XPost: sci.image.processing, rec.photo.digital, comp.graphics.apps.gimp
of how I understand it, RIMM is based on Pro Photo RGB, originally for
film i/o
ERIMM is a bigger derivative for digital camera input
they are both compromises for input profiles, they don't characterize a particular camera/scene
the problem as I see it is that making ONE input profile for a digital
camera requires some appearance considerations, white point for
instance, or a full appearance implementation like CIECAM02
maybe the camera manufacturer could run some gamut of appearance
considerations on the camera, this would of course require the
accompanying selector tool for the photographer to identify such
appearance considerations from tags in the input profile to the
rendering engine, I don't know if ICC, etc., have appearance matching yet
if you can use a target in your scene, there is no need for this,
although it might be nice to have a better "generic" profile
maybe Pro Photo RGB or ERIMM is good enough when you don't have a scene
where you can use a target, leaving the rest to editing
How does ROMM working for translucent/transparency output? or have video
or even modern print replaced these?
--
Dale
http://www.dalekelly.org
--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)