Stochastic screening
joe <[email protected]>
| Newsgroups | gmane.comp.printing.ghostscript.devel |
|---|---|
| Message-ID | <CAF9ycRPnHkjaQdeohCpcczgd28D3TdVj2AQNYACgsVeKGH0H0A@mail.gmail.com> |
Hi. I've been playing around with stochastic screening for art projects (I believe the stochastic screen patents all expired this year?). I'm planning on releasing an OSS tool to generate the things, and I was wondering if anyone could help me figure out how to get a printer (either the one I have, or a new one) to use postscript defined screens. I'd also like to generate new built-in screens for ghostscript, but high quality ones (thus the need for print testing). Also, anyone know if it's worth generating four-plane (no rotation) screens? It's a set of four stochastic screens that do not require rotation. Of the four screen generation methods I've tried so far, only the one with the worst results is capable of generating the things. I'm curious if it's better to just generate four normal, high-quality screens and rotate, then four low-quality ones that don't need rotation. You can see a demo of mine that uses stochastic screens at: https://github.com/joeedh/BlueNoiseStippling It's designed to (eventually) generate silk screen separations, something I think ghostscript could do as well. It's a bit odd in that is uses error diffusion to make separations, which seems to work pretty well if you have enough colors (I was planning on doing an 8-color silk screen print), not so much for basic cmyk. Best, Joe