Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!iuvax!rutgers!columbia!cs!djk From: djk@cs.cs.columbia.edu (David Kurlander) Newsgroups: comp.graphics Subject: Re: Looking for Slide Makers Message-ID: Date: 16 Aug 89 18:44:32 GMT References: <19243.24E6C3DF@urchin.fidonet.org> <11605@cit-vax.Caltech.Edu> Organization: Columbia University Computer Science Lines: 21 In-reply-to: kalra@rodan.caltech.edu's message of 15 Aug 89 21:14:45 GMT In article <11605@cit-vax.Caltech.Edu> kalra@rodan.caltech.edu (Devendra Kalra) writes: > A few years ago, I played with the Matrix QCR-D model. I noticed serious > banding problems. I seemed to be getting only 6 bits of resolution > rather than the 8 that it advertised. > Any comments? We also get only 6 bits of color resolution out of our Matrix QCR-D film recorder. If you examine the tables that map input intensity to colors for this device, you'll see that several input values are often mapped to a single output value. I wrote a program which uses the Floyd Steinberg error propagation algorithm to compensate for the QCR-D's standard lookup table. The results are significantly better -- banding is often no longer a problem, and is always significantly reduced. The folks at NYIT use a similar approach. Still, it would be nice to have a full 8 bits of output! David -- ---------------------------------- David Kurlander Department of Computer Science djk@cs.columbia.edu Columbia University