I think it's time we revive this question, now that there's support for both 
LogC and PLogLin (the Josh Pines formula) in Nuke 6.3.

I'm currently doing a film when we want to read plates in Log C using the Pines 
math. The main reason for doing this in Nuke is that Arri's own Log C LUT (as 
well as the original Cineon LUT), are really designed for video output. The 
gamma issue is of course easy to compensate for, but the formula is still 
designed to set black and white points that give you a contrast suitable for 
display. The whole concept of a black point is not really suited for converting 
to a truly linear color space, as it can (and will) give you negative values. 
What we want is to restore the scene linear lighting information, which is what 
Pines is trying to do in an intelligent way.

The question that remains is what settings are suitable when applying the Pines 
math to a Log C image. The defaults are designed to work with Cineon images, 
not Log C. For this reason one should avoid using the PLogLin option on the 
Read node (as it does not offer any settings to tweak). Instead, read the image 
data raw and use the PLogLin node for the conversion. As far as I can figure, 
the log reference value for the 18% gray point is 400, but I'm not entirely 
sure about the rest. I believe the gamma should be somewhat lower for Log C 
than Cineon (which defaults to 0.6). Unfortunately my plates aren't all that 
suitable references for this kind of thing, so if there is anyone out there 
with awesome math skills, please help us out!

/Andreas



_______________________________________________
Nuke-users mailing list
[email protected], http://forums.thefoundry.co.uk/
http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users

Reply via email to