Epson SC P600 Workflow

[Clark]

Hi all, I have an Epson SC P600 printer and I'm dreaming of getting simple and accurate screen -> print colour reproduction. Sadly I haven't yet achieved this.

Just to give a small intro - I am fairly experienced with the principles of colour management, but this is my first time owning a printer.

I am using an Eizo CG277 screen, El Capitan, and SC P600. I am printing on "Ilford Gallery Semi-Gloss Duo" paper.

When I first began to print, I used a fairly standard screen calibration - 6000K, 90cdm etc. I don't have a spectrophotometer, so I downloaded the canned profile from Ilford, which is made specifically for this exact paper / printer combination, and followed all their instructions for other settings.

I ran out a few prints. They looked pretty good, and I certainly wasn't expecting instant accuracy, but I noticed pretty quickly that there was a real problem with the rendering of subtle, pale reds, yellows and oranges (ie. skin tones). These were coming out too magenta (people looking all pink and flushed).

I kept getting hot skin-tones, whilst the more saturated yellows, reds and magentas were looking close enough to what I expected. I then tested nudging the top end of the curves in my image file, bumping the green a little, pulling down the blue. By the time I had the skin looking okay on paper, other colours in the print had been thrown out.

So here is what I did next:

- Carefully checking all the recommended print settings from Ilford, I printed out a single A3 test sheet with a variety of photos, a downloaded colour wedge and test strips in it, and then placed this next to my Eizo.

- With the image on screen, I turned on soft-proof mode in Photoshop CC 2017. This changes the colour on screen quite dramatically. I set the Proof Setup as below:

       - I used the same profile I am printing with (Ilford SC P600 semi-gloss) as the "device to simulate"
       - Preserve RGB numbers - no
       - Rendering intent - Relative Colorimetric
       - Simulate paper colour - yes
       - Black point compensation - yes

- I then calibrated my Eizo, and manually adjusted the numbers until what I had on my screen matched very closely to the print. However, I needed to dial IN this magenta cast on pale skin tones, which was coming out on the paper. But on screen this would throw the rest of my image out of whack (knocking pale blues much more quickly than it was affecting the actual skin, for example).

- So I made the final adjustments in the form of a few adjustment layers in photoshop. It only needed 3 or 4 clicks in hue/saturation etc, and I had an on screen image with the same colour characteristics as my printer, across a wide range of photos.

So that's how I set up my colour management, and it worked well for the run of prints I needed to make. I'm just wondering if there is a better way??

I am achieving accurate prints, but my screen's colour is now based on Photoshop's soft-proof mode, which is "simulating" colours using a canned Ilford profile. And on top of this, I am having to drop temporary adjustment layers into any print files as an extra level of soft-proofing (to simulate my printer's tendency to turn people pink), and then create additional layers beneath these to compensate, remembering to then switch off my temporary proofing adjustments before printing.

If anyone has any advice, suspicions, or can point out any glaring mistakes in my approach, I'd really appreciate your helP!!! Is it advisable to use Photoshop's soft-proofing mode? And if so, is it recommend to use the settings above, with the printing profile selected as the device to simulate?

[Mark D Segal]

The one thing I think you didn't mention, unless I missed it, is whether you have PROFILED the Eizo (this is different from calibrating it to the settings you mention). On the monitor calibration, I would argue that 6000K is too high - better bring this closer to 5000. The 90 cdm may be a bit low depending on the brightness of the environment. If you have some other illumination somewhere in the room, a setting around 110 or so may be more appropriate, but that wouldn't affect the main problem you are having.

Use a known printer test page with known "memory colours" for assessing the accuracy of the printer/profile combination. I would suggest the Outback Photo image you can download from here: http://www.outbackphoto.com/printinginsights/pi048/essay.html.

You also didn't mention whether you did a nozzle check before printing. Clogged nozzles can cause such issues, exaggerating the opposite colour of the channel that is clogged. The fact that the warm part of the colour spectrum is most affected, suggests a rendition problem with Cyan.

You also didn't mention which Media Type you have selected in the Epson driver. Media Type can make a big difference to accuracy of rendition. Remember, this is a 3rd party paper, so Ilford's profile needs to be "keyed" to a Media Type in the Epson driver. Make sure that the Media Type you specify in the driver is the same one that Ilford used for creating the profile. You can usually detect this in the Ilford profile naming and the accompanying data sheet they provide with their profile downloads.

