i1 UV Cut help - black sublimation very off

[MM1287]

I've been lurking around these forums for a long time now just trying to learn from everyone. I primarily do sublimation work on aluminum plates and fabric and decided to try creating my own profile for my inks. I recently purchased an older i1 pro UV Cut (when it was still under GretagMacbeth) and attempted to create a profile with the Match 3 software which was pretty simple. However, my black is terrible when measuring the data off of an aluminum plate. It's looking brown when using the profile that I created. The original patches with no color management had a better black than the profile I created. The other colors seem fine. I attempted to do this with fabric and it was much better...although not perfect. I'd say about 90% there though.

I looked into whether the UV Cut would be a problem, and from what I read, it should do very well with glossy material like the plates that I used.  Here's what I did:

1. Printed the chart from Match 3 using my Epson printer with color management off, Plain Paper, Quality setting.
2. Sublimated the chart onto the aluminum plate
3. Measured the aluminum plate on the next step with Match 3 and saved the profile
4. Printed some test pages with Photoshop with the same settings as #1 above, but with the profile chosen, black point compensation checked, and Perceptual as the Rendering Intent. My color space was Adobe RGB (1998).

Can anybody help me troubleshoot through this?

[GWGill]

I primarily do sublimation work on aluminum plates and fabric and decided to try creating my own profile for my inks.
...
Can anybody help me troubleshoot through this?

I don't have any direct experience in profiling aluminium, but I'd suspect you would have problems that are nothing to do with UV/not UV.
UV cut only has an effect when there is fluorescence involved, typically the media due to the addition of FWA/OBA to paper. So unless you are using fluorescent ink, that wouldn't seem to be an issue for you.

A much more likely cause of problems is the reflective nature of your media (aluminium). Graphic arts instruments like the i1 pro use a 0/45 or 45/0 measurement geometry, where (in the i1's case) the light is shone on the sample at 45 degrees, and the color measured at 0 degrees to the surface. This completely excludes any specular light. I would expect that when a real person looks at the surface, specular effects play a part in its appearance, so the i1 isn't capturing what you see, and instead is "seeing" the colors as darker and more saturated than it actually appears, resulting in a profile with a "too light" black. You may need a much more specialized instrument to get reasonable measurements - i.e. a d/8 sphere instrument may be more suitable.

See http://www.xrite.com/documents/apps/public/whitepapers/Ca00015a.pdf  and http://www.packageprinting.com/article/color-measurement-metalized-substrates-labels-packaging-materials/1 for more information on this topic.

[MM1287]

That's pretty disappointing. I clearly misunderstood the "UV Cut" part of this tool.

A while back, I rented the ColorMunki and create a profile with it on fabric. That same profile was used on the aluminum plates with excellent results (the black was 95% accurate in my opinion). However, with this much more expensive i1 pro, the black is more of a dark green on the aluminum plate. I'm not sure if that's an issue with my i1 or what but it's been frustrating.

[GWGill]

A while back, I rented the ColorMunki and create a profile with it on fabric. That same profile was used on the aluminum plates with excellent results (the black was 95% accurate in my opinion). However, with this much more expensive i1 pro, the black is more of a dark green on the aluminum plate. I'm not sure if that's an issue with my i1 or what but it's been frustrating.

Hmm. Hard to know why that would be, since the ColorMunki is also a 45/0 type instrument. Does the Aluminium have some texture ? - If so, you may just have been lucky. One of the differences between the ColorMunki and i1pro is that the former has a light source that illuminates from a single direction, whereas the latter illuminates from all directions. The latter is generally better in evening out any slight texture or polarizing effects. But in your particular case, maybe you had the ColorMunki oriented in such a way that it's single illuminant direction gave you a more useful reading. 

[Stefan Ohlsson]

Hmm. Hard to know why that would be, since the ColorMunki is also a 45/0 type instrument. Does the Aluminium have some texture ? - If so, you may just have been lucky. One of the differences between the ColorMunki and i1pro is that the former has a light source that illuminates from a single direction, whereas the latter illuminates from all directions. The latter is generally better in evening out any slight texture or polarizing effects. But in your particular case, maybe you had the ColorMunki oriented in such a way that it's single illuminant direction gave you a more useful reading. 

I had a problem creating a good profile with a glossy canvas. The blacks were too light, and I thought that a instrument with a polarizing filter would help. But is it better to use a sphere spectro, and is there such a thing that could be used to create a profile?

[digitaldog]

Measure a single black patch on this material. What's the Lab values?

[MM1287]

Forgive me for my ignorance but is that the number under "Lab_L" for the target file printed with no color management? I'm a complete beginner with this stuff but I used ProfileMaker's MeasureTool to measure it. I simply follow directions as I did with the ColurMunki expecting a similar result.

I used the TC9.18 RGB target and all I really see for the pure black patch is -0.33 under "Lab_L" on patch A1. If you'd like to look at the measurement file, it's attached in this post.

Edit: I think I figured it out. The Lab values for the profile I created using a pure black patch is 10, 14.6, 10.6

[digitaldog]

1   A1      0.0   0.0   0.0   26.62   2.61   -0.28   
This is what ColorThink reports from RGB 0/0/0, an Lstar of 26.6 which doesn't sound at all kosher.

[MM1287]

Please dumb down that post for me....I have absolutely no clue what that means.

I'm not exactly sure my method of finding the Lab value was accurate. What I did was open up ProfileEditor, chose "One ICC Profile" changed the "Source or Destination Profile" to the one that I created, chose Absolute, and then opened up an image that I frequently print. I got a warning that the embeded profile for the image that I used was AdobeRGB which is different than the color profile that I created so I just clicked OK.

For some reason I got different Lab values on the black now. I got 26.6, 2.7, and -.3. The black on this image is showing up as dark green when I press it onto the fabric.

[digitaldog]

For some reason I got different Lab values on the black now. I got 26.6, 2.7, and -.3. The black on this image is showing up as dark green when I press it onto the fabric.

The 26 value is close enough and not what I'd expect to see by a long shot for measurement of black. Zero is the black hole. 100 white. For example, the Lstar I get for a profile on Luster paper is 6.84, that's more like it. 26? Not good.

The black on this image is showing up as dark green when I press it onto the fabric.

Not sure I understand your entire workflow here. But if you've printed something that gets transferred to something else (fabric?), you want to measure that fabric, not the intermediary process.

[MM1287]

Sorry if I'm causing any confusion. What I'm trying to say was earlier, I printed an image using the profile that I created with the i1. It was basically a black and white image. After pressing it, the black looked dark green on the fabric. When I created the profile, I did take the measurements using the fabric.

With ProfileEditor, there is a function to edit a color for the profile. I can simply choose the color that's giving me the problems and adjust the value of it for the profile. I unknowingly purchased this software with no dongle so I can't save the edited profile. Do you know of any other software that will allow me to do this?