Assuming that a DCG hologram has been red-shifted, it would be necessary to shrink it back down somewhat, to get the color right. Say for example we want a yellow image, but it's orange. And it would be better to shrink it using a "dry" method, so that it doesn't have to go into any liquids again. Also, the hologram must be very dry before it's sealed, so we can't rely on simple moisture-reduction (heating) to get the shrinkage. In other words, if we happen to get the right amount of shrinkage by heating, it's almost certain that it's holding too much moisture. So we need to get it dry enough for sealing, yet still too red, then shrink using some method to get the color right. There may be various methods to do this involving ovens (shrinking), microwave ovens (drying with minimal shrinking), vacuum (?), vacuum ovens (?), etc.
There is an interesting (though rather odd) patent which involves a post-processing method for shrinking a DCG layer, without resorting to re-processing (liquid contact):
http://worldwide.espacenet.com/publicat ... cale=en_EP
The process is to: 1) expose the DCG layer to 68% RH at 30C for 6 minutes, 2) put layer into a rather high (20 microns Hg) vacuum chamber for 2 hours, then 3) expose layer to strong UV light until the right amount of shrinkage is obtained. This is shown on the graph at the end. The authors use the term "Bragg angle deviation" for the factor to be corrected (the replay angle). But expansion of the layer (and subsequent shrinkage) seem to closely correspond to Bragg angle deviation. It's interesting that processed DCG hologram shrinks upon exposure to 68% RH water vapor at 30C (according to the graph). I recall reading elsewhere that water vapor does shrink the processed layer, and something about the Bragg structure "collapsing" due to water penetration.