A further problem (perhaps less likely) could be a mismatch between the colour space of the image file versus the image editing colour space (do you have your Photoshop warnings turned on?). For example, if you are trying to print sRGB photos in a ProPhoto working space that would be appropriate with a large gamut printer profile, exaggeration of certain colours can occur.

[digitaldog]

When I first began to print, I used a fairly standard screen calibration - 6000K, 90cdm etc.

And apparently those settings were wrong as any setting someone recommends that doesn't produce a match. Start again. Download a color reference image like the one below, and view this video on how, with the proper tools, you should be able to get a close(er) match between display and print:


http://www.digitaldog.net/files/2014PrinterTestFileFlat.tif.zip



Why are my prints too dark?
A video update to a written piece on subject from 2013

In this 24 minute video, I'll cover:

Are your prints really too dark?
Display calibration and WYSIWYG
Proper print viewing conditions
Trouble shooting to get a match
Avoiding kludges that don't solve the problem

High resolution: http://digitaldog.net/files/Why_are_my_prints_too_dark.mp4
Low resolution: https://youtu.be/iS6sjZmxjY4

[Frodo]

Hi Clark
I'll venture a response following those of two respected experts.
First, use a Datacolor Spyder or similar to calibrate your screen.
I bought a P600 a few months ago. The colours in my first print were out of whack like you describe using Ilford Smooth Pearl. I was using the downloaded profile, but did not uncheck "printer manages colour". So Lightroom and the P600 were in conflict.
Once that was sorted, the colours in the print match the soft proofed image on my screen.

[Mark D Segal]

................I was using the downloaded profile, but did not uncheck "printer manages colour". So Lightroom and the P600 were in conflict.
Once that was sorted, the colours in the print match the soft proofed image on my screen.

Good point - double colour management could well be the problem. When using an ICC profile in Photoshop make sure in Photoshop's Print Module it says that Photoshop Manages Color and in the Epson driver, make sure Printer Color Management is OFF.

[unesco]

Clark,

Assuming you have some basic knowledge of color management (and do not do double color management) and have already read and watched all videos Andrew has recommended ;-) a few more things come to my mind:

1) You have brilliant monitor - I assume you have used ColorNavigator and colorimeter built into your monitor for calibration...

2) I would recommend to lower a bit temperature of your display, I use 5500K for fine art printing. Also if you do not work in strong day light but using some artificial, 90 nits seems to be a bit too much, I use 60.

3) Mentioned Ilford paper has quite cold white point what could make magenta shift - try to use some warmer paper.

4) Skin tomes being too magenta is a typical effect of some bad profile, so called "orange turns magenta" effect - quite common for generic profiles. Try to make/order a custom profile. Ideally, make it for another, warmer paper, Epson Premium (Semi)Glossy would be a good starting point.

[Clark]

Clark,

Assuming you have some basic knowledge of color management (and do not do double color management) and have already read and watched all videos Andrew has recommended ;-) a few more things come to my mind:

1) You have brilliant monitor - I assume you have used ColorNavigator and colorimeter built into your monitor for calibration...

2) I would recommend to lower a bit temperature of your display, I use 5500K for fine art printing. Also if you do not work in strong day light but using some artificial, 90 nits seems to be a bit too much, I use 60.

3) Mentioned Ilford paper has quite cold white point what could make magenta shift - try to use some warmer paper.

4) Skin tomes being too magenta is a typical effect of some bad profile, so called "orange turns magenta" effect - quite common for generic profiles. Try to make/order a custom profile. Ideally, make it for another, warmer paper, Epson Premium (Semi)Glossy would be a good starting point.

Hi Unesco, thanks for this reply. As I mentioned I do have knowledge of colour management. I understand the basic principle that a digital image is just data, and a certain screen, in a certain state will only show me a certain interpretation of that data.

However, just as one may observe a print in a variety of lighting conditions and conclude that the skin of a person seems quite pink, I am working with images that I have viewed on multiple devices and in various different calibrated states on my Eizo, and have not noticed an imbalance of skin-tones relative to the rest of the image.


