krita cvs with 16 bit individual color channel adjustment CMYK, etc.?

[Boudewijn Rempt]

On Monday 02 January 2006 17:33, Larry Marso wrote:
> On Mon, 2 Jan 2006 2:48 am, Boudewijn Rempt wrote:
> > I'm not promising we'll be able to get it right, though :-).
>
> I haven't seen glasgow, but cinepaint hit the gtk1 wall long ago.
>
> I can provide some technical info on the "right" way to present the UI
> for lab color curves.  I figure the first step is what you're doing
> right now, making it work for alternative color spaces based on the
> profile.  Cinepaint did that and stopped.  And it's fine -- something
> you can adapt to if you have to.  But L A and B curves are generally
> presented upside down in the user interface. And the A B curves are on a
> value space from -128 to 128 in an 8 bit color space, and -65536/2 to
> 63335/2 in 16 bit.  I don't recall offhand whether the L channel does
> the same.

Could you join us on the kimageshop mailing list? Casper and Bart would very 
interested in discussing this further, I'm sure. I guess we could use the 
same zooming trick we use for 16 bit/channel histograms for the A and B 
curves.

> I don't have a stable krita at my fingertips.

Neither have I :-). Casper confirms that our nice per-channel color adjustment 
plugin disappeared about a week ago, but no doubt it'll resurface now that 
Casper has returned from London.

< ...about the area averaged color-under-mouse measurements in real time... >

> But I'm describing a 
> feature that would, in some position or window, give you real time L A B
> channel values for whatever is under the mouse.  And, importantly, you
> should have flexibility to define a radius over which you can see
> averaged values calculated in real time.  I believe the way photoshop
> does it is to present simultaneously the single pixel values and an
> averaged-over radius value.  Moving the mouse over portions of a picture
> where you intend to make modifications in L A B color (an outrageously
> powerful method), you get a clear sense of what portions of the curve
> you should manipulate.

Sounds to me like that's quite easy to do, but maybe I'm missing something. I 
bet we can even do that for bell-pepper shaped regions :-)

> By the way, modifications via curves on the A B channels is often the
> most important in the 10-20% of the curve around the center value.  To
> make krita stupendously useful, it should enable you to resize the curve
> widget so it's very large, enabling microsurgery on these curves.
>
> Thank you for this dialogue.  You're a great advocate for krita.  You've
> been at the top of all my impportant google searches on this topic, at
> fading memories.

Useful input is always very welcome. I hope you'll be able to help us further!

-- 
Boudewijn Rempt 
http://www.valdyas.org/fading/index.cgi
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
Url : http://mail.kde.org/pipermail/kimageshop/attachments/20060102/2d2f8e11/attachment.pgp