Teal/Blue to green printing shift

[James Clark]

Hi - hoping someone has come across this problem and can offer a solution (or even a hint).

When I print this image from LR CC Classic (latest version), I get a dramatic color shift to green from the expected teal/blue.  This happens with ALL icc profiles I lay in but it prints CORRECTLY (i.e. consistently with the soft proof) when managed by printer.  I've soft proofed and the soft proof appears fine in the develop module - no color shift (nothing out of gamut either) when proofing with the icc profile.   

The weird thing is when I go to print preview just prior to printing the color shift IS there - but only with the icc profile applied.  (See attached graphics).

So basically:
* always ok in develop module soft proof with multiple different icc profiles applied
* never ok in print preview popup just prior to print with icc profile applied
* IS ok in in print preview popup just prior to print with managed by printer applied
* final print mimics the image in the final print preview, NOT the soft proof.


TIA for any thoughts.

(PS - I have another image that seems to have the same sort of color shift in the same tonal area.  Is there something odd about teals/greens that I'm unaware of?)

[Mark D Segal]

From what you say, it would appear that perhaps the effect of the profiling is not being picked up in the Print Module. Could you try making this print from Photoshop with colour management and ICC profile applied to see whether it comes out correctly? This would help determine whether the problem is confined to Lr or is instead something more systemic.

[James Clark]

From what you say, it would appear that perhaps the effect of the profiling is not being picked up in the Print Module. Could you try making this print from Photoshop with colour management and ICC profile applied to see whether it comes out correctly? This would help determine whether the problem is confined to Lr or is instead something more systemic.

Good thought. I’ll give it a try and report back. But to be clear, it’s more like the profiling process is borked, vs not being picked up. The print is consistent with the visual and soft proof when the profile is *not* applied in the print (ie when the printer is allowed to manage color)

[virojarvi]

When you set icc profile in LR, do you also set color management off in the printer driver? My initial error with Pro4000 was to have double color management.

[Mark D Segal]

When you set icc profile in LR, do you also set color management off in the printer driver? My initial error with Pro4000 was to have double color management.

That is definitely worth checking, though double colour management would usually show more severe errors than apparent in this case - nonetheless to verify.

James: Yes, I understood your first explanation of what is happening. So, the problem could arise from a number of places and what I suggested is just a start to ruling possibilities in or out. I'm thinking that if the profiling isn't working, one should first know whether it is specific to one application or more generic.

I had thought perhaps if those particular colours were out of gamut it could be a matter of the rendering intent and how OOG colours are being brought to the gamut boundary. But if they come out correctly with Printer Manages Color, then whatever paper and printer you are using the colours would seem to be in gamut. You say you encountered the same problem with a number of profiles, so perhaps ruling out a defective profile being the issue.

[James Clark]

That is definitely worth checking, though double colour management would usually show more severe errors than apparent in this case - nonetheless to verify.

James: Yes, I understood your first explanation of what is happening. So, the problem could arise from a number of places and what I suggested is just a start to ruling possibilities in or out. I'm thinking that if the profiling isn't working, one should first know whether it is specific to one application or more generic.

I had thought perhaps if those particular colours were out of gamut it could be a matter of the rendering intent and how OOG colours are being brought to the gamut boundary. But if they come out correctly with Printer Manages Color, then whatever paper and printer you are using the colours would seem to be in gamut. You say you encountered the same problem with a number of profiles, so perhaps ruling out a defective profile being the issue.

Thanks Mark!   I'm going verify color management today as virojarvi suggested, and I'm going to try printing from my Mac LR install(vs my primary production PC) and from Photoshop on the PC and see what happens.  Appreciate all the responses so far.   Also, I should post out that the actual print variation is more severe that the small jpegs I posted shows.  The base image is definitely blue, and the printed result is very very green.

[Mark D Segal]

James, could you tell us what paper and printer you are using, whether the inks are the printer manufacturer's inks and specifically what profile you would normally use for that paper and printer?

[James Clark]

James, could you tell us what paper and printer you are using, whether the inks are the printer manufacturer's inks and specifically what profile you would normally use for that paper and printer?

Sure - I was using a Canon Pro-10 and originally Hahnemuhle Wiliam Turner paper.  After several hours today, I seem to have narrowed the issue to either the paper or the Hahnemuhle-supplied icc profiles.   I did some test prints with a bunch of other papers I had on hand, some cool and some warm (specifically, Moab Entrada Rag Bright, Canon Pro Platinum and Hahnemuhle German Etching) and all were within reasonable proximity to the screen image (very minor cool/warm color shift as would be expected from the papers in question after adjusting the soft proof).

Interestingly the print preview window in LR still shows a rather vibrant green where the aqua ought to be regardless of profile and paper, but the final prints are still correct (i.e. blue).  I'm left to conclude that I probably either need a custom profile for the William Turner OR that the paper is just so warm that it's really hard to get color fidelity right.   Which is too bad, because I absolutely love the feel and look of the paper.  But, the German Etching is close in feel, and easier to control tonally.

Thanks everyone -  especially to Mark - for following along and offering suggestions!

[Mark D Segal]

Sure - I was using a Canon Pro-10 and originally Hahnemuhle Wiliam Turner paper.  After several hours today, I seem to have narrowed the issue to either the paper or the Hahnemuhle-supplied icc profiles.   I did some test prints with a bunch of other papers I had on hand, some cool and some warm (specifically, Moab Entrada Rag Bright, Canon Pro Platinum and Hahnemuhle German Etching) and all were within reasonable proximity to the screen image (very minor cool/warm color shift as would be expected from the papers in question after adjusting the soft proof).

Interestingly the print preview window in LR still shows a rather vibrant green where the aqua ought to be regardless of profile and paper, but the final prints are still correct (i.e. blue).  I'm left to conclude that I probably either need a custom profile for the William Turner OR that the paper is just so warm that it's really hard to get color fidelity right.   Which is too bad, because I absolutely love the feel and look of the paper.  But, the German Etching is close in feel, and easier to control tonally.

Thanks everyone -  especially to Mark - for following along and offering suggestions!

You are welcome. I think you have a profiling issue in two places: the monitor and the paper. It is possible that your monitor needs calibration and profiling, and that you need better paper profiles than those supplied by the paper manufacturers. Their profiles are generic to the printer they made them with, which may not be performing identically to yours.

[Dan Wells]

Another thing to think about is that smaller printers like the Pro-10 have more unit-to-unit variation than the big machines - they are made in far greater numbers with less individual adjustment, and they have less internal calibration, etc.

Dan

[James Clark]

Another thing to think about is that smaller printers like the Pro-10 have more unit-to-unit variation than the big machines - they are made in far greater numbers with less individual adjustment, and they have less internal calibration, etc.

Dan

Hi Dan - that makes sense.  I do have a Canon 8300 in the studio as well, but it's under the weather at the moment .  It would have been interesting to check the differences in final output!