CHROMiX

Viewing a Rendered Gamut

youtube.com/watch?v=sTxaHhMYSYE

In the above named video an analysis of the rendered gamuts versus the proofing gamuts of a few profiles takes place in ColorThink. In one case, certain colour areas of the rendered gamut of the profile “don’t quite make it out to the boundary of what the printer is capable of.” I recently profiled a Baryta paper in combination with a Canon Pro9000 Mark 2 printer for a client. I analysed the profile in Colorthink and found areas in the oranges and yellows of the rendered Gamut which did not make it out to the proofing gamut. I have a screen shot which I could upload, if you like. There is an easily noticeable margin between the rendered and proofing gamuts in these areas. A nozzle check done prior to printing the profiling charts showed that all nozzles were firing. So, I wonder what would cause this to happen?

To perhaps clarify the question, is this a profile issue or does it indicate a deficiency with the device? I see why it can be an issue, but I fail to understand how it is caused. Does anyone know?

In most cases, this is an issue with the software that created the profile. If the profile was created properly, then this discrepancy between what the printer is capable of, and what it gets sent when the profile tells it to print one of these outer boundary colors, has to be due to the profile. Its not surprising that any software will come up a little bit short when printing these saturated colors, but its always nice to check to see whats actually happening on your own profiles. And some older profile-building software did a horrible job with this. I have seen huge margins as you call them.

Keep in mind that you can plot your measurement data and it will match up closely with your proofing gamut - so you know your printer is capable of printing those colors (without color management.) But then compared to the printing / rendering direction of the profile its sometime surprising and disheartening to see how much will be left out. No matter what you do with that profile, you will never get those colors that are represented in that gap. You would need a new profile, or more likely new profiling software.

Another point worth taking into consideration are the saturation and smoothness controls in profiling apps like i1profiler and MonacoProfiler. Obviously, if the saturation control is set to a lower than standard setting, of course the resulting profile and prints will also have lower saturation than the exact same profile created with the controls at either their defaults or even with the saturation increased.

This might be worth looking into because if you’re able to increase the saturation setting, you might find those edge/border colours and target patches end up being located much closer to what you were initially expecting to see.

Thank you Pat and Aron for your replies. For the most part, profiles which I have made for our own Epson printers have not shown a great deal of this issue. In fact, to quote the video once again, our results have been “jumpin’ out of its skin, it’s so good.” But, as I mentioned, this particular profile was made for a client for their printer with dye based inks and their paper. They printed the charts. So, I can not attest to the exact accuracy there. But, they looked good. Otherwise, I used the same procedure that I use to make our own profiles. That’s why I wondered what effect, if any, the printer or inks or paper could have on the result.

The profiling software we use now is i1Profiler and while I did not increase or reduce the saturation level from the default centre spot, if I had increased the saturation to full, my understanding is that this would only have an effect with the Perceptual intent. But the profile evaluation I used as per the video takes place with the Abs Col intent. So, please correct me if I’m wrong, but would this have the effect in ColorThink’s profile evaluation that you suggest, Aaron?

Thanks again for the help with this.

I have seen that the gap or margin between printed gamut and rendered gamut will increase with measurement data that is not smooth, and tends to be irregular. So the dye sub process might very well be a contributor here.

The best thing to do if you’re wondering what a software’s options will do is to try it! The nice thing about this rendered gamut procedure is that you don’t have to wonder what will happen, you can try it and see the results. Report back here what you find.:wink:

Well, Pat, I did just that and the results are fascinating to me. First, for clarification, the profile is not for a dye sub printer. Instead, the inks used in the Canon Pro9000 Mark 2 printer are dye based as opposed to pigment inks. I have had no profiling experience with dye based inks before now. Hence, I was curious if they may have played a role in terms of the result I got with the profile analysis in ColorThink as described above.

So, I just made 3 new profiles from the same measurements which were used in the creation of the original profile that is at the heart of this subject. The first was with the saturation setting in i1Profiler set to full. The second was with the saturation setting back to default but with the smoothness setting up to full. The third was with both set to full.

I processed the profiles in ColorThink as per the" Viewing a Rendered Gamut" video. As I suspected, the profile with the saturation set to full showed no difference whatsoever. It appeared to be identical to the first profile which had been made with both settings at the default centre positions. But, I was viewing it in ColorThink with the RelCol intent and the saturation setting should show an effect only with the Perceptual Intent.

The second profile with the smoothness set to full showed a surprising difference. Not only was the “margin” that we had seen in the oranges and yellows between the rendered gamut and the proofing gamut gone, but the proofing gamut in the greens and aquas to cyans had now shrunk well inside the rendered gamut. The rest of the rendered gamut was right out at the skin of the proofing gamut for the most part. I did not expect that. You were quite right and I appreciate your suggestion to check it out.

The third profile with both the saturation and smoothness settings set to full showed exactly the same results as the second profile. Obviously, that is because the smoothness setting is all that is showing a difference here.

I would be happy to upload screenshots if I knew how. In the meanwhile, I thank you again for this suggestion which has been quite interesting. Now I just have to try to figure out what the heck is going on here. Any thoughts would be appreciated. Thanks again.

Exploration and discovery. That’s what ColorThink is all about.

Here’s a little info on posting images:
colorforums.com/faq.php?sid= … d6bc5f2#24

You basically need to use a link that goes to a different site where the image is stored. If you highlight that link when you’re editing the post, and click the “Img” button, the image should appear here.