2005.05.28 18:01 "Re: [Tiff] How to Write a tif file with compression =COMPRESSION_JPEG& photometric = PHOTOMETRIC_YCBCR ?", by Joris Van Damme

Jean-Yves,

In my humble opinion, if a multichannel colorspace start with a brightness channel, and the SamplesPerPixel value equals 1, that is a unambigious situation, and may be preferable over storing grayscale as full YCbCr.

Yes, but do this conform to TIFF specifications?

I thought about this further...

Apart from including ICC profiles, our best option to communicate grayscale in a scientifically unambigious way, is probably by using the L* component of L*a*b*. If we were to say that Photometric=YCbCr and SamplesPerPixel=1 is not legit... I don't think we have a legit way to store L* rasters, either.

That may be a to pragmatic argument to be convincing, but I would like to add to that, again, that Photometric=YCbCr and SamplesPerPixel=1, or Photometric=whateverLab and SamplesPerPixel=1 is really unambigious. I would hesitate to write it, for now, unless people here clearly indicate they feel the same about this as I do, but I would certainly think you'll need to be prepared to at least read it.

Joris Van Damme
info@awaresystems.be
http://www.awaresystems.be/
Download your free TIFF tag viewer for windows here:
http://www.awaresystems.be/imaging/tiff/astifftagviewer.html