Thanks also for the pointer that Ilford paper may be a pretty cold white, and something to be aware of. As I said I may understand colour management reasonably well, but I am really a newbie to printing, papers, and printers. I'm quite limited to the choice of paper as I need a double-sided option for my portfolio.

Thanks to others who've replied too. I am making progress towards a better configuration of my workflow. I watched your videos digital dog and learned a few things :-)

Also Mark - just to quickly answer - I know I didn't list all my settings etc but yes I am using Adobe98 as my working space and image profile, I am using Epson Semi Gloss as the media-type in my printer driver as recommended by Ilford for their "semi-gloss" paper. One thing I wanted to ask you though - when you say have I "PROFILED" my screen? I have run a calibration using my screen's built-in colorimeter. I then fine tuned this with some manual tweaks in Eizo's ColorNavigator, whilst comparing side by side to a test-sheet from my printer. Is "profiling" different to "calibrating" to an output reference?

The main thing I have done since my original post is visit the Ilford website. I just happened to go into their profiles download section, and noticed that the profile for my printer model / paper type has an updated name: Previously "GPSGD_EPSCP600_EPS" now "ILFORD_GPSGD11_EPP600_EPSn.icc"..?

I downloaded and installed, and there is an immediate improvement in the "orange turns magenta" tendency. I also noticed the on-screen soft-proof produces a much less dramatic shift in colour than the previous profile.

So I have printed another colour reference image, this time the one suggested by digital dog, and the print shows much less of the unexpected pinky tones. Overall a better balance with what I see in the digital file.

However, I have also run another nozzle check on the advice of Mark. All patterns are perfect except the yellow. There are a few missing bars - The pattern is broken - but my question is - is it worth doing a clean cycle for this? Is this something you guys would definitely address with head cleaning or do you exercise a certain amount of tolerance when you see one or two breaks in the pattern? See image attached.

[ColourPhil]

Hi,
Absolutely worth doing a head clean!
Especially as you have colour problems.
Those blockages can only get even more clogged with time.
Cheers,
Phil

[ColourPhil]

Hi,
I forgot to say, just do one or two cleanings only, then leave overnight.
Otherwise you could 'flood' the heads.
Cheers,
Phil

[Clark]

Thanks ColourPhil. Have just run a single clean cycle, and the nozzle check pattern is unbroken now.   

This afternoon I'll try running the same test print I made yesterday to see if I can see a difference in print! 

[Mark D Segal]

I'm quite limited to the choice of paper as I need a double-sided option for my portfolio.


Also Mark .......... One thing I wanted to ask you though - when you say have I "PROFILED" my screen? I have run a calibration using my screen's built-in colorimeter. I then fine tuned this with some manual tweaks in Eizo's ColorNavigator, whilst comparing side by side to a test-sheet from my printer. Is "profiling" different to "calibrating" to an output reference?

The main thing I have done since my original post is visit the Ilford website. I just happened to go into their profiles download section, and noticed that the profile for my printer model / paper type has an updated name: Previously "GPSGD_EPSCP600_EPS" now "ILFORD_GPSGD11_EPP600_EPSn.icc"..?

I downloaded and installed, and there is an immediate improvement in the "orange turns magenta" tendency. I also noticed the on-screen soft-proof produces a much less dramatic shift in colour than the previous profile.

So I have printed another colour reference image, this time the one suggested by digital dog, and the print shows much less of the unexpected pinky tones. Overall a better balance with what I see in the digital file.

However, I have also run another nozzle check on the advice of Mark. All patterns are perfect except the yellow. There are a few missing bars - The pattern is broken - but my question is - is it worth doing a clean cycle for this? Is this something you guys would definitely address with head cleaning or do you exercise a certain amount of tolerance when you see one or two breaks in the pattern? See image attached.

Hi Clark,

As mentioned above, yes, clean the nozzles if the test pattern indicates a blockage.

Good that you updated the profile. That seems to have looked after a good part of your problem, but is it COMPLETELY resolved now, or do you still see a mismatch?

