CHROMiX

Absolute Movement

When I plot an image as vectors in ColorThink Pro 3.0.1b5 using the Absolute intent, I’m seeing considerable movement, such as delta-Es of over 3, for in-gamut colours. When I plot this same image as vectors (using the same profile) in ColorThink 2.1.2 using the Absolute intent, I see very little movement–i.e. lots of Green (delta-Es less than 1). This happens with every profile and image I try. ColorThink Pro seems to be saying there will be major colour shifting of in-gamuts colours, while ColorThink 2 says there will be little to no shift.

Has anyone experienced this? Is this normal? If it’s the absolute intent, this shouldn’t be normal. ColorThink Pro seems to be giving erroneous vector plots for the Absolute intent. Relative and Saturation seem to look pretty similar between Pro and v2 (and of course Perceptual doesn’t work).

Can I trust ColorThink Pro?

I converted the image in Photoshop using Absolute and brought the converted file into ColorThink Pro. This plot contradicts the vector plot. In other words, the colour points of the converted file are only slightly dislocated from the original image’s colour points, unlike the vector plot which predicts a radical change in position.

Seems like the Absolute intent is remapping the white point, like Relative, within ColorThink Pro.

I know (I think) that this is not the case with your problem, but I figured I’d mention it here in case some readers are just starting out…

When I first started plotting images in ColorThink (2.x) I ran into some strange problems. It turned out it was just because it was resizing my image to 100x100 (or no bigger than 100 in the long direction I guess) prior to plotting the colors. Does it still do this with ColorThink Pro? Is it still 100x100 max?

As a side note, Absolute is supposed to not remap for the different white points? Or are different white points the only time that Absolute does remap in-gamut colors? I know I just read about this somewhere else but I can’t remember because I haven’t had to care about Absolute much!

devlinr wrote:

When I plot an image as vectors in ColorThink Pro 3.0.1b5 using the Absolute
intent, I’m seeing considerable movement

The same is true for me. I just tried it with a 10x34 pixel Lab tiff in CTP 3.0.1.b6 selecting the Euroscale Coated v2 profile.
Btw, my out-of-the-box 3.0 release doesn’t seem to suffer from this problem.

Is this normal?

I don’t think so.

Can I trust ColorThink Pro?

Good question.

Giordano

hussey wrote:

Absolute is supposed to not remap for the different white points?

As I understand it, the only adaptation AbsCol carries out is the clipping of out of gamut colours.

Giordano

I tested this, and you’re right. But also, all the rendering intents plotted differently.

Which version is correct? Now I REALLY don’t know whether to trust ColorThink. Is there someone at Chromix that can shed light on this subject?

I wrote:

As I understand it, the only adaptation AbsCol carries
out is the clipping of out of gamut colours

I must eat my own words: as I’ve since learned, when the profiles don’t share the same PCS white point an adaptive transformation will take place. My apologies to the forum.

That said, as I understand it (oh my! here I come again :slight_smile: this transformation won’t change any colours; it will just substitute the destination PCS’s tristimulus values for the source’s ones. So I think that no point should move in our plots due to it.

Giordano / waiting for refutation

As far as I know, you are right. With the absolute colorimetric intent, there should be no or little movement within the reproducible gamut. So ColorThink 2.2’s rendering should be correct, not the ColorThink Pro’s. I would like to see what Chromix explains about the issue.