Luminous Landscape Forum
Raw & Post Processing, Printing => Printing: Printers, Papers and Inks => Topic started by: fike on January 11, 2022, 03:41:23 pm
-
I got my new P900 last weekend and have been cranking some of my favorite prints through it. I moved down from my 24" 7880 roll printer because I was tired of dealing with the curl of roll paper and such a huge printer taking up space in my studio.
But I'm having trouble with blue-green. It's muddy. Kinda brownish, maybe orange yellow, but it definitely loses the blue green. The difference isn't subtle. My wife notices it, and she usually looks at two comparison images and tells me that they both look the same to her.
- My monitor is calibrated, and I confirmed that the profile is in effect. It is an adobe RGB display.
- I am using the Epson-provided paper profiles.
- I have the same problem with red river papers and moab papers
- I understand how to avoid double color-management (photoshop manages color, not printer).
- Nozzle check looks perfect
- I ran a head cleaning
- I've run at least ten 8.5x11 images through it
- I have soft-proofed and brought all the colors in-gamut
- source image is in adobe RGB
- photoshop working space is adobe RGB
- just for fun, I printed the same images from an non-color managed iPad pro and the same problem exists
I haven't gone so far as to make a custom, homemade paper profile yet, but that is my next move. I don't think that is a good idea. I shouldn't need to make custom profiles to get good (not perfect, but good) results out of off-the-shelf papers.
Have I missed anything?
-
Don't want to be insulting, but have you run a standard test image?
Richard Southworth
-
I'd try dumping your Epson drivers and profiles. There are tutorials on how to do this.
Reinstall them from the Epson site, using the latest info for your operating system.
I've found Macs want to use their own versions of drivers instead of using the Epson drivers.
Good luck.
-
Yes, but in those it is hard to see much because you aren't comparing to anything. The gradations were smooth. I'm not sure what else to look for without a reference. It looked fine.
https://www.northlight-images.co.uk/printer-test-images/#colour_printer_test_images
I did print an x-rite color checker and compare it to my color checker. It was very close. (I am aware of the million ways that photographing and printing a color checker can go wrong, but I did it anyway.)
https://www.babelcolor.com/colorchecker-2.htm#CCP2_images
-
The image you chose is a photograph, a reasonable test image but not one that puts the same stress on a printer as some others. My favorite is the Printer Evaluation Image in the Pro Photo workspace,
http://www.outbackphoto.com/printinginsights/pi048/essay.html
Andrew (Digital Dog) also has some good ones. The point of using a reference image is you don't have to compare to a reference, you should be able to evaluate your printer performance by image examination alone. Read some of the tutorials that go along with the reference images.
Richard Southworth
-
D50 monitor calibration?
-
You should always test output using good color reference images designed for that task. The color reference images RGB values are such they are set for output and are editing and display agnostic. Test the output this way and examine for the same color issues so we know it's not your image specific issues causing the problems:
http://www.gballard.net/photoshop/pdi_download/
http://www.normankoren.com/makingfineprints1A.html#TestPrint
http://www.digitaldog.net/files/2014PrinterTestFileFlat.tif.zip
-
I printed @digitaldog's test image. The individual color ramps all look good. the black and white ramps look good. the one place that seems poor is on the mixed area (I don't know what it is called). In there area where the blue blends to pink and then red, the gradients seem abrupt. It doesn't seem to blend well. (see the attached image.) I don't understand why this would be happening if the individual ramps are good and the original sample image is good.
On a related note, how can it be that almost all the color patches are out of gamut when I soft proof for the paper in photoshop (I used premium luster)? Why does the sample image still look good?
-
Hard to comment on a photo of the print but it looks pretty okay and this Granger Rainbow is a real torture test. You want to examine the pure blues and greens of which you said were muted; are they?
-
Hard to comment on a photo of the print but it looks pretty okay and this Granger Rainbow is a real torture test. You want to examine the pure blues and greens of which you said were muted; are they?
