Possibility to convert csv-file with only L*A*B* values into a ArgyllCMS ti3 file

[ConradH]

Perhaps I should start a new thread, but I wanted to put up some test results on my CM30. It's obviously a CM30 by color and function but nowhere does that label appear. It's a $79 Veykolor Pro from Amazon- I didn't want to mess with overseas sellers.

Know several things. I'm an idiot. A fairly sophisticated one, but I make mistakes and often slide a foot back for every 13" I move forward. I have a mash-up of software and have not worked through to actually profile with the CM30, just comparisons to get a feel for the thing.

OK, so I have an ancient eye-one UVcut and ProfileMaker 5.0. It does have the matching cal base. I also have the Babel programs and the PC app for the CM30. I was fortunate enough to be gifted a set of Labsphere reflectivity targets by a friend who left the color business, though the certs are long gone. They should still be decent.

The first thing that seemed interesting was to compare the CM30 and the eye-one using paper with more and less optical brightening agents. Using my Harbor Freight UV flashlight and a Red River sample pack, I found two likely candidates, UltraPro Gloss (very white) and Premium Gloss C2S (ever so slightly yellow). This image is from the eye-one and since it's UV filtered the difference between the papers isn't huge, though far from zero.



The CM30 software only shows one graph at a time, so here are two images.




You can see the drop at the short wavelength end is substantial with less optical brighteners.
That tells me that to compare the two spectrophotometers I need to use a spectrally flat target, like the Labsphere ones, with no brighteners. (That's not got much Spam in it. What do you mean, you don't want any Spam!) I used the 50% reflectivity one, which I think should have an L of 76.



I don't know why the eye-one varies as much as it does at the 400 end. Here's the CM30.



To save your eyesight, the Lab values are 75.57, 0.17 and 0.66 for the eye-one. Then 74.28, 0.15 and 0.59 for the CM30. Both were set to D50/2deg. I think that's a delta E of less than 1.

Hopefully this was of at least some interest to somebody and certainly do point out whatever blunders I've made.

CH

 

[Epatcola]

I see ConradH asking about CR30 communications on DPReview. I don't participate there.

FYI the CR30 USB interface is just a serial port implemented with a CH340 USB -> serial chip. While connected you will see the CR30 in windows device manager under serial ports as "USB-SERIAL CH340".

There are serial port monitors for windows that can watch communication on serial ports.

I have not yet tried a CR30 bluetooth connection. I would imagine that is also just a serial port using the bluetooth SPP profile.

Hopefully they didn't obfuscate the CR30 communication protocol so it will be possible to figure it out by watching ColorQC 2 talk to it.

On a different subject I am thinking it would not be hard to make a jig to attach a CR30 to the head of a 3D printer and not be hard to write g-code to read targets with it. I envisage using the Z axis to lift the CR30 during movement and to press the button to make a measurement.

 

[ConradH]

The Bluetooth connection is super convenient so I'm starting to look at that with Wireshark. I see the data but I'm still too green to filter it properly and figure out how to talk to it in another program. I've done some serial before and that would be a smarter place to start, but once you get used to wireless...

Great minds think alike! The CR30 is small enough that it could mount on any 3D platform, like a router table or other thing. It would also be pretty easy to build a SCARA style robot with a couple stepper motors, like the very expensive X-Rite scanning table, only smaller.

The thing that got me going on all this was I wanted to measure an IT8 target. I was doing profiles using the DDI Profile Prism program and a scanner, which works very well, but uses an IT8 target as a comparison reference with the printer target. The good Wolf Faust IT8 targets aren't available anymore, so I bought one from the greywhitebalancecolorcard company in the UK. It worked poorly and I suspected that the generic data file wasn't a good enough match. IT8 targets have 6mm square patches, so I bought the used eye-one to re-measure it. Of course, I'm so easily distracted that I've made profiles with that and have never re-measured the IT8 target. Still need to do that just to satisfy my curiosity.

One thought that has crossed my mind is how long the CR30 will be available at a reasonable price. It seems too good to be true and with all the tariff nonsense, it's hard to predict the future and I'd be a bit miffed if I write some utilities that are instantly obsolete.

 

[pharmacist]

On a different subject I am thinking it would not be hard to make a jig to attach a CR30 to the head of a 3D printer and not be hard to write g-code to read targets with it. I envisage using the Z axis to lift the CR30 during movement and to press the button to make a measurement.

The reading can be done too by sending a pulse to the CR30: the ColorQC2 software can perfectly read a patch by pressing on the button "+Test Target", so it should be possible to drive the CR30 with the right software/driver without pressing physically on the CR30 top measuring button.

 

[Epatcola]

The reading can be done too by sending a pulse to the CR30:

That would be much more complicated.

I envisage a jig like this

The CR30 is a loose fit in the jig main part. The collar around the CR30 is fixed. The jig lifts the CR30 by the collar, moves it to the next patch, lowers it to the target then drops some more to briefly press the CR30 button with the top of the jig.

A g-code program to do that for all patches in a target would not be hard to generate from a ti2 file and patch size. Just run that and let ColorQC 2 collect the measurements.

I have an obsolete cheap ender 3 clone I would repurpose if I had a lot of profiles to create.

to be true and with all the tariff nonsense,

You should be happy doing a bit to help reduce the USA's $27 trillion national debt