Hello.
I’m wondering if “color mapping” can always be done in post?
I’ve been tweking the curve in photoshop but results are not that good!
When using “HSV exponential” I have light intensity at 3 and with linear I use 1.
Is it possible to render in linear and then get the same result as exponential in post or is something lost in the process?
Thanks
depends on the color depth you render to…
if you do color mapping while rendering, you still can get the full range within the 8 bit per channel
if you change colors in post using 8bit images, you are stretching the color range
probably thats causing the problems you are having in ps (typically these kind of manipulations show as banding in formerly smooth gradients)
try to render to a 16bit format (f.e. png) or even use openexr for float color
- that should lead to better results
try to render to a 16bit format (f.e. png) or even use openexr for float color
OpenEXR while useful, seemed a bit jumpy contrastwise in PS for me.
On the other hand, how easy is a 16bit png to adjust - does PS give you opening dialog options for this? I take it that you have to apply ‘save to’ prior to rendering. I guess I’m about to find out…
don’t know about openexr, but as far as i have read there is a growing number of people using it for post (esp. in combustion and other post tools)
png will just open in 16bit per pixel mode in photoshop (CS)
no problems so far and you can still use the levels/curves and most other image adjustment tools
most filters wont work in 16bit tho’ (there are special filters like optipix from http://www.reindeergraphics.com/ that work in 16bit and the recently announced photoshop cs2 is supposed to fully support 16bit/32bit/float images)
Wow. I’m never going back to any other format as also knocks out the alpha channel…
…besides, you can always save down.