Pages: [1] 2 3   Go Down

Author Topic: Why are dark blues purplish in test print  (Read 9721 times)

Andy_S

  • Newbie
  • *
  • Offline Offline
  • Posts: 14
Why are dark blues purplish in test print
« on: January 09, 2018, 12:42:19 pm »

I have a new monitor – NEC MultiSync PA302W - that I have calibrated using Spectaview II and the following parameters:

  White Point        D65
  Gamma             2.2
  Intensity           120 CD/M2
  Contrast  Ratio  300:1

I then printed Andrew Rodneys’ Test image on Ilford  Gold Fibre silk paper on my Epson P800  using the profile provided by Ilford.

The brightness of the printed matched the display and all the other colours are fine except for the dark blues which are purplish.

Wondering what I do to fix this.
Logged

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
Re: Why are dark blues purplish in test print
« Reply #1 on: January 09, 2018, 12:51:41 pm »

First make sure there are no nozzle clogs. Yellow is particularly hard to see with the P800 so you need to inspect it under a magnifying glass with bright light at an angle. Cyan, Light Cyan or Yellow nozzle clogs could contribute to this issue. Absent any clogs, it could be that the profile is a bit problematic. Try setting everything up for Printer Manages Color, run the Rodney page again and observe what differences it makes. If the Blue comes out looking more Blue, that would tend to indicate a profiling issue. Even though you are using Gold Fibre Silk, try using the Epson Legacy Baryta profile instead and see whether that improves it. Again, if so, it would point to the Ilford profile.
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

Andy_S

  • Newbie
  • *
  • Offline Offline
  • Posts: 14
Re: Why are dark blues purplish in test print
« Reply #2 on: January 09, 2018, 02:36:03 pm »

Thanks for your suggestions.  I will try them over the next few days.

Andy
Logged

Alan Goldhammer

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4344
    • A Goldhammer Photography
Re: Why are dark blues purplish in test print
« Reply #3 on: January 09, 2018, 04:44:50 pm »

I observed this exact same thing several years ago using the Ilford profile as a comparison to the one I made myself for an Epson 3880.  At the time the Ilford profile was quite poor in other color areas as well.  One would assume that the profile you are using is relatively new because of the new model of printer.  I would follow what Mark told you and that should solve the issue.
Logged

Andy_S

  • Newbie
  • *
  • Offline Offline
  • Posts: 14
Re: Why are dark blues purplish in test print
« Reply #4 on: January 10, 2018, 11:50:21 am »

Appears to be a profile problem.  Printed using the Epson Legacy Baryta profile and had the same problem with the blue.
When I printed with colour management off the blue was much closer to the image on the display.  So looks like I need a custom profile.  Any suggestions on where to get one made? 

Thanks for your help.

Andy


Logged

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
Re: Why are dark blues purplish in test print
« Reply #5 on: January 10, 2018, 02:10:50 pm »

Nope - based on what you are saying here it is most likely not a profiling problem. If there were something wrong with the Ilford profile alone, then the print would have come out well using the Legacy Baryta profile because there isn't much difference in the profiling for either of these papers (Epson Legacy Bryta or IGFS). However, if Printer Manages Color is giving you a good result - much better than BOTH those profiles, this indicates that there is something else amiss uniquely in the colour management route (that is, via application colour management) - other than the profiles - causing this issue. Perhaps a setting somewhere, but off hand I can't offer a reliable view of what it may be; could you post all your colour management settings made through the application (Ps, Lr?) for the print pipeline - perhaps we may discover something there.

(typo corrected)
« Last Edit: January 13, 2018, 08:48:08 am by Mark D Segal »
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

Andy_S

  • Newbie
  • *
  • Offline Offline
  • Posts: 14
Re: Why are dark blues purplish in test print
« Reply #6 on: January 10, 2018, 02:40:38 pm »

I am printing using LR Version 6.13 on an Epson P800.  Attached are screens shots of the LR and printer settings.

