Printer calibration dilemma

[crimbfighter]

So I have been printing at home for a few years with my Canon Pro-100 but always struggle with getting my prints to match my calibrated monitor. After trying virtually every trick known to the internet, I have settle on the fact that I need to calibrate my printer, well more accurately, I need use custom ICC profiles. That leads me to my dilemma. I currently calibrate my monitor with a Spyder Pro 4. I could add the SpyderPrint to my workflow for a reasonable price, but I'm consistently seeing people online saying they got better results with the Colormunki Photo, which is marginally more expensive. I also thought about just sending out for custom profiles but if I got 10 profiles and I'm at the same price as just buying the calibration tool, and I'd have it for future use.

If I go with the Colormunki, that doubles as a monitor calibrator, so I could get rid my Spyder 4.

Has anyone had experience with one or both? The SpyderPrint or the Colormunki Photo for printer calibration?

 

[Jeff15]

I have been using a Spyder for a few years now, I am very pleased with the result....

 

[crimbfighter]

I have been using a Spyder for a few years now, I am very pleased with the result....

Thanks. I assume you mean the Spyderprint? Would you say that your prints matched your edit very closely after calibration?

 

[mrca]

You are trying to match a back lit monitor with a paper that reflects light. Also, the image may exceed the gamut of the paper you are using. I recently switched from epson exhibition fiber to legacy baryta and there is a marked difference in the image tonal range and it closely matches my monitor. You don't specify how it doesn't match but if it is a brighness issue, that goes to the fact that most monitor are really bright so we can easily see on them. That results in a darker print. You could take down the brighness of the monitor but then it is harder to use. If you use lightroom, the print module has a section for kicking up brighness and contrast. Are you doing test prints on cheaper paper. I like to print a half page print on 8.5 x 11 less expensive paper to dial in color/brightness. I can crop a small section of the image and process a half page of it to check sharpening. After printing on half the page, you can run it through again and print on the other half.

 

[Ysarex]

I haven't been able to compare unfortunately. I do use Colormunki Photo for printer calibration and I'm real happy with the results. I used to use an XRite eye1 and might have been even happier with that, what I miss in the Colormuniki Photo software is an option to hand-tune the profile.

Joe

 

[Jeff15]

Every month I use my Spyder 5 Express and I have no problems at all..

 

[Ysarex]

Every month I use my Spyder 5 Express and I have no problems at all..

Tanner is asking about printer profiling with a spectrophotometer -- different products -- he's asking about one of these: Datacolor SpyderPRINT

compared with one of these: http://xritephoto.com/colormunki-photo

Joe

 

[TCampbell]

I do own a ColorMunki Photo. It’s an older one ... I’ve had it for ... probably at least 10+ years. I notice my monitors get pretty accurate color. But I’ve color calibrated my printers and I notice the default ‘automatic’ color management the printer does was doing a better job than the ICC profile I created. This made me wonder if the age of the tool is effecting it’s accuracy. I honestly don’t know.

I considered buying the newer and improved X-Rite i1 Studio to see if it is better.


For a whole lot of reasons, a printer (even a calibrated printer) isn’t going to provide colors that necessarily match your monitor. Remember that paper reflects light. The monitor emits light. As you take a print from room to room, it’s at the mercy of whatever the quality of light is in that room. If the light is a neutral daylight-balance light, then you’re going to get a color cast based on that light.

Also, the color gamut of printer/paper/ink is generally a lot less than the gamut of the monitor (which is less that the gamut of the world around you). You’ll notice that as the price tag of the photo printer goes up, the number of tanks increases along with it... and this is because it generally improves the color gamut possible.

Speaking of color gamut... since invariably the printer can’t produce every possible color you want, there’s a question about how you want it to handle the colors it can’t accurately produce. In the printer software this is usually referred to as the “rendering intent” (often just “intent”) and the choices are usually “relative colorimetric” (often just stated as “relative”) and “perceptual”.

