "G(13250,34500)B(7500,3000)R(34000,16000)WP(15635,16450)L(40000000,50)" --colorprim bt2020 --colormatrix bt2020nc --transfer smpte-st-2084 --range limited --max-cll "4000,290" --input-depth 16

Last edited by mparade; 17th September 2017 at 16:09.
mparade is offline   Reply With Quote
Old 18th September 2017, 01:06   #2524  |  Link
stax76
Registered User
 
stax76's Avatar
 
Join Date: Jun 2002
Location: On thin ice
Posts: 6,837
Movie scene where a jukebox is spinning a stax record...

https://www.youtube.com/watch?v=97qA-Wr1SBU


@mparade

Thanks, it will definitively be supported in the next build (or one after).
stax76 is offline   Reply With Quote
Old 18th September 2017, 04:09   #2525  |  Link
dvd_maniac
Registered User
 
Join Date: Feb 2002
Location: northeast/usa
Posts: 830
Stax,
I upgraded to 1.6.0 and my batch script is broken again.
I opens a staxrip.exe but does not encode or create a temp folder. When I try to run the cli script manually it opens the select Avisynth source filter dialog box. How can I bypass this and set it to automatic for every run?
dvd_maniac is offline   Reply With Quote
Old 18th September 2017, 06:37   #2526  |  Link
stax76
Registered User
 
stax76's Avatar
 
Join Date: Jun 2002
Location: On thin ice
Posts: 6,837
Quote:
When I try to run the cli script manually it opens the select Avisynth source filter dialog box. How can I bypass this and set it to automatic for every run?
This dialog is only shown if the source filter is set to 'Manual', you can set it to 'Automatic'.

There is a project option on the misc tab that will hide any dialog, it's probably better to enable this only via CLI with following CLI switch: -SetHideDialogsOption:true
stax76 is offline   Reply With Quote
Old 18th September 2017, 07:27   #2527  |  Link
mparade
Registered User
 
Join Date: Nov 2013
Posts: 577
Quote:
Thanks, it will definitively be supported in the next build (or one after).
Thank you. I am waiting for the CUDA bug to repair because I prefer GPU assisted decoding:

https://devtalk.nvidia.com/default/t...ing-is-broken/
mparade is offline   Reply With Quote
Old 18th September 2017, 12:39   #2528  |  Link
stax76
Registered User
 
stax76's Avatar
 
Join Date: Jun 2002
Location: On thin ice
Posts: 6,837
Quote:
Originally Posted by mparade View Post
Thank you. I am waiting for the CUDA bug to repair because I prefer GPU assisted decoding:

https://devtalk.nvidia.com/default/t...ing-is-broken/
The only other source filter that support hardware decoding is DSS2/DirectShowSource but it's probably 8-bit only.
stax76 is offline   Reply With Quote
Old 18th September 2017, 13:29   #2529  |  Link
ABDO
Registered User
 
Join Date: Dec 2016
Posts: 65
Hi frank, thank you for your hard work in staxrip, i have a question please, can i import image sequence to staxrip.
ABDO is offline   Reply With Quote
Old 18th September 2017, 13:34   #2530  |  Link
stax76
Registered User
 
stax76's Avatar
 
Join Date: Jun 2002
Location: On thin ice
Posts: 6,837
Hi, what kind of image and filename pattern do you have? One filter I found is this:

http://avisynth.nl/index.php/ImageSource
stax76 is offline   Reply With Quote
Old 18th September 2017, 13:48   #2531  |  Link
ABDO
Registered User
 
Join Date: Dec 2016
Posts: 65
thanks for your quick reply, my image is png or Bmp, i use adobe premiere pro to export it, and want to use staxrip to encode it because it much better in this case,the filename pattern is depend on the number of frams exported , if less than 3 digits from 1:100 ( 01-02-03-99), if larg than 3 digts then (001-002-003) and so on.
ABDO is offline   Reply With Quote
Old 18th September 2017, 14:29   #2532  |  Link
stax76
Registered User
 
stax76's Avatar
 
Join Date: Jun 2002
Location: On thin ice
Posts: 6,837
You can try to customize staxrip or write a avs/vpy script and open the script file as source file. I'll try to add direct support for image sequences. Would exporting it in a video format with adobe premiere not work?
stax76 is offline   Reply With Quote
Old 18th September 2017, 14:48   #2533  |  Link
ABDO
Registered User
 