Andy
Logged

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
Re: Why are dark blues purplish in test print
« Reply #7 on: January 10, 2018, 02:59:41 pm »

I don't see anything in these settings that should be causing such a problem. As an aside I could recommend turning off Finest Detail and High Speed in order to optimize print quality of normal photographs, but those have nothing to do with this issue. However, if Blue looks Blue on your display and in the print using Printer Color Management, then there is some other application-based colour management issue at play, apart from these settings in the print pipeline. Not clear to me what it may be - yet. I'm wondering about the version of Lr you are using - I'm on the subscription plan so the numbering is different from the permanent license versions; I think yours is the latter. If your's dates from mid-2016, there was a problem with one of those versions that could be explaining this; if so, perhaps try another version of Lr and see whether it makes a difference.
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

Andy_S

  • Newbie
  • *
  • Offline Offline
  • Posts: 14
Re: Why are dark blues purplish in test print
« Reply #8 on: January 10, 2018, 05:17:02 pm »

Printed using LL Classic CC 7.1 and same problem.  Also printed on Epson Premium Photo Paper Glossy and same problem.  Appreciate your suggestions and help.  Not sure where to go from here.

Andy

Logged

Doug Gray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2197
Re: Why are dark blues purplish in test print
« Reply #9 on: January 10, 2018, 05:46:30 pm »

Printed using LL Classic CC 7.1 and same problem.  Also printed on Epson Premium Photo Paper Glossy and same problem.  Appreciate your suggestions and help.  Not sure where to go from here.

Andy

Just on the off chance there is a gamut issue, does the soft proof differ? Prints should match the soft proof more closely than just the image. If the blues are out of gamut there could be a hue shift going on but it should show up in the soft proof if the profile is correctly made.

Also, it might help if you could identify the specific file name and location of the blue shift in the image.
« Last Edit: January 10, 2018, 05:55:42 pm by Doug Gray »
Logged

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
Re: Why are dark blues purplish in test print
« Reply #10 on: January 10, 2018, 05:56:29 pm »

Just on the off chance there is a gamut issue, does the soft proof differ? Prints should match the soft proof more closely than just the image. If the blues are out of gamut there could be a hue shift going on but it should show up in the soft proof if the profile is correctly made.

Funny you should mention that - I was going to generate a post about colour space and gamut, but I refrained because I thought it unlikely to be the problem insofar as IGFS has a very large gamut volume in the SC-P800 printer and one trusts the profile to bring OOG colours to their nearest value on the gamut boundary (Relative Intent) or within (Perceptual Intent) - but one never knows, so that brings to mind, thanks to your contribution - one more idea. Try making the same print with the Rendering Intents you have NOT been using. So for example if you were printing in RELCOL, try PERCEPTUAL, try ABSOLUTE, and see if one of those works better. If so we would be back to a profiling issue in the sense of how the profiles are handling OOG colours. But I have to add that I've printed many blue skies in that printer model with that paper using both canned and custom profiles and not experienced this problem.
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

Wayne Fox

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4237
    • waynefox.com
Re: Why are dark blues purplish in test print
« Reply #11 on: January 10, 2018, 10:13:45 pm »

i guess my question from the original post is whether poster thinks the print is inaccurate, or if it just doesn't match. If printed correctly, the print should be accurate but purplish blues could easily be a weak profile.

It can be a little challenging to determine this. One possible way is obtain a print of the same file from a reputable source, which should offer a nearly perfect match to one printed. Another step might be to use a standard high quality paper that has been manufactured and profiled by the printer manufacturer (maybe in this case Epson Premium Luster would be a good candidate). while even manufacturer profiles might not be optimum, most printer makers to a pretty good job supplying decent profiles whereas 3rd party paper profiles are sometimes problematic. If there is a difference perhaps that would indicate the Ilford profile might not be optimum. Once confident the printed output is accurate, if the output doesn't match then adjusting the brightness and white point of the display is necessary until a match is achieved.  This all assumes an appropriate and repeatable viewing condition with accurate lights.  To some extent the viewingconditions may also be somewhat adjustable.

