I've made a video that was filmed with a Canon 60D. It has been graded and edited in Davinci Resolve.

The original footage from the camera is 8-bit. But what happens in the grading process? Will I use nuances from the grading if I archive it as 8-bit?

I understand there's no point in converting the 8-bit video to 16-bit when unedited, but how about when it's graded?

1 Answer 1


Yes, you will lose information from the grade if you save as 8-bit.

Here's a thought experiment to illustrate. Say your original red level for a pixel was 128 out of 256 (00001111 / 11111111), and then in Resolve you grade it (using the internal 32-bit floating point colour in Resolve) to be a little bit brighter, say 128.25. Now you'll need to save it in 10 bits or more to keep that information (e.g 000111101 / 1111111111 where the two right-hand-side bits 01 / 11 = 1/4 = 0.25). If you saved it as 8-bit it would have to be rounded to 00001111, meaning it would go back to being 128.

The upshot is, if your new colours don't fit into the 256 values per channel of 8-bit colour they will get rounded to the closest fit. So while you don't gain anything from up-depthing (that's a thing now, because I just invented it) ungraded footage, you do gain by not collapsing the higher-bit-depth graded footage back down to 8-bit.

  • Of course, if the new red is 128.13, then 10 bits is insufficient. Considering that the end-user display is almost always 8-bit, the question is, do the extra storage bits make a difference to the final 8-bit raster that's displayed, or does it simply shift the burden of the rounding step downstream?
    – Gyan
    Sep 5, 2017 at 14:01

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.