Calibrating and profiling a display are two different things. Calibration sets certain parameters - usually White Point, Gamma and Brightness. Profiling actually characterizes the display's behaviour and creates a profile for it containing this characterization, which helps guide how the display renders the colours thrown at it. Very much like what a printer profile does, except it's for a display. I don't know how Eizo implements these two functions in their displays because I'm using an NEC with BasicColor software and never owned an Eizo. But one way or another, both those functions need to be implemented, and once done, it should not be necessary, or even desirable most of the time, to tweak anything.

Dual-sided paper: your Ilford choice is fine, but you may also wish to try Moab Lasal Dual Semigloss (I reviewed it recently on this website) if you want a luster finish, or Moab Entrada Rag Natural 2-sided 190gsm if you want matte finish (also in the same review). Both nice papers.

Andrew's test image is of course very useful but I recommend you also conduct your printer tests using the Outback Photo image I referenced, paying attention to the guide that is downloadable with it explaining how to use each section of the image in your printer performance evaluation; this is particularly helpful.

[Clark]

Hi Clark,

As mentioned above, yes, clean the nozzles if the test pattern indicates a blockage.

Good that you updated the profile. That seems to have looked after a good part of your problem, but is it COMPLETELY resolved now, or do you still see a mismatch?

Calibrating and profiling a display are two different things. Calibration sets certain parameters - usually White Point, Gamma and Brightness. Profiling actually characterizes the display's behaviour and creates a profile for it containing this characterization, which helps guide how the display renders the colours thrown at it. Very much like what a printer profile does, except it's for a display. I don't know how Eizo implements these two functions in their displays because I'm using an NEC with BasicColor software and never owned an Eizo. But one way or another, both those functions need to be implemented, and once done, it should not be necessary, or even desirable most of the time, to tweak anything.

Dual-sided paper: your Ilford choice is fine, but you may also wish to try Moab Lasal Dual Semigloss (I reviewed it recently on this website) if you want a luster finish, or Moab Entrada Rag Natural 2-sided 190gsm if you want matte finish (also in the same review). Both nice papers.

Andrew's test image is of course very useful but I recommend you also conduct your printer tests using the Outback Photo image I referenced, paying attention to the guide that is downloadable with it explaining how to use each section of the image in your printer performance evaluation; this is particularly helpful.

Thanks Mark, so then I am "profiling" my Eizo display. It has a built-in colorimeter, and software that works together with this to create profiles. I only use the screen with profiles created in this way.

[Mark D Segal]

Sounds fine.

[unesco]

Thanks Mark, so then I am "profiling" my Eizo display. It has a built-in colorimeter, and software that works together with this to create profiles. I only use the screen with profiles created in this way.

If you are using Eizo's Color Navigator software (the only reasonable option for this kind of top displays) you are doing both: calibration and profiling. Calibration changes your monitor settings directly in hardware, profiling makes icc profile saved in your operating system.

As for previous problems, glad they seem to be resolved - it looks to be a combination of several problems in one place. Printing is a very complex process, and requires time and try and error approach. A lot of time...
Good luck! :-)

[Garnick]

I downloaded and installed, and there is an immediate improvement in the "orange turns magenta" tendency. I also noticed the on-screen soft-proof produces a much less dramatic shift in colour than the previous profile.

So I have printed another colour reference image, this time the one suggested by digital dog, and the print shows much less of the unexpected pinky tones. Overall a better balance with what I see in the digital file.

However, I have also run another nozzle check on the advice of Mark. All patterns are perfect except the yellow. There are a few missing bars - The pattern is broken - but my question is - is it worth doing a clean cycle for this? Is this something you guys would definitely address with head cleaning or do you exercise a certain amount of tolerance when you see one or two breaks in the pattern? See image attached.

Hi Clark,

After taking a close look at the yellow channel you have presented, I must say that this is anything but a few gaps.  I have learned over many years of printing with Epson ProGraphic printers that a couple of small gaps near the bottom of the nozzle check can be overlooked and generally not show in the final print.  However, what you have shown us is what I would call a major problem with the yellow channel, and one that would most definitely present an issue in a print.  As a matter of fact I'm surprised that you haven't mentioned any sort banding issues that would often accompany this type of nozzle "blockage".  However, please forgive me if it has been mentioned and I have overlooked it.  A simple nozzle check and cleaning cycle is, in my opinion, the foundation of a good printing workflow.  Once the printer has booted up each morning, the first order of business is to run a nozzle check.  If there are any noticeable issues I run a cleaning cycle, followed by another NC.  In the traditional "Wet Lab" scenario it was referred to as QC(Quality Control), something that was done every morning for every piece of equipment that would have any part in the processing of film or print. 

