Talk:Color quantization: Difference between revisions
m
→J solution
(J solution: a suggestion) |
m (→J solution) |
||
| (5 intermediate revisions by 2 users not shown) | |||
Line 6:
== J solution ==
Is it possible to change the metric, to put more weight on green, less on red, even less on blue? The result would probably look better (currently the palette isn't able to resolve the toes under the frog's belly). --[[User:Ledrug|Ledrug]] 23:50, 12 August 2011 (UTC)
:Yes, I should think about that. Currently, I am using "color distance". I'll think about other mechanisms. --[[User:Rdm|Rdm]] 00:07, 13 August 2011 (UTC)
:I have posted a simple approach... perhaps I should also consider using different color representation schemes? --[[User:Rdm|Rdm]] 01:17, 13 August 2011 (UTC)
::I now have four implementations out there, and I have fixed a bug in my initial implementation. --[[User:Rdm|Rdm]] 01:44, 13 August 2011 (UTC)
::: Without dithering, I think the second palette looks the best. Kinda funny since that one probably required least amount of work. --[[User:Ledrug|Ledrug]] 05:27, 13 August 2011 (UTC)
:::: That is the case surprisingly often. Anyways, I have removed the alternate implementations. If anyone cares, they can find them in the [http://rosettacode.org/mw/index.php?title=Color_quantization&oldid=117230#J edit history for this page] --[[User:Rdm|Rdm]] 11:44, 14 August 2011 (UTC)
| |||