Re: Digital Camera Profiling...Why can't we all just get along?
Re: Digital Camera Profiling...Why can't we all just get along?
- Subject: Re: Digital Camera Profiling...Why can't we all just get along?
- From: Graeme Gill <email@hidden>
- Date: Wed, 23 Apr 2003 12:27:19 +1000
Roger Breton wrote:
>
What kind of 'model' will separating characterization from intent afford us?
One that simplifies the purposes of each part of a color transformation.
The device profiles let us work in a device independent space, and then
in that space we can choose the intent/look/adjustment we want.
Just as separating color transformations into a pair of device profiles
reduced the problem from an N! problem to an N problem (given N different
devices, there are N factorial direct color conversions, but you only
need N device profiles if you use them in pairs), given N devices and
M intents, combining the intents into the profile gives you an N x M
problem, while separating them gives you an N + M problem.
>
Do you foresee profiles only carrying characteriztion data, leading us into
>
"intelligent CMM"? Meaning that rendering is best left as a 'run-time'
>
decision?
Yes. The "dumb CMM" reached its limits a while ago.
>
Have you attempted that in Argyle?
Argyll has always been an "intelligent CMM". The gamut
mapping results, accuracy and flexibility (but not speed)
are noticeably superior using smart linking, and it is
also much easier to know what is going on, without
mysterious unknown transforms being embedded in the
profile data.
Graeme Gill.
_______________________________________________
colorsync-users mailing list | email@hidden
Help/Unsubscribe/Archives:
http://www.lists.apple.com/mailman/listinfo/colorsync-users
Do not post admin requests to the list. They will be ignored.