ArgyllCMS 875-patch printer profile with i1Pro 2

[pkk]

Yes these are 924 patch targets. I will try your 875 target soon. I use Epson K3 ink except in the 3880 I have Cone carbon black in the matte black channel. But that is not a factor in either of these as they are both done with PK. Thanks again, Pharmacist! I'll be doing more of these next week.

 

[Ink stained Fingers]

I'M not an ArgyllCMS user but had a look to the finished profiles - yes, the gamut is pretty good overall - with OEM Epson paper and Epson inks, I'm comparing here just the -smatt type with a profile which I created longer time ago for my P7600 , discarded in the meantime. I was using other non-Epson inks , a mix of pigment inks actually - leftovers from other tests, and this on a -smatt paper by a 3rd party in Germany. The gain of a larger gamut - specifically in the a -b quadrant is visible , your 9880 printer comes with many more inks - OEM inks - supporting that - the gamut is wider and reaches down much more to darker colors. The yellow line as the border line of the displaed gamuts is for the Arg924 9880 profile. light blue is for a P7600.

The image shows a horizontal cut through the gamut body - at a luminance of L*=50 in this case, so far - so good, but there is some iregularity at the bottom side.


Here is an image of the cut at L*=13 - rather dark in this case - this just for the Arg924 semimatte profile.

This could be an issue with the scanning process, you may do a re-scan and re-calculation of the profile. The other Arg924 semigloss 3880 profile is o.k. - as displayed here as an overlay in blue.

 

[pharmacist]

softproofing (choose relative colormetric and NO black point compensation) is essential to see of the profile data is scanned properly. Scanning errors will be visible in unexpected color changes in color and b&w gradients. I sometimes see unexpected stripes in a color gradient from white to 100% color and this is an indication of scanning errors.

 

[Ink stained Fingers]

Please do a rescan; @pharmacist is addressing scanning errors. Yes - they are there , they may be caused by some variations in manual scanning speed and/or pressure, and if you may get different results (with tolerances outside the equipment spec) it is a handling problem. I'm using the i1Io scanning table which avoids such errors - or repeats the same errors so they are not visible in number variations.
But there is one other reason rarely discussed and documented, I had a kind of similar problem long time ago and several discussions with XRite - glossy papers with a fine structure - like semiglossy - can cause such problems - papers with a pretty strong gloss, there is straylight from such paper into the sensor module. The spectros are built to a particular standaed which does not cover this special condition, it is as such not a spectro device failure but a paper characteristic causing this problem. There are rarely any examples documented on the Internet, but the problems were there and XRite brought a new spectro on the market - the I1Pro 3 Plus - the 'Plus' makes the difference - it is using polarized light to suppress such unwanted side reflections. You may read the product description of the Plus model here - in comparison to the regular I1Pro3

https://www.xrite.com/-/media/xrite.../l7-760-i1-brochure/l7-760_i1_brochure_en.pdf

Have a look to page 4 , you can read here

' ... The I1Pro3 .... works better on harder to profile materials such as lightly textured paper,...specialty photo papers ........ ' which just means that such papers cannot be profiled correctly with all previous spectro models . I remember that Konica Minolta with the Myiro Spectro is using polarized light as well.

 

[pkk]

I will rescan the semimatte targets. All of this discussion is good for me, gentlemen, as I have much to learn about making printer profiles. I want to learn to read them myself. One thing at a time though. Thanks again for the input.

 

[pharmacist]

Please do a rescan; @pharmacist is addressing scanning errors. Yes - they are there , they may be caused by some variations in manual scanning speed and/or pressure, and if you may get different results (with tolerances outside the equipment spec) it is a handling problem. I'm using the i1Io scanning table which avoids such errors - or repeats the same errors so they are not visible in number variations.
But there is one other reason rarely discussed and documented, I had a kind of similar problem long time ago and several discussions with XRite - glossy papers with a fine structure - like semiglossy - can cause such problems - papers with a pretty strong gloss, there is straylight from such paper into the sensor module. The spectros are built to a particular standaed which does not cover this special condition, it is as such not a spectro device failure but a paper characteristic causing this problem. There are rarely any examples documented on the Internet, but the problems were there and XRite brought a new spectro on the market - the I1Pro 3 Plus - the 'Plus' makes the difference - it is using polarized light to suppress such unwanted side reflections. You may read the product description of the Plus model here - in comparison to the regular I1Pro3

https://www.xrite.com/-/media/xrite.../l7-760-i1-brochure/l7-760_i1_brochure_en.pdf

Have a look to page 4 , you can read here

' ... The I1Pro3 .... works better on harder to profile materials such as lightly textured paper,...specialty photo papers ........ ' which just means that such papers cannot be profiled correctly with all previous spectro models . I remember that Konica Minolta with the Myiro Spectro is using polarized light as well.

Would it possible to put a polarizing filter into the i1Pro2. I know you can disassemble the scanning eye and there is glass filter inside the scanner.

 

[Ink stained Fingers]

The i1Pro1 came in two versions - a regular one and a UV-cut version - similar like the ColorMunki spectro. The UV-cut Pro1 has a small circular UV filter disk in front ot the sensor, such filter is supposed not to block visible light but only UV-light which is close to the visible light. The spectro calibration takes care of that and a flag in the spectro firmware indicates which UV -version it is.

Would it possible to put a polarizing filter into the i1Pro2.