I was pretty sure that looking at a photo of it was going to be mostly worthless. It's good to know that it is called a Granger Rainbow. Hearing that it is a torture test makes me feel better, but doesn't get me closer to solving my problem.
As for the blues and greens. They look clean and smooth. They don't show anything that would lead me to believe they aren't perfect. I talked to Epson. They keep asking me to check the driver and make sure that I have color adjustment set to off in the driver. I keep telling them that this is my fourth Epson printer and that I know how to use the driver (unless there is something new in the P900).
The strange thing with the muddiness is that it is consistent across papers and profiles while all the basic functions of the printer look perfect. It is as if some firmware was offsetting the profile (I know that is not a real thing). At this point I think I will try to make a custom profile for epson premium luster with my old colormunki to see if that changes anything.
-
Here are the images I am using to abuse the printer. They are very green.
https://flic.kr/p/JDMnYt
(https://live.staticflickr.com/1751/27994548567_7dd796bc6a_c_d.jpg)
https://flic.kr/p/MVhAjD
(https://live.staticflickr.com/1938/30139066097_72a181e702_c_d.jpg)
-
I printed @digitaldog's test image. The individual color ramps all look good. the black and white ramps look good. the one place that seems poor is on the mixed area (I don't know what it is called). In there area where the blue blends to pink and then red, the gradients seem abrupt. It doesn't seem to blend well. (see the attached image.) I don't understand why this would be happening if the individual ramps are good and the original sample image is good.
On a related note, how can it be that almost all the color patches are out of gamut when I soft proof for the paper in photoshop (I used premium luster)? Why does the sample image still look good?
The granger chart, like a series of circles sometimes called "Bill's Balls" after Bill Atkinson is a way to look for discontinuities. Probably best to view them in device RGB space rather than Adobe or especially ProPhoto RGB. This can be done by assigning the printer's profile to the image. Transitions will appear much smoother since the image is now always in gamut. But be warned, the image's colors will be heavily shifted so only do this to examine granger and other synthetics like Bill's Balls.
The other part is to use a reference image that is in gamut. There are many such images. Ideally, you print a reference image in Relative Colorimetric and compare it side by side to one printed by someone that can provide a physical reference print. When I print such an image using custom profiles on my Canon and Epson printers I cannot tell one from the other*. So long as the image is in gamut for both printers.
* Big caveat. This is true only when the print is illuminated at an angle with no reflected glare. If I can see any reflections, then bronzing/gloss differential differences between the printers stand out.
Ideally, suppliers of custom profiles could, for an additional charge, include a printed reference image made on their own calibrated/profiled printer. Since, when the profile is made, the vendor could automate gamut checking. This validates the supplied image is in gamut for the customer's printer/paper combo. So when the customer makes a print of the reference image with the profile they can view them side by side.
-
Are the images in your last post the source images for printing? I downloaded the first one, did not have an embedded profile. Please post one with the embedded profile intact, trying to see where your problem lies. Don't worry about how it looks on the web, I'll load it into Photoshop for evaluation.
Richard Southworth
-
Those are from flickr and probably had the profile stripped.
Here are the tiffs. One is adobe RGB and the other sRGB.
https://trailpixie.net/ll_samples/llanberis_waterfall.tif
https://trailpixie.net/ll_samples/white_mountains.tif
-
I downloaded the first one, brought it into Photoshop, and soft proofed with various printers, on luster type papers. All looks ok, and as you stated everything is apparently in gamut. It is a dark image, with only the bottom of the waterfall in the upper range. What type of light are you using to view the print? Take it outside and evaluate. It's easy for green to go muddy under some types of illumination, particularly florescent.
Richard Southworth
-
I think the next thing I would try is to print using ‘Printer manages colour’ thus eliminating the colour managed workflow and see how this looks.
If the results are still similar this would suggest that it may be something to do with the printer itself.
-
I think the next thing I would try is to print using ‘Printer manages colour’ thus eliminating the colour managed workflow and see how this looks.
If the results are still similar this would suggest that it may be something to do with the printer itself.
I did try that with similar results.