As I started this I referred to your nozzle check pattern as indicating a "major problem".  I will however offer a caveat on that statement, to say that I believe it is an issue that definitely calls for a cleaning cycle or two, and will likely show some very good results following those cycles.  If not completely fixed, you should perhaps let the printer rest overnight and start again the next morning, as has been mentioned previously.  Sometimes too many clean cycles can have an adverse effect on the print head, as we all know. 

Gary           

[Clark]

Hi Clark,

After taking a close look at the yellow channel you have presented, I must say that this is anything but a few gaps.  I have learned over many years of printing with Epson ProGraphic printers that a couple of small gaps near the bottom of the nozzle check can be overlooked and generally not show in the final print.  However, what you have shown us is what I would call a major problem with the yellow channel, and one that would most definitely present an issue in a print.  As a matter of fact I'm surprised that you haven't mentioned any sort banding issues that would often accompany this type of nozzle "blockage".  However, please forgive me if it has been mentioned and I have overlooked it.  A simple nozzle check and cleaning cycle is, in my opinion, the foundation of a good printing workflow.  Once the printer has booted up each morning, the first order of business is to run a nozzle check.  If there are any noticeable issues I run a cleaning cycle, followed by another NC.  In the traditional "Wet Lab" scenario it was referred to as QC(Quality Control), something that was done every morning for every piece of equipment that would have any part in the processing of film or print. 

As I started this I referred to your nozzle check pattern as indicating a "major problem".  I will however offer a caveat on that statement, to say that I believe it is an issue that definitely calls for a cleaning cycle or two, and will likely show some very good results following those cycles.  If not completely fixed, you should perhaps let the printer rest overnight and start again the next morning, as has been mentioned previously.  Sometimes too many clean cycles can have an adverse effect on the print head, as we all know. 

Gary           

Hey Gary, thanks for your reply. It's really interesting for me to hear your experience and opinions on this. I now know that this kind of break of pattern is definitely indicative a problem and not to be ignored. Also happy to learn the easy way (from you guys) that this may get worse over time if not cleaned out. Likewise very glad to know that too many clean cycles can cause problems?

Happily, it only took one clean cycle yesterday, and I printed an unbroken yellow nozzle check pattern. I'll definitely follow your advice and run nozzle checks routinely before any printing.

Interestingly, I followed up by printing a few tests, but really couldn't see any difference in images from before and after the cleaning. I'm glad to have this at least under control and managed though. When I discovered this I really suspected it may have revealed itself to be another factor in the "orange turns magenta" problem in my print.

I would say now that the overall balance of colours in my print output is very close to what I expect to see. Coming out a little denser and cooler, but this seems to be a global characteristic and I hope when I re-profile my screen I can bring it very closely in-line with what I see on paper 

[Garnick]

Hey Gary, thanks for your reply. It's really interesting for me to hear your experience and opinions on this. I now know that this kind of break of pattern is definitely indicative a problem and not to be ignored. Also happy to learn the easy way (from you guys) that this may get worse over time if not cleaned out. Likewise very glad to know that too many clean cycles can cause problems?

Happily, it only took one clean cycle yesterday, and I printed an unbroken yellow nozzle check pattern. I'll definitely follow your advice and run nozzle checks routinely before any printing.

Interestingly, I followed up by printing a few tests, but really couldn't see any difference in images from before and after the cleaning. I'm glad to have this at least under control and managed though. When I discovered this I really suspected it may have revealed itself to be another factor in the "orange turns magenta" problem in my print.

