patman86 / x265-mod-by-patman Goto Github PK

Dear @Patman86, when you find some time, please remember to release SandyBridge x264/x265.

Thx! :)

Would you wish us a Merry Xmas with a new compile? ;)

Thanks in advance!

StaxRip v.2.29.0

Profile 422 does not work in 265 codec - completely

and when selecting 422 in the 264 codec, you must manually specify 422 for input and output - it does not work in automatic mode

using visual studio 2019
cmake shows Looking for strtok_r - not found and then stop
pretty weird that cmake works no problem when compile msg7086's x265 which shows Looking for strtok_r - not found as well

Hi,
I was wondering if a copy feature can be added that can copy source metadata (similar to nvenc hw encoder for example). Some sources that are HDR10+ (SMPTE ST 2094 App 4, Version 1, HDR10+ Profile B compatible) are only HDR after re-encoding since no support for SMPTE ST 2094 is available (SMPTE ST 2086, HDR10 compatible). Would be nice to have the HDR10+ if possible.

I know there are tools to extract HDR10+ metadata and probably can be patched/merged but it would be nice if it was done in one place if possible. Thank you for considering the feature.

Would it be possible to get a csv (per frame) or the encoding log (percentage of all frames) entry for the selected AQ modes when auto AQ is enabled?

I have done some tests with the last AVX version and I have found that in most tests, it has different VMAF metrics than plain version:

Sintetic ones such as PSNR and SSIM are almost identical.

I use FFMetrics, it you want to replicate them.

Please update the CLI help text for --tune to include the "animation" option, i.e.,

current:

-t/--tune <string>               Tune the settings for a particular type of source or situation:
                                 psnr, ssim, grain, zerolatency, fastdecode

proposed:

-t/--tune <string>               Tune the settings for a particular type of source or situation:
                                 psnr, ssim, grain, zerolatency, fastdecode, animation

Thanks much.

CRF bitrates are much lower at the same settings than the previous +153 build. First pass bitrates can be ~20% over the specified bitrate as well, although second pass final rate is accurate.

Would you please add the --auto-aq patch?

I'd like to ask for this feature in StaxRip but they couldn't having a proper build :)

I am really confused about the +patch +version inside x265 builds.

x265-3.5+104+15 by Patman — yesterday

x265-3.5+131-aMod by DJATOM — 3 weeks ago

x265_r3_5_0_103 by jpsdr — May 14

How can we understand what is "the newest" regarding main branch, additional features, etc?

Could you please release Comet Lake optimization of your mod?

Hi, I noticed this weird indentation

https://github.com/Patman86/x265-Mod-by-Patman/blame/ba4e7a2cb3266d9cd59ab70ad5832d520a882254/source/encoder/slicetype.cpp#L614

Is this just an indentation errror, or a bug?

Hello, would it be possible for you to make an optimized version for ryzen zen 3 and zen 2 like @DJATOM did!
Thanks

Hello,
i use StaxRip with your latest version on x265 encoder and there is something on my 2017 LG OLED TV:

When i do encode existing DV profile 5/8 to reduce video size my TV can trigger and play new smaller file but i can't do FF or RW.
When i do FF or RW my TV start playing from begining.

this has been tested on other DV TV capable by other user.

When i do encode with NVENC there is no such behavior.
I have tested other x265 releases:

• x265 binaries for Windows: 3.5+68-40e37bc
• x265-x64-v3.5+20-aMod-gcc10.2.1+opt.7z (https://github.com/DJATOM/x265-aMod/releases/download/3.5%2B20/x265-x64-v3.5+20-aMod-gcc10.2.1+opt.7z)

hello,

i would like to ask what is the best version (compilation) of your last release to Ryzen 7-7840HS ?