Join Date: Dec 2016
Posts: 65
Would exporting it in a video format with adobe premiere not work? i can use adobe premiere to exporting it in a video, but the Problem is the (hevc encoder) in premiere is Useless, and give very bad result compara it (x265) in staxrip, would be great if you can get some time and effort to add this Feature dirctly to staxrip,Unfortunately i do not understand or use scripts
ABDO is offline   Reply With Quote
Old 18th September 2017, 15:10   #2534  |  Link
StephanD
Registered User
 
Join Date: Sep 2017
Location: Germany
Posts: 2
Heyho,

Dunno if this was already reported, couldn't find anything in the search, but there is a mistake in the automatically created command line für AV1.

If you use the threads-flag, its automatically creating "t=x". This is not correct. There are two flags for it. If you use "--threads", you can use the "=". if you use the short "-t" it should be "-t x" without "=".

As this codec is still in progress, I wouldn't use the shortcodes, because it is not clear, if they remain in later versions.

For example you get a failure when using the "--best"-flag in your programm. In the latest repository of AV1 this flag is gone.

greetz
StephanD is offline   Reply With Quote
Old 18th September 2017, 15:57   #2535  |  Link
mcjordan
Registered User
 
Join Date: Nov 2010
Posts: 124
Updates

mClean v1.7c
* Slight adjustments and slightly better speed

NVEnc v3.20
Summary: Addition of automatic selection function of edge emphasis filter/device, correction of defects, etc.

Common
· If you do not specify otherwise, make device selection automatically when you execute.
In a multi-GPU environment, automatically determine the GPU to be used, especially when there is no designation of the device to be used.

The judgment standard is:
· Whether the specified encoding is possible (check codec, 8bit/10bit, YUV420/YUV444 etc)
· If --avhw is specified, it is possible to decode HW of the input file
· Low use rate of Video Engine
· Low GPU usage rate
· New generation of GPU
· People with a large number of GPU cores

The usage rate of the Video Engine and GPU usage rate are acquired by executing "nvidia -smi.exe -q" because the x64 version can not use the NVML library and the x86 version can not use NVML.
Nvidia-smi.exe is normally installed in "C:\Program Files\NVIDIA Corporation\NVSMI\nvidia-smi.exe" together with the driver.

· Added Unsharp filter. (-vpp-unsharp)
Unsharp filter for ordinary contour and detail emphasis (I think that's it). At first I tried using it because it is in the NPP library ... I tried it a lot but I thought that usage was bad after all, but it did not work well, so I decided to implement it myself with CUDA as there was no choice.

Parameters
-radius = <int> (default = 3, 1-9)
Scope of contour / detail detection. By setting it to a larger value, it will emphasize in response to a wider range of detail.

-weight = <float> (default = 0.5, 0-10)
Strength of contour and detail emphasis. By setting it to a larger value, it becomes strongly emphasized.

-threshold = <float> (default = 10.0, 0-255)
Threshold for contour and detail detection. Outline emphasis is performed on pixels having a difference equal to or larger than the threshold value.

Example: Somewhat stronger
-vpp-unsharp weight = 1.0

· Added edge level adjustment filter. (-vpp-edgelevel)
Added by adding CUDA as an edge level adjustment filter after Unsharp. The default value of the parameter is slightly weaker than the Aviutl version.

Parameters
-strength = <float> (default = 5.0, - 31-31)
Strength of contour emphasis. By setting it to a larger value, it will emphasize in response to a wider range of detail.

-threshold = <float> (default = 20.0, 0-255)
Noise threshold to avoid outline emphasis. As the larger the value, the larger the luminance change is treated as noise.

-black = <float> (default = 0.0, 0-31)
For the black part of the outline, shoot black more and emphasize the outline.

-white = <float> (default = 0.0, 0-31)
For white portions of the outline, shoot more whitely and emphasize the outline.

Example: Somewhat stronger (Aviutl version default)
-vpp-edgelevel strength=10.0, threshold=16.0, black=0, white=0

Example: Strengthening the black part of the outline
- vpp-edgelevel strength=5.0, threshold=24.0, black=6.0

NVEncC

· Fixed crop being doubled when using crop and resize when using avhw reader.