Personally I find d65 has never offered me colors that match my prints, but then again that is all based on the light used to view the prints once printed.   To me it's always been a process of trial and error, tweaking the profile white point until the screen offers a match.  With Spectraview II on my NEC, I've ended up dialing in a manual white point, currently set to x=0.346 y=0.373.  According to spectraview that equates to @4700 white point.  Sounds really warm, but matches by 3500 solux or 3000k Soraa Vivid LED (which surprisingly are cooler than the Solux) viewing lights very nice and really doesn't look extremely warm.  My second display is set to 6500 sRGB to simulate something closer to what most display manufacturers factory settings produce and side by side it is obviously "cooler" but it's not as extreme as you might think.
.
« Last Edit: January 10, 2018, 10:17:11 pm by Wayne Fox »
Logged

Andy_S

  • Newbie
  • *
  • Offline Offline
  • Posts: 14
Re: Why are dark blues purplish in test print
« Reply #12 on: January 11, 2018, 12:04:17 pm »

I have soft proofed the test image.  Attached are three screenshots.
  1.  The test image as displayed in the Print module
  2.  A soft proof of the image to print
  3.  Out of gamut display

Since the soft proof does not match the print module image and there are out of gamut colours does that mean I should recalibrate my monitor?  If so what values would you suggest?

Logged

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 20630
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: Why are dark blues purplish in test print
« Reply #13 on: January 11, 2018, 12:06:22 pm »

I have soft proofed the test image.  Attached are three screenshots.
  1.  The test image as displayed in the Print module
  2.  A soft proof of the image to print
  3.  Out of gamut display
Since the soft proof does not match the print module image and there are out of gamut colours does that mean I should recalibrate my monitor?  If so what values would you suggest?
1. Not soft proofing there, ignore what you see.
2. That's a soft proof within the confines of display gamut!
3. OOG overlay is buggy and nearly useless.

The Out Of Gamut Overlay in Photoshop and Lightroom
In this 25 minute video, I'll cover everything you need to know about the Out Of Gamut (OOG) overlay in Photoshop and Lightroom. You'll see why, with a rare exception, you can ignore this very old feature and still deal with out of gamut colors using modern color management tools.

YouTube: https://www.youtube.com/watch?v=00O-GTDyL0w
High resolution: http://digitaldog.net/files/OOG_Video.mp4
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
Re: Why are dark blues purplish in test print
« Reply #14 on: January 11, 2018, 01:28:16 pm »

1. Not soft proofing there, ignore what you see.
2. That's a soft proof within the confines of display gamut!
3. OOG overlay is buggy and nearly useless.

.................

1. Agreed
2. The display gamut may well exceed the printer gamut (for example see my IGFS/SC-P5000 gamut versus my PA271W gamut in the appended screen grab.
3. Agreed

These views are useful because they indicate that the "true blue" elements of the image, especially the lapus beads on her hand, really are blue, not purple. So the colour management between the display and the target photo and the soft-proofing portion of the profile seem to working properly. So this would bring the discussion back to the printer - if Andy has a colour-managed reflective scanning capability it could be useful to see unedited scans of the photo printed with the profile and the one printed with <Printer Manages Color>.
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

Doug Gray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2197
Re: Why are dark blues purplish in test print
« Reply #15 on: January 11, 2018, 02:15:34 pm »

1. Agreed
2. The display gamut may well exceed the printer gamut (for example see my IGFS/SC-P5000 gamut versus my PA271W gamut in the appended screen grab.
3. Agreed

These views are useful because they indicate that the "true blue" elements of the image, especially the lapus beads on her hand, really are blue, not purple. So the colour management between the display and the target photo and the soft-proofing portion of the profile seem to working properly. So this would bring the discussion back to the printer - if Andy has a colour-managed reflective scanning capability it could be useful to see unedited scans of the photo printed with the profile and the one printed with <Printer Manages Color>.

Yep. Not a question of OOG for the blue jewelry the woman is wearing.

A couple possibilities given the monitor is color managed and profiled.