“Relative colorimetric” tries to produce as many colors as possible to be accurate. Out-of-gamut colors are shifted to the closest color that the printer can reproduce. But if you’ve got an image with two hues (say they are reds - say the subject is wearing a red dress) and these are adjacent colors ... the printer can accurately produce one, but not the other. Those colors should appear slightly different. And because the printer can accurately produce one, it does. And since it cannot accurately produce the other, it shifts to the nearest reproduceable color ... and those two hues might “print” as the identical color even though they *should* have looked a bit slightly different. That loss of tonal difference (between hues that *should* appear different) can reduce the overall look of the image (as if you lost detail ... the camera had the detail but the printer can't reproduce it).

“Perceptual” tries to preserve the natural relationship between the colors ... not the accuracy of the colors themselves. So in the previous red-dress example, the printer would shift the non-producable color to something it can produce ... but since that color is identical to a nearby color, it shifts that color as well... so that the relationship appears natural.

“Relative” will give you as many accurate colors as possible... at the cost of not preserving the relationship between colors.

“Perceptual” will preserve the relationship... at the cost of not having perfectly accurate color.

Which is better depends on what you are printing (you may have to produce test prints to decide).

When trying to check your printer for color accuracy, you should probably select "Relative Colorimetric" for such tests (because that's the one where the printer will try to maximize color accuracy for all in-gamut colors.)

 

[mrca]

Tim does an excellent job of summarizing the issues in getting a print to "look like" the monitor. Intensity and color of the light the print is viewed under plays a large roll. Try viewing the print under window light, ambient room light, incondescent and florescent. So one consideration in the print is the type of light where the print will be displayed. Also, perhaps making the print, rather than the monitor, match your vision.

 

[crimbfighter]

You are trying to match a back lit monitor with a paper that reflects light. Also, the image may exceed the gamut of the paper you are using. I recently switched from epson exhibition fiber to legacy baryta and there is a marked difference in the image tonal range and it closely matches my monitor. You don't specify how it doesn't match but if it is a brighness issue, that goes to the fact that most monitor are really bright so we can easily see on them. That results in a darker print. You could take down the brighness of the monitor but then it is harder to use. If you use lightroom, the print module has a section for kicking up brighness and contrast. Are you doing test prints on cheaper paper. I like to print a half page print on 8.5 x 11 less expensive paper to dial in color/brightness. I can crop a small section of the image and process a half page of it to check sharpening. After printing on half the page, you can run it through again and print on the other half.

Thanks for the reply. I'm aware of the difference between the back lighting of the monitor and how it's different than the environmental lighting on a print. I actually calibrate my monitor to the ambient room brightness so I know my monitor is not too bright. The prints do not match in terms of color reproduction, and it's obviously an issue with the ICC profiles not being a perfect match for my printer. Despite using Canon paper, a Canon printer, and the canon ICC profiles, every printer will work slightly differently than the one before it and the one after it off the same production line. As a result, the best, and only sure way, to make the prints match the colors is to use a custom ICC profile that profiles your printer to your monitor. The closest I've been able to get is to use a plug-in by Canon called Print Studio Pro which prints out test pages and you make selections off that to tell the software how to compensate for color, contrast, etc. However, it's too subjective because it's based on the eye's observations and interpretations. Hence the need for an actual measuring device to read what your printer is actually printing when it thinks, for example, it's printing red. What it thinks it's printing as red might actually not be red so it can measure and compensate appropriately. I am also looking to profile my printer to eliminate exactly what you're describing. Printing test page after test page wastes ink, paper, and adds to my level of frustration. When using a properly profiled printer for a given paper, the first print should, in theory, be a good match for what you see on the monitor. This of course is also done in conjunction with soft proofing, which I already do so that I can compensate for out of gamut colors and make adjustments to my shadows, highlights, and contrast to match the image to the capabilities of the printer. At the end of the day, in order to overcome my problems, I need to be profiling my printer to each type of paper I use.

I haven't been able to compare unfortunately. I do use Colormunki Photo for printer calibration and I'm real happy with the results. I used to use an XRite eye1 and might have been even happier with that, what I miss in the Colormuniki Photo software is an option to hand-tune the profile.

Joe

Thanks for the info! I didn't think about being able to adjust the profile. I bet I could use the Canon plug-in and that would allow me to make any adjustments at print to compensate for anything.

Every month I use my Spyder 5 Express and I have no problems at all..

