F55 RAW Kelvin does not match DaVinci Resolve

We have some F55 RAW clips. When debayering the clips we noticed that the color does not match DaVinci Resolve (when using the same settings).

The clip was debayered as S-Gamut3.cine, S-Log3. Exposure was 1250. Variety of kelvin tests with 3200, 4300 and 5500. What I noticed is the 3200K debayer in Cortex has less red/blue than a debayer from Resolve (waveforms doing an A/B comparison show the green channel is about the same, but red/blue is lower from the Cortex debayer at 3200K. This makes the image look more green overall, slightly darker).

Same test at 4300K. It’s less noticeable, but Cortex is still slightly less red/blue than Resolve (according to waveforms).

Once again with 5500K. This time, the Cortex debayer has more red/blue than the Resolve debayer (looking at waveform). It’s almost like there’s a weird scaling with red/blue and the center point is somewhere around 5000K.

Here’s some side-by-side examples. Resolve on left, Cortex on right. No color has been applied. These are just the debayered images.

All these stills are from DPX renders from Resolve and Cortex, combined in Photoshop. I have uploaded the DPX files to your FTP site. I have also trimmed the RAW MXF file to a single frame and uploaded it as well.

Have you compared these in Sony RAW Viewer?

1 Like

Ahh, I see what you’re hinting at. The Cortex debayer does, indeed, match Sony’s RAW Viewer debayer. (Slight difference, but only detectable on a scope. Visually there’s no difference.)

So this is more of a bug with Resolve. Yikes. I’ll try to debayer this clip through Resolve 11 beta and see if anything has changed or if this Resolve bug has been fixed.