TIFF and LibTiff Mail List Archive

1998.03.06 18:42 "CIE L*a*b*, particularly Adobe Photoshop's concept of it", by Dale Carstensen

Hello All,

I have some files that Adobe Photoshop wrote in LAB colorspace. I need to use them with programs that only support RGB. The tifftopnm program rejects them saying

 tifftopnm: unknown photometric: 8

Tiffcp recognizes the colorspace

   Photometric Interpretation: CIE L*a*b*

but doesn't have any option to convert it to RGB that I see. My own programs that read TIFF can stupidly copy L* to R, a* to G, and b* to B, of course, but they look kind of dark and strange that way, almost like negatives but not quite.

I found several resources about CIE L*a*b* via the web and one book I have, the XIElib book by Syd Logan published by Prentice Hall. The best web resource, so far, seems to be Charles Poynton's FAQ on color. He also has one on gamma, and both can be accessed from

in text, PDF and other forms.

But either I am unable to understand how to apply the formulas (I am confused about what Xn, Yn and Zn are supposed to be) or Adobe has cooked up some entirely different version of "LAB." A few messages I found on dejanews accused Adobe of the latter. But none had a conversion formula, other than the standard ones of CIEL*a*b* to CIEXYZ to RGB that don't work, at least as I understand them, for Photoshop LAB to RGB.

Anybody have a Photoshop LAB to RGB method that works?


  Mr Dale Carstensen
  LAMP, Inc.                   office (505)662-2524
  1504 S Sage                     FAX (505)662-3588
  Los Alamos, NM 87545-3037