dwatola@NEXTASY2.EECS.WSU.EDU (David Watola) (06/06/91)
(to my description of a problem with background colors on monochrome vs. color machines): Eric P. Scott writes: > Your image has an alpha channel and you're not compositing it properly. and Izumi Ohzawa writes: >So, it is one of the following: >[1] NeXT goofed. >[2] Pixar goofed. >[3] It's a feature of NeXT alpha channel. > >My guess is [3]. yes, it does appear (in the literal sense) that the alpha channel is the source of the problem. by my images are NXEPSImageReps that are rerendered from encapsulated postscript (generated by mathematica, actually) whenever they are redrawn. this is definately true in all the circumstances under which i have used the program--the rerendering was fairly obvious to the eye, i.e. my images were obviously not merely being composited from an offscreen window since they are only redrawn when they are rescaled. unless mathematica output uses alpha, this is not the problem. of course, there is the possiblity that the drawRepresentation:inRect: method of NXImage uses alpha (improperly?). anybody got a definitive answer to this? also, Eric's original response raises another point. hypothetically, if i had an image with an alpha channel and i composited it 'improperly' (hmmm... actually, i used NXImage as documented, for its intended purpose-- how much more proper can you get), i don't really understand the different output. after all, display postscript means device independence, does it? even if results aren't what i expect, they should be consistent beteen two 'identical' display postscript implementations, yah? or are there some subtle differences i am unaware of? hey, since this topic isn't extremely critical to me, i'll settle for idle speculation as well as wizardly analyses. thanks, all. dave watola dwatola@yoda.eecs.wsu.edu