1. The printer no longer matches the profiles. A custom profile should fix this if it's the cause.
2. There is metameric failure. Perhaps the print is being viewed in light from a CFL or poor quality LED light. The print should be viewed near a window to see if there is a significant difference.
Logged

Wayne Fox

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4237
    • waynefox.com
Re: Why are dark blues purplish in test print
« Reply #16 on: January 11, 2018, 02:43:18 pm »

The brightness of the printed matched the display and all the other colours are fine except for the dark blues which are purplish.
Just trying to clarify ... are you saying the print has purplish blues (which shouldn’t happen no matter what the display looks like) or the display has purplish blues.  I assume you haven’t tried to modify any colors or settings in the file.

What is the lighting you are using to evaluate the print?
Logged

Andy_S

  • Newbie
  • *
  • Offline Offline
  • Posts: 14
Re: Why are dark blues purplish in test print
« Reply #17 on: January 11, 2018, 03:28:17 pm »

Mark, I have an Epson V600 scanner but it is not colour managed.  You mentioned that you have printed using the P800 on Ilford GFS paper with both an Ilford and custom profile and that the blues were corrected.  This would seem to suggest there is something wrong in my workflow rather then the profiles.  Is it possible that the PA302W is calibrated incorrectly for this paper?

Wayne,  the prints have purplish blues and the colour on the display is correct.  The test images has not been modified.  I do not have a Solux light source but have viewed the prints under a varied of lighting conditions and the blues are purplish.

Any suggestions where I go from here?

Logged

Wayne Fox

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4237
    • waynefox.com
Re: Why are dark blues purplish in test print
« Reply #18 on: January 11, 2018, 05:29:24 pm »

Wayne,  the prints have purplish blues and the colour on the display is correct.  The test images has not been modified.  I do not have a Solux light source but have viewed the prints under a varied of lighting conditions and the blues are purplish.

Any suggestions where I go from here?
my recommended step would be to print the image onto a standard popular epson paper using Epson’s profiles, such as Premium Luster.  I guess to me the first step is to eliminate the possibility that the paper profile has a problem.  It doesn’t matter what settings you have the display and it doesn’t matter if the display matches anything with a test print, it should print correctly.  If it doesn’t and you have correctly entered the information into the printing dialog boxes, the color should appear correct on the print.  Using two different papers and profiles helps eliminate whether a profile might be problematic. If they don’t match then one profile might be an issue.  If they are very close, then maybe the print is correct and the color is what we should expect, so then we move to the process of getting the display to match that output. 

Blues such as that are notorious for showing the problem you describe unless using a very well made profile.  I use 4101 patch targets when I make my profiles just to help the color management system handle nuances such as this.  I’m sure Andrew was quite deliberate in the choice of what he included in his target to help reveal such problems.
Logged

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
Re: Why are dark blues purplish in test print
« Reply #19 on: January 11, 2018, 07:41:35 pm »

Mark, I have an Epson V600 scanner but it is not colour managed.  You mentioned that you have printed using the P800 on Ilford GFS paper with both an Ilford and custom profile and that the blues were corrected.  This would seem to suggest there is something wrong in my workflow rather then the profiles.  Is it possible that the PA302W is calibrated incorrectly for this paper?

Wayne,  the prints have purplish blues and the colour on the display is correct.  The test images has not been modified.  I do not have a Solux light source but have viewed the prints under a varied of lighting conditions and the blues are purplish.

Any suggestions where I go from here?

Andy, if you have a V600 scanner, the software driving the scanner, whether it is EpsonScan or SilverFast would have built in profiles, in this case for reflective colour scanning, so you can likely use it without worrying much about the colour management - worth giving it try anyhow.

The P800 and the profiles for Ilford GFS are wide enough gamut that the blues in Andrew's target print should not come out purple. The Ilford profile is pretty decent, but a well-made professional custom profile would be likely superior, if only because it is customized to that very printer, rather than Ilford's printer notwithstanding that the model is the same; manufacturing tolerances are pretty tight these days, but one never knows what performance variance there may be between Ilford's P800 and yours.

The reason for using the target print is that the colours are familiar and known, so what you see on the display doesn't matter, as Wayne mentioned.

As I mentioned above, when I looked at your settings I didn't see anything obviously wrong with your workflow.

The one thing we haven't seen yet is what those purple blues really look like - as I suggested, could be helpful if you could make those scans and let us see the problem "in the flesh".
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."
Pages: [1] 2 3   Go Up