-
I made some progress this week. One of the images (looking down the autumn creek in New Hampshire) is in sRGB. My monitor is in Adobe RGB mode. I thought it would be enough to put photoshop in the sRGB working color space because (I reasoned) sRGB is a subset of aRGB. Nope. I needed to change the LCD mode AND the photoshop working space to see the appropriate soft proof rendering. This laptop display is the first adobe RGB display I have owned, so I have never dealt with changing display modes before. To further the problem, the original laptop driver didn’t have an option to switch display modes, so I didn’t even have the switches and knobs to do what I needed until I discovered that my display driver utility needed to be updated. (This is an MSI creator 17, an otherwise decent laptop and display.)
The waterfall photo still isn’t printing as I think it should, but it isn’t as far off. It required more work to bring it in gamut, so this image may be more challenging from a operator-skill point of view than a printer functionality point of view. Of course, the process of bringing images in gamut without ruining the vibrancy can be tedious and challenging.
-
I made some progress this week. One of the images (looking down the autumn creek in New Hampshire) is in sRGB. My monitor is in Adobe RGB mode. I thought it would be enough to put photoshop in the sRGB working color space because (I reasoned) sRGB is a subset of aRGB. Nope.
Nope, indeed and green is a big difference here:
The benefits of wide gamut working spaces on printed output:
This three part, 32 minute video covers why a wide gamut RGB working space like ProPhoto RGB can produce superior quality output to print.
Part 1 discusses how the supplied Gamut Test File was created and shows two prints output to an Epson 3880 using ProPhoto RGB and sRGB, how the deficiencies of sRGB gamut affect final output quality. Part 1 discusses what to look for on your own prints in terms of better color output. It also covers Photoshop’s Assign Profile command and how wide gamut spaces mishandled produce dull or over saturated colors due to user error.
Part 2 goes into detail about how to print two versions of the properly converted Gamut Test File file in Photoshop using Photoshop’s Print command to correctly setup the test files for output. It covers the Convert to Profile command for preparing test files for output to a lab.
Part 3 goes into color theory and illustrates why a wide gamut space produces not only move vibrant and saturated color but detail and color separation compared to a small gamut working space like sRGB.
High Resolution Video: http://digitaldog.net/files/WideGamutPrintVideo.mov
Low Resolution (YouTube): https://www.youtube.com/watch?v=vLlr7wpAZKs&feature=youtu.be
1285 characters, 210 words
-
Thanks for posting those @digitaldog. After watching them I understand why this is a nonsense.
I made some progress this week. One of the images (looking down the autumn creek in New Hampshire) is in sRGB. My monitor is in Adobe RGB mode. I thought it would be enough to put photoshop in the sRGB working color space because (I reasoned) sRGB is a subset of aRGB. Nope.
I spent a few more hours with this yesterday, the yellowish cast remains. I added another image to the test. It's another of my deep forest greens, and it behaves the same way. I finally got a chance to try brute force by using my colormunki to make a new profile on epson premium luster. Two interesting things happened:
1. The color gamut shown when soft proofing was far smaller
2. There was NO CHANGE to the color cast after printing with the new profile
Finally, I have viewed the image on two additional displays. They are quality manufacturer calibrated displays, though without custom profiles. They look identical to my profiled laptop aRGB display.
I keep hoping that I will find something that leads me away from a problem with the printer. As a matter of fact, the printer is so consistently off, that I feel like it must be something I am doing wrong.
On a possibly related tangent, I ran a head alignment routine on the printer, and it seemed to be way off. Every adjustment line was at the minimum setting of 1 out of 9 when the default appears to be 5. After running it a second time, all the values clustered between 4 and 7. There was a noticeable increase in sharpness (or local contrast) afterward. Could this printer be a sample that is far out of spec?
-
if you can post the icc profile you made it may help to get a better understanding
-
if you can post the icc profile you made it may help to get a better understanding
If the results are the same with the printer managing the colour (as stated by the OP) then it's unlikely to be the profile; more likely to be something to do with the printer itself.
-
i'm not asking because i think the profile is the problem necessarily, but it will help visualize what's happening since it can be opened and analyzed in colorthink or the like
-
Here's the profile.