I would say now that the overall balance of colours in my print output is very close to what I expect to see. Coming out a little denser and cooler, but this seems to be a global characteristic and I hope when I re-profile my screen I can bring it very closely in-line with what I see on paper 

First Clark, you are very welcome and I hope you can get your printer under control very soon.  Even though you would like to be able to print what you see on your display and have a total match, in my opinion that scenario is analogous to a dream.  Not a nightmare by any means, but definitely a dream.  We all strive for that degree of control, but really it all comes down to some level of Hard Proofing, in step with Soft Proofing.  Personally I would never, under any circumstances, make what I believe to be the final adjustments to an image and then commit it to print without at least one test strip.  This way of thinking and working dates back to the late 60s when I first started working in a custom colour lab in Toronto, where every "negative" had to be tested, usually more than once, before a full print was exposed.  That system has carried through many years of owning and operating my own colour lab.  In the days of video analyzers for colour printing we would also run a small test print first and of course it would usually be considerably closer to a final print, but still further adjustments had to be made in many cases.  The digital darkroom/lightroom is no exception.  There are simply too many variables to overcome in order to arrive at a display that approaches the final print in every way.  Even Andrew Rodney mentions this in his "Why Is My Print Too Dark?" video.  Nothing is perfect, but some things are "more perfect".  By that I mean that "perfect' seems to be a matter of degree, and sometimes we just have to saw it off at some point.  I will offer another caveat here; all of this is of course simply my opinion, based on many years of printing, both traditionally and digitally. 

And now a word about not running too many consecutive cleaning cycles.  I believe it was Mark Segal who first brought that valuable piece of information to the forum several years ago, after discussing the "issue" with Epson Techs.  I try not to offer any sort of advice without also including the original author of that advice, something that I did in my initial reply.  Please excuse my faux pas Mark.

Gary

             

[Mark D Segal]

No "faux pas" to worry about Gary, all's cool. I would only suggest that whether or not one needs test strips in the context of a properly implemented colour management workflow is open for discussion. While I can imagine some arcane or unusual circumstances in which this can provide value-added, as a general proposition I for one would find it rather redundant these days. But as usual - "it all depends....". Cheers, Mark

[Garnick]

No "faux pas" to worry about Gary, all's cool. I would only suggest that whether or not one needs test strips in the context of a properly implemented colour management workflow is open for discussion. While I can imagine some arcane or unusual circumstances in which this can provide value-added, as a general proposition I for one would find it rather redundant these days. But as usual - "it all depends....". Cheers, Mark

Hi Mark,

I appreciate your comments as always.  However, and of course there's often a "however" involved.  Although I do calibrate and profile my NEC displays on a regular basis, I also realize that not everything I can see on the display will necessarily be transferred to the print without implementing various adjustments.  Therefore my workflow has always, and probably will always involve at least one 2x8" test strip of an indicative area of the image.  On occasion, depending on the size of the final print, I will test a larger area of the image just to be as certain as possible that all is well and ready for a full sized print.  I suppose I may have exaggerated the necessity of the test print to some extent, but habits are often difficult to break.  Often one small test will suffice, in place of a full print that could possibly end up in the trash.  My customers are very particular about their work, as am I.  That's not to say that every full print is indeed a final print ready for the customer, but the test usually gets me a lot closer to where I want to be.  And of course when I'm reproducing original art work there are often many tests of different areas of the image, since not all colours etc. will necessarily reproduce properly.  In that case I may have to isolate certain areas of the image and make adjustments accordingly.  One thing I will mention, and it's something that I have been thinking about since I changed my viewing/evaluation lighting.  During all of my colour printing lab life I've been using 5000K fluorescent bulbs for evaluating the colour consistency and density of the prints.  Since my move to my home location I have introduced 4000K LED Bars from Lumicrest as my light source, and have been very pleased with the results.  What I haven't done yet is calibrate and profile the display to 4000K as well, something I'll be doing soon.  I usually calibrate at 5500K, but since I am now using a somewhat warmer evaluation light source I feel a 4000K - 4500K display would probably serve me better.  At least something to consider and put to the test.  Of course with SpectraView it's also very easy to run two or more separate calibrations and then switch back and forth for different purposes.   

Just a little comment about redundancy Mark.  Sometimes redundancy can actually save time, as well as materials.  But, as you mentioned, it all depends.  Agreed! 

Gary     

[Mark D Segal]

Reproducing original artwork is one of those special cases where extra-ordinary measures are often required. In this area, so much depends on the tone and colour correctness of the original capture. This takes special kinds of capture-stage profiling and lighting which I suspect many don't always get right, making the printing that much more challenging.