Our workflow involves huge 8K clips(i.e., 8192x8192). The inputs are mostly very high bitrates 264/265 clips, though we are using ProRes or DNxHD for post-production within the process, we do not want to archive such huge files as the "master" copies(easily over 100GB for 1 minute).

My current approach is, using FFMPEG to compress the ProRes/DHxHD clips into H.265 clips using a very high-quality setting (crf=12, slow profile).

Another quality related setting is the pixel format config. Then my question comes up:

Targeting the same clips(5 of them), using the same compress settings, one with yuv420 8-bit, one with yuv444 10-bit, yuv444 outputs are constantly smaller than yuv420 outputs(5%-10%), i.e, lower bitrates.

If yuv444 10-bit needs to store more information, how could this be possible?

  • I can't answer your question, but may I ask which camers you are using? As not even RED's r3d codec at 8k uses so much storage. Thanks. Jan 8, 2019 at 23:00
  • Yes, this is not from a single camera, they are VR content, a multiple camera setup.
    – Jerry Tian
    Jan 9, 2019 at 2:39

1 Answer 1


From Why does 10-bit save bandwidth (even when content is 8-bit)?

When encoding with the 10-bit tool, the compression process is performed with at least 10-bit accuracy compared to only 8-bit otherwise. So there is less truncation errors, especially in the motion compensation stage, increasing the efficiency of compression tools. As a consequence, there is less need to quantize to achieve a given bit-rate.

The net result is a better quality for the same bit-rate or conversely less bit-rate for the same quality: between 5% and 20% on typical sources.

See the link for additional info.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.