I am coming to the conclusion that I need to ship the darn thing back to B&H. That is a big pain in the ass, and I will need to pay shipping, but I really can't find anything wrong with my workflow.
-
i'm not asking because i think the profile is the problem necessarily, but it will help visualize what's happening since it can be opened and analyzed in colorthink or the like
We need the profile and we need an image (low rez; 300x300 pixels is fine) to plot both in CTP.
-
These are the ones I posted up further in the thread.
https://trailpixie.net/ll_samples/white_mountains.tif
https://trailpixie.net/ll_samples/llanberis_waterfall.tif
-
...and while using that custom profile was a useful diagnostic for me, I really think I should be able to get adequate quality out of a standard epson profile.
-
These are the ones I posted up further in the thread.
https://trailpixie.net/ll_samples/white_mountains.tif
https://trailpixie.net/ll_samples/llanberis_waterfall.tif
In sRGB; that's a deal breaker right there.
-
One of them is in sRGB, the other is in aRGB.
but....I thought you could still print-match sRGB albeit with am smaller gamut.
-
One of them is in sRGB, the other is in aRGB.
but....I thought you could still print-match sRGB albeit with am smaller gamut.
You're clipping data in both which is pointless and isn't fair to compare to the profile gamut.
Render the raw into ProPhoto RGB, post a 300x300 pixel version, then I can plot that image gamut to the profile; we can see what doesn't 'fit'.
-
It's a panoramic stitch so I had to remake it. I haven't printed it here yet. I am going to do that now.
-
Here's the other one in ProPhoto.
-
The solid plot is your profiles gamut.
The 'dots' are the gamut of the image.
As you can see, a lot of greens fall outside printer gamut.
Is that normal or ideal? Don't know as the profile doesn't tell me much about the actual profile process. I can see in the tags, it was created with: 4CreatorAppi1Profiler. Copyright tag X-rite.
-
Not sure this is useful but I'll point out that the Gamut Volume of my Luster profile (3880) is 759,745, the gamut volume of you P900 is 665,481. I'd kind of expect the P900 would have a larger color gamut than the 3880.
-
...and that's the rub. My custom profile shows a smaller gamut than the default epson profile. I think we have come full circle and I have to send this P900 sample back to the manufacturer.
Can you plot what that image gamut would be on the epson-made premium luster profile. You may recall, the profile I made appeared to have a much smaller gamut because the gamut warning showed far more gray on the image than the standard epson profile when I soft proofed.
Thanks for plotting that. This is helping me a lot!
-
Can you plot what that image gamut would be on the epson-made premium luster profile.
What profile would that be?
-
This is the one that is installed with Epson's driver.
-
This is the one that is installed with Epson's driver.
Gamut volume is 784,522
This is the plot (not 100% identical rotation but you get the idea):
-
Can I conclude that because the custom profile I made with the X-rite colormunki and software had a smaller gamut than the out of the box one, then my printer isn't achieving the expected performance of the manufacturer specifications.
--or--
Am I pushing it too hard in the green channels and it is giving the image a brownish-yellow hues because it can't replicate that range?
-
From the comparison with the Epson profile and your profile, obviously SOMETHING is wrong... I would be very hesitant to say it would be a hardware problem unless you're completely missing a channel, as the process of mixing cyan and yellow to make green isn't all that complicated in itself. If it were a problem of hardware, it would most likely manifest itself in some other way (such as banding, or posterization when a channel is completely gone). Either there's a serious problem with the inks, or there's something in the color management workflow that's still not quite right...
Comparing your profile with Epson's, it seems like you're likely at a lower TIL, which shouldn't be happening, but there's no evidence of any individual channel actually missing. I would try profiling again using "Epson Glossy" as your media type, and at least 720x1440 or even 1440x2880 for resolution (i'm not sure what kind of head the P900 uses, maybe it's in multiples of 300 now?). That should give you the highest ink load possible without needing a RIP, and allow you to get more saturation out of those mixtures.