Hello @Patman86 ,

may I suggest using the correct binary prefix as shown here:
https://en.wikipedia.org/wiki/Binary_prefix#Adoption_by_IEC,_NIST_and_ISO
https://de.wikipedia.org/wiki/Bin%C3%A4rpr%C3%A4fix#IEC-Pr%C3%A4fixe_zur_Basis_2

As I saw Ki and Mi would be the appropriate prefixes instead of Kand M because the bytes are divided by 1024:

On the other side using KB``and MB` would be better, when values would be divided by 1000. 😁

I'm glad you created these repositories. 👍 😄

Hey @Patman86 ,

are you going to update your code to v3.5 FINAL in the near future? 😃

I am doing some real scenario benchmarks to choose what flavor of your builds to use on my good old Sandy Bridge.

I have the following results:

3,62 fps — x265-msvc1938
3,65 fps — x265-avx-msvc1938
3,69 fps — x265-sandybridge-gcc13.2.0

Unfortunately x265-gcc13.2.0 gives me the following error, such as if you forced some instruction set. Shouldn't be the most generic build, such as the x265-msvc1938 one?

Video encoding returned exit code: -1073741795 (0xC000001D)

It's unclear what this exit code means, in case it's
a Windows system error then it possibly means:

# for decimal -1073741795 / hex 0xc000001d
  STATUS_ILLEGAL_INSTRUCTION                                     ntstatus.h   
# {EXCEPTION}
# Illegal Instruction
# An attempt was made to execute an illegal instruction.
# as an HRESULT: Severity: FAILURE (1), FACILITY_NULL (0x0), Code 0x1d
# for decimal 29 / hex 0x1d
  ERROR_WRITE_FAULT                                              winerror.h   
# The system cannot write to the specified device.

P.S: As a curiosity, would be possible to have a msvc sandybridge build to benchmark against gcc sandybridge one?

how to enable this mode? --- StaxRip software

QTGMC and interframe have become more efficient when processing UHD

Hi @Patman86

I have already posted something similar @DJATOM 's repo, months ago, and I wish to repeat my request, adding some more info:

1. average FPS has a major flaw: if you suspend processing (powerful feature in Staxrip) then resume 1 hour later, the average FPS will have no meaning at all, and the ETA will be completely unrealistic! Example: I process a file, I suspend processing 5 minutes before it ends, and I resume processing 2 hours later. The ETA will show something like 1 hour or more, when it is expected to be 5 minutes.
2. average FPS vs Instant FPS can both be defended as good indicators.

My request is based on fixing the issue pointed out in (1).

The "Instant" FPS would be an average FPS calculated on a window of frames, or a number of buffers.
And the ETA will be based on that "Instant" FPS.

So in case the processing is suspended and resumed, the ETA will quickly give a more realistic reading.
I hope you may provide a solution !
Thanks!!

Sometimes my encoding gets stuck at 100%. I don't finish even after a few hours. It doesn't consume CPU power in that period.
So I closed the encoding and checked the log file. The log file doesn't show any errors. It's just like a properly done encoded log file.
The problem occurs when I write a command to start another encode after finishing one. But since the first one doesn't get finished (Stuck at 100%), the second one doesn't start. :(

https://bitbucket.org/multicoreware/x265_git/commits/aa7f602f7592eddb9d87749be7466da005b556ee

I have the following error when adding parameters to --auto-aq, i.e. --auto-aq 10.

D:\Eseguibili\Media\StaxRip\Apps\Encoders\x265\x265.exe --crf 20 --preset slow --output-depth 10 --keyint 24 --colorprim bt709 --colormatrix bt709 --transfer bt709 --range limited --qpfile "M:\In\Vinland saga ~BDJP\02 ~847p\11 aqauto 10.qp" --auto-aq 10 --output "M:\In\Vinland saga ~BDJP\02 ~847p\11 aqauto 10_temp\11 aqauto 10_out.hevc" "M:\In\Vinland saga ~BDJP\02 ~847p\11 aqauto 10_temp\11 aqauto 10.avs"

x265 [FLAW]: M:\In\Vinland saga ~BDJP\02 ~847p\11 aqauto 10.qp qpfile not found or error in opening qp file
x265 [WARN]: extra unused command arguments given <M:\In\Vinland saga ~BDJP\02 ~847p\11 aqauto 10_temp\11 aqauto 10.avs>