x265 HEVC Encoder

[RanmaCanada]

Originally Posted by Mr.Rippley

Please, help me!

This particular topic is for discussing releases of x265. If you want actual help in learning how to encode with it you could either read the forum itself as there is a tonne of information in it, or create a topic in the forum requesting help while posting as much information as possible.

[Barough]

x265 v3.5+18-747a079f7 (32 & 64-bit 8/10/12bit Multilib Windows Binaries) (GCC 10.3.0)

https://bitbucket.org/multicoreware/x265_git/commits/branch/master

[tuanden0]

Everyone can access to this site http://msystem.waw.pl/x265/?
I got error 404

[RanmaCanada]

Originally Posted by tuanden0

Everyone can access to this site http://msystem.waw.pl/x265/?
I got error 404

works here.

[charliebaby]

x265-3.5+19-8003e43_gcc112-AVX2
https://www.mediafire.com/file/zms2i...2-AVX2.7z/file

[Dogway]

What bitdepth and packing format is used for UHDs? I ask because I plan to do encodes in ICtCp or maybe its variant IPTPQc2. Are there any consumer media in these color models? Do they use 10 or 12-bit, I420, NV12 or something else?

[quietvoid]

Originally Posted by Dogway

What bitdepth and packing format is used for UHDs? I ask because I plan to do encodes in ICtCp or maybe its variant IPTPQc2. Are there any consumer media in these color models? Do they use 10 or 12-bit, I420, NV12 or something else?

UHDs? As in UHD Blu-ray discs?

The only media in IPTPQc2 is premium content on streaming services. It requires access to the Dolby Engine for encoding.
It is a 10 bit video file in the end along with Dolby Vision metadata, and requires playback on a Dolby Vision capable device and display.

There's no consumer media over 10 bit, and it's probably all YUV420.

[Dogway]

I haven't looked into IPTPQc2 prerequisites for encoding, but I know x265 supports ICtCp. With these models 12-bit is encouraged but I'm not sure what kind of hindrances this carries later for compatibility (personal use). As I see all them use Planar 420 (fourcc: I420)

[benwaggoner]

Originally Posted by Dogway

I haven't looked into IPTPQc2 prerequisites for encoding, but I know x265 supports ICtCp. With these models 12-bit is encouraged but I'm not sure what kind of hindrances this carries later for compatibility (personal use). As I see all them use Planar 420 (fourcc: I420)

I don't know of any hardware that can decode 12-bit ICtCp. That's the nominal internal color volume used by DoVi in its intermediate stage, but I've never heard of it being directly encoded nor directly displayed. It's the input for DoVi's tone mapper to get to native panel space.

[Dogway]

As per the papers it's a consumer oriented storage color model since it saves bits (and bandwidth) specially for low luma values and luma/chroma decorrelation is better than YCbCr or YcCbcCrc for HDR.
Yes I don't expect any hardware to be on the edge but I guess if premium stream services include it it must be the streaming apps that decode it to RGB. I personally use the LAVFilters+madVR combo so as far it can be decoded with that I'm fine except madVR's future isn't very bright (not FLOSS).

[nevcairiel]

ICtCp is definitely used by streaming services for Dolby Vision. Its just metadata though, and you can decode it with any ordinary 10/12-bit capable HEVC decoder (depending on the stream), interpretation of the color data is then typically up to the implementation (ie. by using shaders), or using pass-through to the display.

[Barough]

x265 v3.5+21-7155a58f7 (32 & 64-bit 8/10/12bit Multilib Windows Binaries) (GCC 10.3.0)

https://bitbucket.org/multicoreware/x265_git/commits/branch/master

[charliebaby]

Originally Posted by Barough

x265 v3.5+21-7155a58f7 (32 & 64-bit 8/10/12bit Multilib Windows Binaries) (GCC 10.3.0)

https://bitbucket.org/multicoreware/x265_git/commits/branch/master

Thank you for GCC 11.2

[LigH]

New upload: x265 3.5+20-17839cc0d

[Windows][GCC 11.2.0][32/32XP/64 bit] 8bit+10bit+12bit

Bugs not yet fixed:

• NASM 2.15.05 multi-line macro warnings (reported August 2020)

[tormento]

Is there any option to optimize greyscale (i.e. BW movies) encoding. mostly about speed?

[benwaggoner]

Originally Posted by tormento

Is there any option to optimize greyscale (i.e. BW movies) encoding. mostly about speed?

That was something that MCW had been looking at some years ago, but it doesn't look like it made it in.

The good news is that, with 4:2:0 chroma subsampling, only 1/3rd of the samples in a frame are color. And there is coarser quantization as well, so chroma compute is probably less than 33% on net. Setting --cbqpoffs and --crqpoffs to 12 would provide chroma with much coarser quantization, which speeds up some things (but would make actual chroma look pretty terrible).

You'd thus want to filter your source to luma-only before encoding to avoid that kind of problem. And IIRC there are some early exits that turn on when a chroma plane is packed with zeros. Give it a shot.

[asarian]

Originally Posted by LigH

New upload: x265 3.5+20-17839cc0d

[Windows][GCC 11.2.0][32/32XP/64 bit] 8bit+10bit+12bit

Bugs not yet fixed:

• NASM 2.15.05 multi-line macro warnings (reported August 2020)

FYI, I am currently using this version. Even though I have --frames 211420 on the command line, your version says 'unknown frame count'.

[LigH]

Don't blame me, I only run the compilation. The reason may be related to processing a raw video input stream via pipe instead of physical source file. Using x265 linked in a current ffmpeg may give you a better environment.

[asarian]

Originally Posted by LigH

Don't blame me, I only run the compilation. The reason may be related to processing a raw video input stream via pipe instead of physical source file. Using x265 linked in a current ffmpeg may give you a better environment.

Isn't the '--frames 211420' supposed to tell x265 how many frames there are?

[Boulder]

Originally Posted by asarian

Isn't the '--frames 211420' supposed to tell x265 how many frames there are?

The message you are seeing is from the input handling. I don't think I've ever seen a number of frames there when using piping.

The number of frames set in the parameters is only used for calculating the ETA.