On Sat, 15 Jan 2005 23:42:55 +0100, Gerhard Gauling 
> If you apply a conversion to the file there will be a loss of color
> information, so it's necessary, that we avoid unneeded conversions to the
> original file.
> 

For some filters and operations to be correct, the image data needs to
be in a color space with linear light. (e.g. not gamma corrected for
CRT display devices).

Operations that need linear light to be correct:
  gaussian blurs
  smudge
  anti aliased brush strokes
  soften
  brightness / contrast adjustments
  resampling (scaling an image up/down, rotating etc.)

To avoid extra conversions whenever such a filter is applied to an
image, having a standard working space that has linear light will
decrease the total number of conversions needed. Doing the custom
conversion and countermeasures to account for gamma in all these
operations will introduce quantification whenever such an operation is
performed, less conversions are better, and a consistent high range
internal working space is the path to the lowest number of conversions
whenever any of the above listed operations are done.

This is why I am advocating that gimp/GEGL should prefer to work, save
and load in a preferred working color space.

For multiple loads/edits of a project the conversion to a preferred
internal working space already have happened when you load an image
using gimps own file format. other image file formats are not suited
as intermediate formats for saving an image that is being edited (for
the same reasons that you shouldn't use lossy compression,.
compression through generations/multiple unneeded conversions degrades
your image)

When exporting from gimps internal file format to display / printer /
other file format with embedded icc profile, the image data shouldn't
be touched by gimp again.

/yvind K.

-- 
Software patents hinder progress | http://swpat.ffii.org/ 
Web :  http://pippin.gimp.org/
_______________________________________________
Gimp-developer mailing list
Gimp-developer@lists.xcf.berkeley.edu
http://lists.xcf.berkeley.edu/mailman/listinfo/gimp-developer

Reply via email to