> ----------
> From: Post David L Civ AFRL/HECV
> Sent: Wednesday, October 6, 1999 9:07 AM
> To: 'Color and Vision Network'
> Subject: RE: CVNet - query on color conversions
>
> There are many ways to do this, but the basic, standard approach is:
>
> 1. Measure the chromaticity coordinates of the display's RGB channels.
>
> 2. Establish functions relating the RGB values to RGB luminances or
> (following the Berns, Motta, & Gorzynski method) RGB display tristimulus
> values. This is where most of the variations in methodology occur. You can
> sample the RGB channels separately, or sample the display's neutral point
> (R=G=B/=0) and decompose the measurements into their assumed RGB
> components using the info from step 1. If your measuring instrument has
> sufficient sensitivity, it's a good idea to include a measurement at
> RGB=0,0,0 and correct the other measurements and later computations for
> this "black light." As for functions, you can use piecewise linear
> interpolation in either the linear or log-log domains (the latter works
> better), or you can use closed-form functions -- Berns et al.'s gives the
> most accurate results of those that have been offered so far.
>
> 3. Use the info from steps 1 and 2 to predict the CIE tristimulus values
> that will be produced by given RGB inputs.
>
> 4. Convert to L*a*b* in the usual way.
>
> Dave Post
> --------------------
> From: "Victor Rush" <rush@ceeb.uky.edu>
> To: cvnet@lawton.ewind.com
> Subject: converting RGB to L*a*b standard
>
> I was wondering if anyone can tell me how to convert raw RGB
> values into L*a*b standard. We have the data from Photoshop as
> RGB and dont want to reanalyze the samples (well over 1000
> pictures). If someone knows of a conversion program etc... it
> would be appreciated.
>
> I do understand that different RGB values from different machines
> will cause variation in the L*a*b values. But actually that is part
> of the question that we are interested.
>
> Thanks.
>
> Victor Rush
> Victor Rush
> Research Ass. Prof.
> Center for Ecology and Evolutionary Biology
> T. H. Morgan School of Biological Sciences
> University of Kentucky Lexington KY 40506
> ph 606-257-8742 fax 606-257-1717
> email rush@pop.uky.edu
>