DaVinci Wide Gamut & DaVinci Intermediate info

I can't seem to find any documentation showing what the DaVinci Wide Gamut primaries and white-point actually are, nor can I seem to find any documentation on what the DaVinci Intermediate OETF/EOTF is. The closest is simply a passing image reference in the manual.

Can someone from BMD please provide more info, or a reference to more info?

I posted this question in the Developer section, but there's an overabundance of tumbleweeds there.

Seems no any specs for DaVinci Intermediate Log curve yet, but you can see DaVinci Wide Gamut in the CIE scope.
Some info was posted by CaptainHook here:
viewtopic.php?f=2&t=125084#p686105

Thanks Dmitry. Got the Color Space, now need the gamma info.


Can someone from BMD please post the Gamma EOTF/OETF equations for Intermediate?


This information is quite important, given we need to align with the working space, especially for Developers.

Can someone from BMD please post the Gamma EOTF/OETF equations for Intermediate?

While at it, gamuts and curves of colorspaces used by BMD cameras would be nice too.

You can get the Gamut info thru the CIE chart.

Gamma details still seem to be kept under lock and key though.

Still no info on the DaVinci Intermediate Gamma curves. Can we please get some clarification.

This is now available on the Support Page

I've just noticed a discrepancy between the documentation and Resolve for DaVinci Wide Gamut.

In the DaVinci Wide Gamut Intermediate documentation, it lists the color primaries as:

red 0.8000 0.3130
green 0.1618 0.9877
blue 0.0790 -0.1155
white 0.3127 0.3290


Inside Resolve, the CIE Chromaticity scopes list the color primaries as:

red 0.8000 0.3130
green 0.1682 0.9877
blue 0.0790 -0.1155
white 0.3127 0.3290


Which is correct?

And the plot thickens...

I double checked a RGB XYZ matrix I had originally calculated using the primaries derived the Resolve 17 Chromaticity readout and my calculations are an exact match (plus some extra precision) of the RGB/XYZ matrices that have been quoted in the documentation.


WG RGB to XYZ:

Code: Select all

0.7006223920936714,     0.1487748151231968,     0.1010587198348032
0.2741185109066489,     0.8736318959404368,     -0.1477504068470858
-0.09896291288323111,   -0.1378953250755433,    1.325915988718653

WG XYZ to RGB:

Code: Select all

1.516672042024044,      -0.2814780478789692,    -0.146963633236778
-0.4649171012332758,    1.251423775681712,      0.1748846088650907
0.06484904706715998,    0.1091393437105699,     0.7614146215498789

However, If I use the chromaticity coordinates noted in the Wide Gamut documentation, the calculated matrices are way off.


The documentation is clearly incorrect.

Either the quoted Matrices are wrong, or the chromaticity coordinates are wrong.

Looking at the pdf revision history, it mentions that the incorrect matrix values were fixed, so my assumption would be that the primaries are wrong (specifically the green x value which should be 0.1682).

Problem also discussed here https://github.com/cameramanben/LUTCalc/issues/17