PhotoComix wrote:
@patdavid
i left a comment to your decompose script in the registry but i fear he went as lost (or better disjointed from the topic
)
I used often a similar decompose script and i agree with most of your selection of "useful" channels " that may contain good BW interpretation of colour image
except for what about CMYK
I very seldom find the only one you included ,K useful
While very often the C M and Y give surprisingly good results...
( from a artistic point of view )
if not directly as positive as negativesThe word negative and positive is not the most adapt here , let say that often look and would work as good BW negatives and so if inverted they give a interesting results
Gmic would allow to enlarge the number of "useful channels" because support much more image mode,
and basically in each image mode the channel corresponding to the ... Luminosity or Luminance or Lightness or Value or similar may be worth a look.
I didn't try much the Samj bw filter except to see that can give very nice results and much control but i don't know if uses a similar principle .
About filters i would like one that just allow to browse and confront the different channels and just select one or more to render.
..a bit as your script but with a preview
all RGB ,the (inverted) CYM of CYNK ,and for all others modes the correspondent of the Lightness channel ..a bit as your script but with a preview
I only just saw your question on the registry this morning, sorry about that!
I answered you there, but I'll answer here to be complete.
I think the script decomposes to everything that you want already?
I didn't include the CMY from the
CMYK decompose, because the resulting pure color channels were gross.
I did include the CMY channels from the
CMY decompose, because those channels are much "cleaner" in their output. (Unless something is very wrong with my GIMP install?)
For example, here is the C channel from CMY decompose:
Attachment:
CMY-C.jpg [ 125.13 KiB | Viewed 1747 times ]
Here is the C channel from the CMYK decompose:
Attachment:
CMYK-C.jpg [ 93.97 KiB | Viewed 1747 times ]
I can only guess that the removal of K from the C channel in the CMYK decompose has led to the nasty quantization artifacts I am seeing in the result?
Are you seeing something different if you decompose to CMYK?