Topic: Light cyan banding on print margins (Read 6363 times)

mshea · « **on:** April 06, 2011, 11:27:59 pm »

I seem to remember a posting about this problem some time ago, but I can't locate it. I'm getting light cyan bands in the margins of my prints, sometimes on opposite margins, sometimes on all four margins. However, these only occur on Hahnemuhle Photo Rag Ultra Smooth paper, for which I've made a profile using Colormunki. I see no banding when I use the Hahnemuhle supplied profile. The trouble is, I get a slight green caste in my B&W prints when using the canned profile.

Epson 3880
Macbook Pro, OS 10.6.7
CS5

What gives? Any thoughts?

Merrill

Schewe · « **Reply #1 on:** April 06, 2011, 11:39:18 pm »

Are the profiles version 2 or version 4 ICC profiles? (hint, you don't want to use v4 a this point because of the issues you are seeing).

stefano · « **Reply #2 on:** April 07, 2011, 05:53:21 am »

Merril,
I too have been experiencing this issue. It is due to a known issue between Munki generated v4 profiles and Mac os X. The solution is to reprofile your paper after setting the Munki software to create v2 printer profiles. You can do so in the preferences tab. It defaults to v4 for both screen and printer profiles, but once you change it it will remember the new setting.
If you want to know more, search for 'scum dot' in the forum or online.

Stefano

mshea · « **Reply #3 on:** April 07, 2011, 06:24:25 pm »

Thanks for your input, fellas. I reset the preferences to V2 and reprofiled the paper, but sorry to say, the scum dots are still there. Now, I didn't reprint the target charts; just rescanned from the original charts. Would reprinting them make any difference?

Merrill

stefano · « **Reply #4 on:** April 07, 2011, 06:34:59 pm »

Merril,

I would not expect that reprinting the targets would make a difference, even if I did reprint as I had not kept the old ones around. Make sure you set teh v2 preference for your printer profile - the Colormunki software preference screen has two sections - the top one is for display profiling and the bottom one is for printer profiling, and the profile version is set independently for either.

It did work for me, and I can reliably go from cyan border (v4 profile) to no border (v2 profile). Can you open the profile with colorsync and see what version it claims, just to make sure?

Like you I use mac os 10.6.7 and Photoshop CS5, printing via the latest Epson driver.

Stefano

mshea · « **Reply #5 on:** April 08, 2011, 11:30:33 am »

Stefano,

Yes, I made sure to check V2 for the printer profile.

Guess I'll have to contact Colormunki. I hope their customer service responds; I'm not encouraged by the reports I've read.

Thanks again,
Merrill

stefano · « **Reply #6 on:** April 08, 2011, 02:16:22 pm »

Best of luck in getting it resolved, I am glad the v2 fix worked for me. Our setups are pretty similar, so I am curious to know what turns out to be the fix for you, in case the dreaded scum dots reappear at the next driver or OS update...

Stefano

elolaugesen · « **Reply #7 on:** April 10, 2011, 02:07:57 am »

see discussions here... lots more at apple etc.... I develped a temporary bypass as you will see and then finally resolved the problem by spending some money on a Spyder3 Print profiling package. it was obviously an xrite software issue although theydid not wish to admit it.... I am up and running with no problems...

http://www.luminous-landscape.com/forum/index.php?topic=52185.0
http://forums.adobe.com/message/3172530#3172530

JimGoshorn · « **Reply #8 on:** April 10, 2011, 12:19:30 pm »

Does anyone know if i1Profiler has any issues with generating v4 profiles?

Jim

elolaugesen · « **Reply #9 on:** April 10, 2011, 12:49:29 pm »

my problems with the cyan light band etc. happened with both 2.4 and 4.0 profiles generated by xrite profiler. Lightroom no problem, Aperture no problem because of the way they handle paper margins.... etc... not one problem with colorvision/datacolor profiles ...

cheers elo

digitaldog · « **Reply #10 on:** April 10, 2011, 01:52:22 pm »

Quote from: JimGoshorn on April 10, 2011, 12:19:30 pm

Does anyone know if i1Profiler has any issues with generating v4 profiles?

No scum dot. But no reason to build V4 profiles, they provide nothing over V2 since PRMG isn’t supported.

rpuller · « **Reply #11 on:** April 12, 2011, 02:29:54 pm »

I have contacted ColorMunki support and they state that this is a well known problem involving only Adobe and Epson.