A pol-filter takes quite a share of the visible light, and I'm afraid the self-calibaration process cannot take care of such modified spectro just with a pol filter in front of it. And there is another question if and how the pol-filter should be aligned to the diffractive element in the sensor.

Here is another publication by XRite

https://www.xrite.com/service-support/450045andannularringilluminationoptics

I'm copying a line of text from there:

'If an object has a specular component of reflection (glare), it will be excluded by 45°/0° illumination automatically. Why? Light that is reflected specularly will be reflected at the same angle as it strikes the object. With 45°/0° illumination optics, light that is reflected away at 45 degrees will not be "viewed" by the spectrophotometer.'

So far so good but if that reflecting surface is not flat like a normal mirror you have a different situation - just assume a glossy textile fabric - which is causing such measuring problems since the specular component of the reflected light goes into every direction .

The problem is not new, and XRite knows about that effect but XRite does not mention it in context of paper profiling at all indicating that there is a functional limit to the 'classical' spectrometers.

I had a semiglossy paper longer time ago which caused problems during profiling - irritating deformations of the gamut volume. It was a pretty glossy paper with a rather coarse structure - I didn't like the look of it anyway but you don't know how a paper looks until you buy it anyway.
I discussed it with the XRite support, and I sent them the scanned data, and they asked me after a longer discussion which paper type was causing the problem. The final answer was that I could not profile that paper correctly with the Pro1 or Pro2 spectros at that time and I should not use that paper.

I'm not implying that the Argyll 875 or 924 profiles by @pkk are victim of this effect but you should keep it in mind as a possibility . There is one more thing which can be done - creating the profiles with a lower patchcount of the targets - e.g. to start with 100 patches and go to 300 . The above relection effect does not occur at all colors of the target sheet but only at a rather small range of colors and lightness values, and changing the patchcount may just get you around it.

 

[pharmacist]

The i1Pro1 came in two versions - a regular one and a UV-cut version - similar like the ColorMunki spectro. The UV-cut Pro1 has a small circular UV filter disk in front ot the sensor, such filter is supposed not to block visible light but only UV-light which is close to the visible light. The spectro calibration takes care of that and a flag in the spectro firmware indicates which UV -version it is.




A pol-filter takes quite a share of the visible light, and I'm afraid the self-calibaration process cannot take care of such modified spectro just with a pol filter in front of it. And there is another question if and how the pol-filter should be aligned to the diffractive element in the sensor.

Here is another publication by XRite

https://www.xrite.com/service-support/450045andannularringilluminationoptics

I'm copying a line of text from there:

'If an object has a specular component of reflection (glare), it will be excluded by 45°/0° illumination automatically. Why? Light that is reflected specularly will be reflected at the same angle as it strikes the object. With 45°/0° illumination optics, light that is reflected away at 45 degrees will not be "viewed" by the spectrophotometer.'

So far so good but if that reflecting surface is not flat like a normal mirror you have a different situation - just assume a glossy textile fabric - which is causing such measuring problems since the specular component of the reflected light goes into every direction .

The problem is not new, and XRite knows about that effect but XRite does not mention it in context of paper profiling at all indicating that there is a functional limit to the 'classical' spectrometers.

I had a semiglossy paper longer time ago which caused problems during profiling - irritating deformations of the gamut volume. It was a pretty glossy paper with a rather coarse structure - I didn't like the look of it anyway but you don't know how a paper looks until you buy it anyway.
I discussed it with the XRite support, and I sent them the scanned data, and they asked me after a longer discussion which paper type was causing the problem. The final answer was that I could not profile that paper correctly with the Pro1 or Pro2 spectros at that time and I should not use that paper.

I'm not implying that the Argyll 875 or 924 profiles by @pkk are victim of this effect but you should keep it in mind as a possibility . There is one more thing which can be done - creating the profiles with a lower patchcount of the targets - e.g. to start with 100 patches and go to 300 . The above relection effect does not occur at all colors of the target sheet but only at a rather small range of colors and lightness values, and changing the patchcount may just get you around it.

Good to know. I can optimize my 480 patch target in such a way that more samples are taken from the same colour patch when strip reading. My 480 patch target already can produce very good profiles and good neutral BW-prints. This way the reflection problem maybe averaged more when the scanning eye moves over the colour patches and more samples are taken from the same colour patch. With the 875 and 924 patch targets the patches are rather small but feasible when the paper does not have that strange reflection problem described by you.

 

[Ink stained Fingers]

Yes, changing the patchcount can get you around the problem , but the first problem is that you need to identify it - and the number and type of tools is very limited for us as non-commercial users.
I was testing that problem-paper longer time ago and I changed the patchcount - I don't remember the details anymore - the problem section with strange indentations moved from one location of the gamut voume to another location so it was a pretty strong effect overall, and the 4th or 5th attempt with a different patchcount finally gave a 'clean' normally shaped gamut voume - it was the lowest patchcount of all, it could have been an averaging effect or jumping over the problem zones whatever. This all creates the fun you get with icc-profiling the papers yourself.
I got a similar paper some years later but just in a sample box of a supplier so it was not a loss overall. I got two of such problem papers over a long time of self-profiling - more than 20 years so I can assume that the vast majority of other users profiling their papers never will find such papers or are not in the position that they can even identify such problems with a profile so there is possibly a gray zone of unknown cases.

 

[pkk]

I'm paddling hard to keep up with you gentlemen so I'll have to ask: Is this problematic reflected light coming from the devise or is it ambient light?