Re: Uniform Perceptual Lab
Re: Uniform Perceptual Lab
- Subject: Re: Uniform Perceptual Lab
- From: Graeme Gill <email@hidden>
- Date: Fri, 18 Jul 2003 11:43:59 +1000
"Bruce J. Lindbloom" wrote:
>
IPT is one of eight models evaluated and does not have any outstanding rank
>
compared to others, either overall, or even with respect to hue linearity in
>
particular (p. 11). All models are evaluated using the Munsell renotation
>
data. As I said before, the only way to really achieve the ideal performance
>
is to use a LUT as I have done. My ICC profile would have basically a zero
>
error for all metrics discussed in the above paper. The only error would be
>
due to interpolation errors that always accompany LUTs, but these errors
>
would be extremely small compared to those of any of the models.
A similar idea was used by Braun & Fairchild to Hue-Linearise
the CIELAB Color Space. For those interested in the details,
see <
http://www.cis.rit.edu/people/faculty/fairchild/PDFs/GusCIC98.pdf>
I notice in the work surrounding CIECAM97, that there are other data
sets being used, as well as the Munsell data, for evaluating and
tuning color appearance models.
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.