I have been following this issue for quite a while (at least since CS5 was initially released) and Mark Dubovoy, Eric Chan and other "experts" (that I believe to be knowledgeable) have variously attributed the fault to some combination of Adobe, Apple, ColorMunki and Epson.

Various workarounds, including use of v2 profiles instead of v4, have been proposed, but I think the fact that v4 profiles were developed suggest that, at least at some point in the future, we could benefit from having the ability to use them. To that end, I'd like to see this issue resolved.

1. Is there any factual information (as opposed to opinion) as to which suppliers are at fault here?

2. Can some of our own respondents, like Jeff Schewe and Andrew Rodney (whom I suspect to be influential as well as knowledgeable), apply any pressure to the suppliers to fix the issue? Maybe Luminous Landscape could assemble a petition, headed by those luminaries and backed by the rest of us, addressed to all potentially involved suppliers, to fix the problem.

Rob Puller

digitaldog · « **Reply #12 on:** April 12, 2011, 04:34:12 pm »

Quote from: rpuller on April 12, 2011, 02:29:54 pm

Various workarounds, including use of v2 profiles instead of v4, have been proposed, but I think the fact that v4 profiles were developed suggest that, at least at some point in the future, we could benefit from having the ability to use them.

That may never happen. Currently, at least in X-Rite products (I can’t speak for all profiling software), there is nothing V4 brings to the party. For all practical purposes, they are V2 profiles in sheep clothing. Just make V2 profiles.

rpuller · « **Reply #13 on:** April 12, 2011, 09:23:22 pm »

Quote from: digitaldog on April 12, 2011, 04:34:12 pm

That may never happen. Currently, at least in X-Rite products (I can’t speak for all profiling software), there is nothing V4 brings to the party. For all practical purposes, they are V2 profiles in sheep clothing. Just make V2 profiles.

Current status seems to indicate, as you say, "that may never happen". Why? Is it because there is insufficient pressure from influential and knowledgeable folk to cause an effect?

I can't refute, on a technical basis, your disregard for v4 profiles, but I believe that Jack Holm, Ingeborg Tastl, Ann McCarthy, Max Derhak, and others who are working on them are not just spinning their wheels. Don't accuse me of name dropping here; I have read some of their comments about PRMG, which you said in an earlier post is not supported by the i1 Profiler and I was wondering why that was so. Again, is it because there is insufficient pressure on i! to get it right?

If v4 profiles are just v2 profiles in sheep clothing, that would indicate that we are being short-changed by a part of the photographic community.

I have generated v4 profiles without the scum dot problem manifesting itself. I'm using OS X 10.6.7, CS5, an X-Rite ColorMunki and an Epson 3880 printer, and, no, I don't know why some are seeing the problem and I am not. I do think that I'm seeing better image quality with the v4 profile, but without some form of objective measurement, i'm willing to admit that my perception may be wishful thinking. It would be nice, however, and reasonable, to know that v4 profiling (with it's full intent intact) is available to me.

digitaldog · « **Reply #14 on:** April 12, 2011, 09:34:28 pm »

Quote from: rpuller on April 12, 2011, 09:23:22 pm

Current status seems to indicate, as you say, "that may never happen". Why?

Because the advance of color management moves at a snail’s pace. We waited 6+ years for X-Rite to bring a new level of pro profiling apps and it doesn’t support V4 PRMG. Now maybe the next version will. But I wouldn’t hold my breath.

Quote

I have generated v4 profiles without the scum dot problem manifesting itself.

Right, the scum dot is the result of buggy code (code that may be in the app making the profiles or the app accessing the profiles, I can’t say for sure). As you say, not all V4 profiles produce this result.

rpuller · « **Reply #15 on:** April 12, 2011, 10:24:16 pm »

Quote from: digitaldog on April 12, 2011, 09:34:28 pm

Right, the scum dot is the result of buggy code (code that may be in the app making the profiles or the app accessing the profiles, I can’t say for sure). As you say, not all V4 profiles produce this result.

That's a good comment regarding which app has buggy code. I think it implies that it's Apple's, Epson's or ColorMunki's app rather than Adobe's or Epson's as has been suggested. Those that have picked Adobe and Epson seem pretty convinced, though. Personally, I see the Adobe tie-in as more tenuous, yet they have come out with Adobe Color Printer Profile, presumably because they acknowledge some culpability.

