Without hardware acceleration, a typical command would be ffmpeg -i input.mkv -pixfmt yuv420p10le -c:v libx265 -crf 21 -x265-params profilemain10 out.mkv. You can use the format filter, or the legacy -pixfmt option, to set the pixel format. I am trying to encode a 10-bit H.265 video from a 8-bit H.264 source using ffmpeg with CUDA hardware acceleration. In ffmpeg yuv420p is called a pixel format.

yuv420p is a chroma subsampling scheme, and the p stands for planar, not progressive. Handler_name : IsoMedia File Produced by Google, 5-11-2011įfprobe output_264.mp4 Input #0, mov,mp4,m4a,3gp,3g2,mj2, from 'output_264.mp4':ĭuration: 00:07:50.02, start: 0.000000, bitrate: 1006 kb/s Use yuv420p There is no such thing as yuv480p or yuv720p because it is not tied to the video height. How do I get an RTMP stream out of this without encoding to h264 again.

Im picking this thread up again because now Im facing the third and hopefully final challenge. Stream #0:0(und): Video: h264 (libx264rgb) (avc1 / 0x31637661), rgb24(progressive), 636x360, q=-1-1, 27.05 fps, 13524 tbn, 27.05 tbc (default)Ĭpb: bitrate max/min/avg: 0/0/0 buffer size: 0 vbv_delay: -1 So basically decoding h265, writing to disk as h264 and forwarding a downscaled yuv420p stream to detection.

H264 yuv420p