From: Jean-Marc V. <ver...@if...> - 2000-06-12 22:53:03
|
As promised a while ago I have worked more on merging my work with the current repository. Unfortunately I ran out of ink for several days, hence the delay. Another reason it took that long is that I had overlooked recent dithering progress, which is far too complex for me to understand easily. This whole thing helped me improve my code a little. However, what was bothering me most was fluffy rendering in some areas or other artefacts depending on dither used. I found where this came from. If I understand well, this is from the (now not so) new matrices unfortunately. i.e., either (in print_color()): default: vmatrix = DITHERPOINT(d, x, y, 6); (vs DITHERPOINT(d, x, y, 0)) or else if (rangepoint >= DITHERPOINT(d, x, y, 5)) (vs DITHERPOINT(d, x, y, 3)) or both produce this result. I don't know what to think about this and I would like some advice. To me it looks like these matrices semi-randomly spread out ink on a relatively large surface area. To me again, that makes perfect sense for printers with few ink levels. I'm not that convinced that it should be the case to the same extent for 6 ink level printers such as the Epson 870. In my opinion these printers have the ability to stick to colors more efficiently. That's my little understanding and I may be completely wrong. The bottom line is that these matrices yield strange light dot clusters on the Epson 870 in my hands. I've not posted a new version because I feel that I need to understand the above for any successful merging of ideas. What are your comments ? -- -- Jean-Marc Verbavatz <ver...@if...> 5, rue La Fontaine "http://perso.cybercable.fr/verbavat" F-75016 Paris |