I've had several positive exchanges with ColorMunki support and I have discussed some color management issues on the Nikonians website. As you can see, I'm trying to push this issue a bit. May I have your permission to copy your responses (only the ones to me) to ColorMunki and to post them on Nikonians?

Rob Puller

Farmer · « **Reply #16 on:** April 13, 2011, 12:31:17 am »

Quote from: rpuller on April 12, 2011, 10:24:16 pm

That's a good comment regarding which app has buggy code. I think it implies that it's Apple's, Epson's or ColorMunki's app rather than Adobe's or Epson's as has been suggested. Those that have picked Adobe and Epson seem pretty convinced, though. Personally, I see the Adobe tie-in as more tenuous, yet they have come out with Adobe Color Printer Profile, presumably because they acknowledge some culpability.

Couple of things:

Adobe released Color Printer Utility (not Profile) because it was too hard to keep updating the no-profile path for target making in Photoshop. By breaking it out to its own app, they can more easily update and manage it. Absolutely nothing to do with this issue.

Which Epson app would you be thinking is involved in this? The driver? The driver works - it prints what's rasterised and sent to it. Again, I can't see how it is related to this in any way.

rpuller · « **Reply #17 on:** April 13, 2011, 08:09:45 am »

Quote from: Farmer on April 13, 2011, 12:31:17 am

Adobe released Color Printer Utility (not Profile) because it was too hard to keep updating the no-profile path for target making in Photoshop. By breaking it out to its own app, they can more easily update and manage it. Absolutely nothing to do with this issue.

Which Epson app would you be thinking is involved in this? The driver? The driver works - it prints what's rasterised and sent to it. Again, I can't see how it is related to this in any way.

As I said, I think the Adobe connection is unlikely, especially since CS5 is not even invoked when producing profiles using ColorMunki; however, it is difficult to feel confident given the finger-pointing by many laymen (include many of the so-called experts) and the lack of confirmation by the suspects, admittedly identified as such by laymen.

The Epson driver doesn't just print what's sent to it. It analyzes the data and resamples as necessary to match the mechanical requirements of the printer, so we know that some computation is involved. What we don't know is how limited, or extensive, those computations are. Do they include response to instructions imbedded in the v2 or v4 profile; I think it's likely.

Rob Puller

Farmer · « **Reply #18 on:** April 13, 2011, 07:01:26 pm »

Quote from: rpuller on April 13, 2011, 08:09:45 am

As I said, I think the Adobe connection is unlikely, especially since CS5 is not even invoked when producing profiles using ColorMunki; however, it is difficult to feel confident given the finger-pointing by many laymen (include many of the so-called experts) and the lack of confirmation by the suspects, admittedly identified as such by laymen.

The Epson driver doesn't just print what's sent to it. It analyzes the data and resamples as necessary to match the mechanical requirements of the printer, so we know that some computation is involved. What we don't know is how limited, or extensive, those computations are. Do they include response to instructions imbedded in the v2 or v4 profile; I think it's likely.

The Adobe connection is not unlikely, it's just "not". The Epson driver deals with rasterised image data. The profile isn't seen by the driver. It makes assumptions about the colour space if you set it to anything other than colour management off, but that's about it.

Schewe · « **Reply #19 on:** April 13, 2011, 07:11:35 pm »

Quote from: digitaldog on April 12, 2011, 09:34:28 pm

Right, the scum dot is the result of buggy code (code that may be in the app making the profiles or the app accessing the profiles, I can’t say for sure). As you say, not all V4 profiles produce this result.

Oh, it's not hard to figure out the problem. Bottom line, it's ColorSync and the way some profiles (V4 is more likely to break) interact with some printers in certain print pipelines. Note, the problem does NOT appear anywhere on Windows...it's a Mac only issue. So...it's pretty clear that the root cause is something ColorSync is doing when using certain V4 ICC profiles in certain print drivers. Can either the color management software or the printing applications work around the issue? Prolly...but at this point, there's simply no compelling reason to use V4 profiles at this point in time on Mac (and several good reasons not to do so).

Author Topic: Light cyan banding on print margins (Read 6363 times)