· Adjust the argument of --transfer to that used in x264 etc.
smpte-st-2048 -> smpte 2048
smpte-st-428 → smpte 428

Last edited by mcjordan; 18th September 2017 at 16:01.
mcjordan is offline   Reply With Quote
Old 18th September 2017, 16:41   #2536  |  Link
stax76
Registered User
 
stax76's Avatar
 
Join Date: Jun 2002
Location: On thin ice
Posts: 6,837
@ABDO

I tried the filter, everything needs to be defined manually so it's too much work to integrate the filter in a way that things work as simple as opening a video file.

D:\Temp\aaa\test_000001.jpg
D:\Temp\aaa\test_003040.jpg

ImageSource("D:\Temp\aaa\test_%06d.jpg", 1, 3040, 25)

Maybe there is a easier way with ffmpeg or vapoursynth.

From a tool like premiere I would expect that it can export to a suitable video format.

@StephanD
@mcjordan

Thanks for reporting, I'll fix/update nvenc/aomenc support.
stax76 is offline   Reply With Quote
Old 18th September 2017, 18:05   #2537  |  Link
ABDO
Registered User
 
Join Date: Dec 2016
Posts: 65
ok, i will export video from premiere with ( AVC -H264 ) and it did Excellent job.thank you frank, always willing to help
ABDO is offline   Reply With Quote
Old 18th September 2017, 19:30   #2538  |  Link
mcjordan
Registered User
 
Join Date: Nov 2010
Posts: 124
Hmm... About new NVEnc 3.2.0
With this addition to my custom NVEnc profile aka
--vpp-unsharp radius=2,weight=0.65,threshold=20 --vpp-edgelevel strength=6.0,threshold=16.0,black=0,white=0
I've get a massive color shifting towards the red spectrum area.
Тhis is not observed when using any other --vpp filter.
Maybe I'm losing something somewhere...
---
Part of job processing info:
Input Info Avisynth 2.60(yv12)->p010 [AVX], 1920x816, 24000/1001 fps
Vpp Filters copyHtoD
cspconv(p010 -> yv12(16bit))
unsharp: radius 2, weight 0.6, threshold 20.0
edgelevel: strength 6.0, threshold 16.0, black 0.0, white 0.0
cspconv(yv12(16bit) -> p010)
---
Source media info:
ID : 1
Format : AVC
Format/Info : Advanced Video Codec
Format profile : Main@L4
Format settings : 2 Ref Frames
Format settings, CABAC : No
Format settings, RefFrames : 2 frames
Codec ID : V_MPEG4/ISO/AVC
Duration : 2 min 31 s
Width : 1 920 pixels
Height : 816 pixels
Display aspect ratio : 2.35:1
Frame rate mode : Constant
Frame rate : 23.976 (24000/1001) FPS
Color space : YUV
Chroma subsampling : 4:2:0
Bit depth : 8 bits
Scan type : Progressive
Language : English
Default : Yes
Forced : No
Color range : Limited
Color primaries : BT.709
Transfer characteristics : BT.709
Matrix coefficients : BT.709

Help me - where am I wrong?
mcjordan is offline   Reply With Quote
Old 19th September 2017, 08:15   #2539  |  Link
Schattenspieler
Registered User
 
Join Date: May 2012
Posts: 42
Quote:
Originally Posted by ABDO View Post
…, but the Problem is the (hevc encoder) in premiere is Useless, and give very bad result compara it (x265) in staxrip, …
You could also try to export using a lossless codec or maybe a codec with lossless settings. There are many threads on the topic over at Adobes site, like this one:

https://forums.adobe.com/message/7686319#7686319
Schattenspieler is offline   Reply With Quote
Old 19th September 2017, 16:10   #2540  |  Link
mcjordan
Registered User
 
Join Date: Nov 2010
Posts: 124
NVEnc v3.2.1
* fix issue of vpp-unsharp, which made the color of output video becoming unstable, going either red or blue.

Problem of color shifting /my previous post/ is resolved.
mcjordan is offline   Reply With Quote
Reply


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 21:09.


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