some notes to all the good info already suggested:
- if you do your gamma correction in photoshop, use the gamma in
"exposure" rather then "levels". last time i checked, the one in
levels was not a mathematical gamma correction but has some
adobe visually pleasing stuff added.
- also, do your gamma correction in 32bit, then first drop down
to 16bit to undo the gamma, then drop to 8it. this will avoid
banding issues.
- and lastly, note that photoshop introduces dither when
dropping to 8bit. this might or might not be what you want.
so basically, you're probably better off comping in sRGB in nuke
(so that it looks slightly wrong), then write out exr, import
into PS, dropping to 8bit (so they look like you would have
comped in nuke linearly). can't see a way to avoid the dither
issue if you take the exr route though.
++ chris
On 3/8/12 at 5:17 PM, [email protected] (Simon Björk) wrote:
After, as it will make the layers blend in linear light
instead of sRGB. The overall gamma will be the same as the
sRGB viewer lut you have in Nuke. Although, you might
introduce problems with banding as your really bending the
colors of 8-bit. Of course you will also clamp values
above 1. It might be better to do your compositing in sRGB
space in Nuke as other suggested, but it's worth a try.
Essentially, this would be somewhat the same thing as
using a linear ICC profile in PS.
_______________________________________________
Nuke-users mailing list
[email protected], http://forums.thefoundry.co.uk/
http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users
