Profile Assessment

[pharmacist]

Hi Emulator,

I just discovered that ArgyllCMS version 1.6.1 seems to push up magenta/reds too much. I did some tests with version 1.6.2 and the profiles are now identical too the ones created with the Colormunki for the reds/magenta. So please upgrade to version 1.6.2 to correct for this strange behaviour.

 

[Emulator]

Hi Emulator,

I just discovered that ArgyllCMS version 1.6.1 seems to push up magenta/reds too much. I did some tests with version 1.6.2 and the profiles are now identical too the ones created with the Colormunki for the reds/magenta. So please upgrade to version 1.6.2 to correct for this strange behaviour.

Hi pharmacist,

The Argyll CMS Sihl profile I am using is the one you generated some time ago, which I believe is 1.6.1. It does appear a bit strong on the reds, but not as unpredictable as the CM profiles taken as a whole and the Canon profiles which seem very low on saturation.

Like Rodbam and Roy, I have not setup Argyll CMS, we are a timid lot, and as "retirees" we have all the time to do it.

Trying to make judgements on profile quality is not easy and as many have said the final proof is the visual image.

 

[Emulator]

Incidentally, I see the latest version of Argyll is now 1.6.3

 

[Emulator]

@RogerB, I had forgotten how interesting your guidance on profiles and Gamutvision was. I noticed some headline images that Nifty's system put up and found this thread again.

We weren't all using Argyll then, so things have moved on a bit, but well worth reading through again for those interested.

 

[Smile]

The Gamutvision BW density assessment screen is very hard to read, I mean there is no usable zoom? And the scale is +20 -20?

The scale is should be like +5 -5 or at least +10 -10, yes you can zoom it in, but then whole length of scale is not viewable.

Yes the log scale is nice but I had viewed one profile shown as very good grayscale then when checked in colorthink pro the opposite is shown.

 

[Ink stained Fingers]

o.k., I'm not really sure, what the last test/picture is supposed to tell me, I guess most of the
colors of the color band are saturated and a major part of it outside the printer gamut.
This becomes visible with the rel. col, setting, the color saturation of the colors had to be
reduced visibly to squeeze them into the gamut of the selected paper. 'There is one effect which should not become visible - banding is an indication of some local weakness of the profile at the affected color/saturation range. An evaluation with Gamutvision may show that more clearly

 

[Roy Sletcher]

The Gamutvision BW density assessment screen is very hard to read, I mean there is no usable zoom? And the scale is +20 -20?

The scale is should be like +5 -5 or at least +10 -10, yes you can zoom it in, but then whole length of scale is not viewable.

Yes the log scale is nice but I had viewed one profile shown as very good grayscale then when checked in colorthink pro the opposite is shown.

Glad to find I am not the only one to find Gamutvision unreliable and hard to view.

I just can't bring myself to send them US$100 for a product that to me seems to be in serious need of fine tuning and a tech support option.

(or maybe they have - been a couple of years since I visited their site)

RS

 

[Smile]

Gamutvision is more like scientific tool, for certain areas that colorthink pro does not show.

To judge the printer profile accuracy a proper RGB certification patch sheet needs to be printed, measured, evaluated by measuring it. There are no software substitutes that can do this reliably imho.

 

[RogerB]

Gamutvision is more like scientific tool, for certain areas that colorthink pro does not show.

To judge the printer profile accuracy a proper RGB certification patch sheet needs to be printed, measured, evaluated by measuring it. There are no software substitutes that can do this reliably imho.

Yes, I agree. The ultimate test of accuracy involves the measurement of a printed target, but that takes quite a long time. A tool like Gamutvision (or Colorthink?) can check the internal consistency of a profile very quickly. It usses the profile to predict the printed colours and calculates the deviation from the actual image colour. If the initial measurements are accurate then the process quite reliable. I often use a 288-patch target for this, and Gamutvision gives a result like the one below.

For this particular profile I have printed and measured the same target - the method that you advocate. The measured results are not quite as good as the Gamutvision prediction, with a mean deltaE of 1.23 but I take that as a good result. (I already posted this information http://www.printerknowledge.com/thr...ling-on-your-computer.8570/page-13#post-69573)

The Gamutvision BW density assessment screen is very hard to read, I mean there is no usable zoom? And the scale is +20 -20?

The scale is should be like +5 -5 or at least +10 -10, yes you can zoom it in, but then whole length of scale is not viewable.

Yes the log scale is nice but I had viewed one profile shown as very good grayscale then when checked in colorthink pro the opposite is shown.

I'm not really sure what your problem is with the B&W density display. Gamutvision displays L* from 0-100 and automatically scales the axis for a* and b* to suit. In the best case the scale for a* and b* is -2 to +10, like the one below. (The parameters plotted are 10x a* and b*)

I should say I don't get commission from Imatest for Gamutvision sales. I use it because I find it's quick, easy and reliable and it displays the information in a way that I can easily understand. (and I can't justify paying $300 more for Colorthink Pro).

 

[Smile]

Gamutvision is free now, GPL license.

But you do need to install the v1.3.7 then on top of that v1.4 otherwise installer of version 1.4 throws me errors.