HDR color balance = strange math

After watching

, I decided to try his approach on some footage I'm grading. I discovered something strange about the difference between using the Primary Balance's Lift ball on my mini-panel to correct the color balance of the black point vs. using the HDR Dark or Shadow balls. In short, the Primary color balance operation works as expected, with the whole plot moving around as expected. Using the HDR balls works not as expected, immediately shrinking the plot to half its normal size before subsequently moving it around. It also radically shifts the extents, from skewed between Y/R to a large M excursion.

Here is a screen capture of the two methods. Both vector scopes are showing the Low range (0-0.30).

https://www.dropbox.com/sh/ic6ibd75d3it ... giQca?dl=0

This is with DR 17b8 (build 27).

BTW, the big YouTube banner above is NOT what I saw in the Preview. The preview correctly showed my text linked to a YouTube URL. The text said "Darren Mosten's excellent tutorial on his use of vectorscopes". I regret that the forum software has not maintained the integrity of my editorial decisions.

Hilariously, as I post this, the text of my previous post is shown below, and is editorially consistent. So I'm really just apologizing to those of you who read these forums, but do not comment.

I had trouble viewing your video, but what color management are you using, how are the clips tagged color-space-wise, etc? This is beta and there could be a bug, but the HDR controls are colorspace-aware while I think Primary isn't, which could make a difference.

To me this ”jump” looks like a mismatch between default state, where some processing is probably bypassed, and full processing. Doing a minuscule change shouldn’t produce such an instant difference.

I'm using Davinci YRGB (default) color management, with RedWideGamutRGB/RedLog3G10 timeline and Rec2020/ST2084PQ output color space. The media is R3D media.