[Tool] ChromIQ – a macOS and Windows GUI for ArgyllCMS printer profiling (v4.0.0)

[Im-back-Arno]

A message from Knut:

Could you please remind Arno and the users that there are 130 presets with ready made patch sets ready for a user to use, so they do not need to create their own, or can use them as starting point for editing the patches using the patch editor… Also, they would start by reading the landing page

https://itsab1989.github.io/ChromIQ/

And read the first help card “Getting started (a tour of ChromIQ)”


So now you know where to look at

Thanks and sorry lol. I had 3 days in between travels and unpacked the i1pro2, plugged it in, pressed make patches, scanned patches, pressed make profile and printed a Colorchecker24 with that and then a real picture.

I do realize that I’m only touching the tip of Pandora’s box here, but it’s deeply impressive that it’s so intuitive that I managed to get a reasonably decent profile without any studies. I looked at Argyll and also QuadtoneRIP for B&W before and boy did I get lost (and I’m not easily scared or giving up quickly). To me, it shows how well ChromIQ is done and the UI.

Thanks Knut for the instructions and heads-up. Will certainly play with that when my travel load quiets down. For now, ChromIQ seems to be the best thing since sliced bread for profiling.

 

[nertog]

@Basti & Co.: thanks a lot for all the effort. This really is a nice piece of software!!

I do run into some issues when using my CR30: when measuring the most saturated patches on my chart, I get the message below. The error is reproducable and happens only when measuring on glossy or satin papers. The same patches printed on matte paper measure just fine.

I believe this is not a real error, and that the 0-bands might just be coming from a very saturated colorimetric value.

Currently using the latest beta + Canon IP8770 with oem inks.

 

[itsab1989]

@Basti & Co.: thanks a lot for all the effort. This really is a nice piece of software!!

I do run into some issues when using my CR30: when measuring the most saturated patches on my chart, I get the message below. The error is reproducable and happens only when measuring on glossy or satin papers. The same patches printed on matte paper measure just fine.

I believe this is not a real error, and that the 0-bands might just be coming from a very saturated colorimetric value.

Currently using the latest beta + Canon IP8770 with oem inks.


View attachment 19248

Hi, I put this into Claude and here is its reply. Reading it myself it sounds optimistic. I try to release the next beta as soon as possible and hopefully with a fix to this. I just came home from my nightshift though and am a little tired. But although I am sorry for causing you trouble I am glad you found this out and even more that you have reported it with this amount of detail

Hi nertog — thank you, and **your hunch was exactly right**. That is our bug, not
your instrument and not your paper. The screenshot was the whole clue: nice one.

Here is what was happening.

Your CR30 measures 31 bands of light from 400 to 700 nm. When a sample reflects
less light than the instrument's own zero point, the firmware doesn't return a
negative number — it clamps the value to exactly 0.0. That is normal and correct
behaviour; we have measured it on our own unit.

Now the paper. A saturated ink on **glossy or satin** paper sits on the surface
and goes very dark indeed — around 0.2–0.4 % reflectance in the band the ink
absorbs. The same ink on **matte** soaks in and scatters, so it only reaches
about 1.3–2.5 %. The CR30 has no black tile; it takes its dark reference from
open air, and that reference can sit a fraction of a percent high. On glossy
your most saturated patches therefore land at or below zero and get clamped; on
matte they never do. That is the paper dependence you spotted, and it is the
correct explanation for it.

And then ChromIQ got it wrong. We had a guard whose job is to throw away an
*incomplete* reply from the instrument — a real hazard, the device sometimes
answers before it has finished and hands back a half-written buffer. The guard
asked the wrong question. It said: *"three or more bands in a row are exactly
zero, so this reply must be truncated."* Your patch C16 had exactly three. It
was a perfectly good measurement of a very saturated colour, and we refused it —
six times, and then gave up on the patch, which means that chart could never be
finished. Your "0-bands from a very saturated colorimetric value" is precisely
it.

What made this a bad guard is that the number was never right. Every genuinely
truncated reply we have on record has a run of 5, 16 or 31 zero bands — nothing
has ever come in at 3. Three was a guess, and it sat inside the range that real
ink produces.

**The fix, in the next beta:** the guard no longer counts zeros at all. It asks
whether the *reply* is complete, which we can now establish exactly rather than
by a threshold — a truncated reply loses the colour value that sits after the
spectrum in the packet, so it gives itself away without us having to guess where
"too dark to be real" begins. Genuine truncated replies are still caught; your
saturated patches now read first time, however deep they go. We reproduced your
exact error message from a synthetic saturated reading, watched it disappear
under the fix, and it is pinned by tests so it cannot come back.

Two small things worth knowing:

- **Nothing you measured was ever at risk.** The window was right about that
part — a refused reading is never written to your measurement file, and every
patch you had already read was already safely on disk.
- When a patch does clamp, those bands are the instrument's floor rather than a
true measurement, so the profile is a touch optimistic in that ink's very
darkest corner. That is a limit of any colorimeter against a very dense glossy
patch, not something the fix changes — but ChromIQ now records it in its log
instead of refusing the patch.

Thanks again for the report, and especially for the screenshot and for saying
what you thought was going on. That is the difference between "a user has an
error" and a fix on the same day.

 

[itsab1989]

Thanks and sorry lol. I had 3 days in between travels and unpacked the i1pro2, plugged it in, pressed make patches, scanned patches, pressed make profile and printed a Colorchecker24 with that and then a real picture.

I do realize that I’m only touching the tip of Pandora’s box here, but it’s deeply impressive that it’s so intuitive that I managed to get a reasonably decent profile without any studies. I looked at Argyll and also QuadtoneRIP for B&W before and boy did I get lost (and I’m not easily scared or giving up quickly). To me, it shows how well ChromIQ is done and the UI.

Thanks Knut for the instructions and heads-up. Will certainly play with that when my travel load quiets down. For now, ChromIQ seems to be the best thing since sliced bread for profiling.

But this makes me happy to hear because it sounds like although you were in a hurry you managed to get a profile out of this