<br><br>On Saturday, November 6, 2021, Terje J. Hanssen via Cin <<a href="mailto:cin@lists.cinelerra-gg.org" target="_blank">cin@lists.cinelerra-gg.org</a>> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><br>
<br>
Den 05.11.2021 11:55, skrev Andrea paz:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
@Terje<br>
If I understand correctly, you used only the h264.mp4 and h265.mp4<br>
presets, changing the "Pixels" option from "420 8-bit" to "422 10-bit"<br>
each time. Also, try using the 8, 10 and 12-bit h265 presets; they are<br>
Andrew's new ones that work for me in the non-multibit version.<br>
I've tried non-multibit and I can render h264.mp4 at 8 and 10-bit and<br>
h265.mp4 at 8 and 10-bit. In short, in my case the non-multibit<br>
version always behaves as a sum of multibit and non-multibit.<br>
<br>
</blockquote>
<br>
@Andrea and All<br>
I had a look into the Manual: Modifying FFmpeg Format Options inside CINELERRA-GG<br>
Figure 9.2: FFmpeg wrench, video preset, view and format options<br>
<a href="https://cinelerra-gg.org/download/CinelerraGG_Manual/Modifying_FFmpeg_Format_Opt.html" target="_blank">https://cinelerra-gg.org/downl<wbr>oad/CinelerraGG_Manual/Modifyi<wbr>ng_FFmpeg_Format_Opt.html</a><br>
<br>
and tried to indicate cin version and parameters in my test file names (no warranty the syntax is quite consistent), i.e<br>
<br>
hd01_cin_appimage_ffmpeg_h264_<wbr>yuv422p10le.mp4<br>
File > Render | File format: FFMPEG mp4 | Video Wrench<br>
> Video Preset | Compression: h264-10bit.mp4 | Pixels: yuv422p10le<br>
<br>
While this rendered OK on one of my workstation, another installation wouldn't render at all with the following<br>
Message log:<br>
virtual void Render::handle close event(int):<br>
Create new at labels checked, but no labels<br>
(or other Failure)</blockquote><div><br></div><div><br></div><div>guess in this case you (accidently?) selected rendering option making new file at each label (3rd radiobutton out of four) instead of rendering whole file/ in-out region.... </div><div> <br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<br>
------------------------<br>
<br>
Regarding ffmpeg (before going further with testing):<br>
Is it a somewhat correct understanding that rendering (encoding) inside Cin-GG (AppImage) works as a GUI front-end for its statical linked ffmepg?</blockquote><div><br></div><div>usually yes, you can try to build it differently</div><div><br></div><div>./configure --help | grep thirdparty</div><div> --with-thirdparty use thirdparty build (yes)</div><div><br></div><div>but this usually will have chance to work only if system ffmpeg at same version as internal copy... Also, internal ffmpeg patched quite heavily... </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<br>
Even if my local system's ffmpeg is not used, I think 3 ffmpeg commands (applied from stackexchange) possibly add understanding also for rendering via Cin-GG:</blockquote><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<br>
---------------------<br>
1) To see what pixel formats and bit depths are supported by libx264:<br>
<br>
ffmpeg -h encoder=libx264 | grep Supported<br>
<br>
ffmpeg version 4.4 Copyright (c) 2000-2021 the FFmpeg developers<br>
built with gcc 7 (SUSE Linux)<br>
[ffmpeg text header .........] --enable-libx264 --enable-libx265 --enable-librtmp --enable-libxvid<br>
<br>
Supported pixel formats: yuv420p yuvj420p yuv422p yuvj422p yuv444p yuvj444p nv12 nv16 nv21 yuv420p10le yuv422p10le yuv444p10le nv20le gray gray10le<br>
<br>
(It seems to me that both libx264 and libx265 are enabled in this case)<br>
<br>
Is it possible that Cin and Cin-multi have statical linked ffmpeg with both libx enabled?</blockquote><div><br></div><div><br></div><div>see above, there is dependency in thirdparty/Makefile so x264/x265 built first, and then ffmpeg built pointed at those libs. Still, I need to double-check for possible missing static switch.. On termux x265 built with only 8 bit depth for system, so I need patch and CinGg renders h265/h264, so I guess this part works here... </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<br>
----------------<br>
2) and by libx265, here with suppressed ffmpeg text header (-v quiet):<br>
<br>
ffmpeg -v quiet -h encoder=libx265 | grep Supported<br>
<br>
Supported pixel formats: yuv420p yuvj420p yuv422p yuvj422p yuv444p yuvj444p gbrp yuv420p10le yuv422p10le yuv444p10le gbrp10le yuv420p12le yuv422p12le yuv444p12le gbrp12le gray gray10le gray12le<br>
<br>
In both cases 10-bit pixel formats are those that end with 10le.<br>
<br>
---------------<br>
3) ffmpeg with the -codec switch, you will get an output of (all) codecs it understands. The codecs are prefaced with letter codes that describe their function. 'D' means Decode, meaning that particular codec has decoding capability (read). While 'E' means Encode, or compiling/writing capability using that particular codec.<br>
<br>
ffmpeg -v quiet -codecs | egrep "x264|x265"<br>
<br>
<a href="http://DEV.LS" target="_blank">DEV.LS</a> h264 H.264 / AVC / MPEG-4 AVC / MPEG-4 part 10 (decoders: h264 h264_v4l2m2m h264_qsv ) (encoders: libx264 libx264rgb h264_qsv h264_v4l2m2m h264_vaapi )<br>
DEV.L. hevc H.265 / HEVC (High Efficiency Video Coding) (decoders: hevc hevc_qsv hevc_v4l2m2m ) (encoders: libx265 hevc_qsv hevc_v4l2m2m hevc_vaapi )<br>
<br>
<br>
----------------<br>
<br>
Terje J. H<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
-- <br>
Cin mailing list<br>
<a href="mailto:Cin@lists.cinelerra-gg.org" target="_blank">Cin@lists.cinelerra-gg.org</a><br>
<a href="https://lists.cinelerra-gg.org/mailman/listinfo/cin" target="_blank">https://lists.cinelerra-gg.org<wbr>/mailman/listinfo/cin</a><br>
</blockquote>