[issue_resolved] But you're right. The old MPC-HC build 1.3.1299 might be the culprit of wrong report. Even thou LAVVideoFilter V.0.49 does the decoding, oldy mpchc1.3 gives back that weird report. Newer 1.5/1.6 builds report same streams as 10bit. But does decode of 10bit really depends on movie player w/ internal filters disabled?
Author: Spec-Chum <Spec-Chum@10f7b99b-c216-0410-bff0-8a66a9350fd8>
Date: Fri Oct 9 21:47:39 2009 +0000
Changed: "DirectShow Audio" label on playback->output now says "Audio Renderer"
Somehow I'm not surprised a Mediainfo from 2009 wouldn't "see" the difference between different bit depth H.264 and would just show everything as "8bit"
In fact, i really have four long reencodes w x86 build that can be obtained from fushizen before i noticed that MPC-HC really show 8bit resolution for all of them and i thought that fft3dgpu induce something to appears as color banding at low bitrates (nevertheless that fft did quite the opposite in the past for me)> fft3dgpu
I cannot use x86-64 builds because fft3dgpu is still 32bit only
> wouldn't induce banding
It's a denoiser. Depending on usage, it is bound to do damage to gradients and related things. On the other hand, if you can actually find something that looks like a broken encoder feature, I'd be glad to check out a short y4m dump of the source that reproduces the issue, and a minimal command line for the actual act of reproducing it
. But somehow I am relatively sure this is not a case of an actual breakage
Also, 1) Don't use 64bit Avisynth, and 2) Check out avs2pipemod
for piping stuff from 32bit Avisynth to 64bit x264.
I just thought this one might be wrongly patched or something by someone else as the blog went offline. In fact im glad to know these are still yours patched builds.
Bugs in patches are always possible, which is why I do a couple of test encodes every time I compile with every build
. Of course, I can't be testing everything, so proper bug reports are always welcome