Idaho Airships, Inc.

No doubt-this is probably only useful in an instructional environment. I was recently doing some Photoshop consulting for the F. Edward Hébert School of Medicine in Bethesda, Maryland…

Our challenge concerned RGB Channel Separations in Photoshop, specifically for some fluorescent microscopy. Along the way I stumbled across something that is pretty interesting from both Photoshop and Color Science instructional perspectives, though it is of little or no value in commercial photography.

Here’s an Action (.ATN file) that takes an RGB image (not tested with Alpha Channels) and splits the Channels into standalone Documents. Simple enough. It then colors the Documents just as you would see in Edit>Preferences>Interface>Show Channels in Color. Incidentally, there’s a slight system performance boost by showing Channels in Grayscale…just in case that’s an issue.

Now you have the option (I can’t do it with Actions) to reconstruct a Document that has Layers representing the original Channels! Simply use the three new Documents from the above Action as Layers (by Drag & Drop if you wish), then set the top two Layers to the Difference Blending Mode. I did some cursory checking with Color Samplers (no stats) and everything looks very solid. Have your Notes Window open for an explanation (Window>Notes).

Now, I’m not really qualified to explain (guess) exactly what’s happening behind the Channels scene. We’re all aware that the new Haeberli Vectors are: R = .212671; G = .715160; B = .072160…ain’t we?

Also-if you’re really wacked about Photoshop you might try reassembling the image using Apply Image, but you’ll end up with a single Layer. Where’s the fun in that?

My non-Rube Goldberg solution was to simply blacken the various Channels in a Document (or Document Copy) to isolate Channels. In other words, if you want a Layer representing both G and B, just darken the R Channel-color it in with Black. Viola’.

Here’s the Action (CS4).

L