Tanner is asking about printer profiling with a spectrophotometer -- different products -- he's asking about one of these: Datacolor SpyderPRINT

compared with one of these: http://xritephoto.com/colormunki-photo

Joe

You are correct, I am only asking about the printer calibration. The upside to spending the money on the Colormunki would be that it doubles as a screen calibrator so I could get rid of my Spyder Pro 4. Thanks for the comment!

I do own a ColorMunki Photo. It’s an older one ... I’ve had it for ... probably at least 10+ years. I notice my monitors get pretty accurate color. But I’ve color calibrated my printers and I notice the default ‘automatic’ color management the printer does was doing a better job than the ICC profile I created. This made me wonder if the age of the tool is effecting it’s accuracy. I honestly don’t know.

I considered buying the newer and improved X-Rite i1 Studio to see if it is better.


For a whole lot of reasons, a printer (even a calibrated printer) isn’t going to provide colors that necessarily match your monitor. Remember that paper reflects light. The monitor emits light. As you take a print from room to room, it’s at the mercy of whatever the quality of light is in that room. If the light is a neutral daylight-balance light, then you’re going to get a color cast based on that light.

Also, the color gamut of printer/paper/ink is generally a lot less than the gamut of the monitor (which is less that the gamut of the world around you). You’ll notice that as the price tag of the photo printer goes up, the number of tanks increases along with it... and this is because it generally improves the color gamut possible.

Speaking of color gamut... since invariably the printer can’t produce every possible color you want, there’s a question about how you want it to handle the colors it can’t accurately produce. In the printer software this is usually referred to as the “rendering intent” (often just “intent”) and the choices are usually “relative colorimetric” (often just stated as “relative”) and “perceptual”.

“Relative colorimetric” tries to produce as many colors as possible to be accurate. Out-of-gamut colors are shifted to the closest color that the printer can reproduce. But if you’ve got an image with two hues (say they are reds - say the subject is wearing a red dress) and these are adjacent colors ... the printer can accurately produce one, but not the other. Those colors should appear slightly different. And because the printer can accurately produce one, it does. And since it cannot accurately produce the other, it shifts to the nearest reproduceable color ... and those two hues might “print” as the identical color even though they *should* have looked a bit slightly different. That loss of tonal difference (between hues that *should* appear different) can reduce the overall look of the image (as if you lost detail ... the camera had the detail but the printer can't reproduce it).

“Perceptual” tries to preserve the natural relationship between the colors ... not the accuracy of the colors themselves. So in the previous red-dress example, the printer would shift the non-producable color to something it can produce ... but since that color is identical to a nearby color, it shifts that color as well... so that the relationship appears natural.

“Relative” will give you as many accurate colors as possible... at the cost of not preserving the relationship between colors.

“Perceptual” will preserve the relationship... at the cost of not having perfectly accurate color.

Which is better depends on what you are printing (you may have to produce test prints to decide).

When trying to check your printer for color accuracy, you should probably select "Relative Colorimetric" for such tests (because that's the one where the printer will try to maximize color accuracy for all in-gamut colors.)

Thanks for the detailed response! That's an interesting observation you made about your older unit. From what I have read in reviews, the actual spectrophotometer used in the Colormunki is very good, and has remained unchanged for years, and it mostly comes down to the software that has benefited from improvements over the years. Have you been able to update/upgrade your software?

I always soft proof my images (well, images I care about, anyway) that I am printing so I can see when colors, contrast, brightness, etc. may be outside the color gamut or capabilities of the printer and medium. However, even soft proofing will fall short when the ICC profile the computer uses doesn't match what the printer is physically capable of printing. Also, for clarification, I use I believe 5600K bulbs in my office lighting and block outside light, so the lighting is consistent. I know the temperature of light that falls on the print drastically alters how the colors are perceived by the eye, so all of my printing is evaluated under the same light as the monitor. Then, if it look different under different light, I at least know it matched how my original edit looked. I also do use relative colorimetric for my prints. To me it produces the best results for what I do.

Tim does an excellent job of summarizing the issues in getting a print to "look like" the monitor. Intensity and color of the light the print is viewed under plays a large roll. Try viewing the print under window light, ambient room light, incondescent and florescent. So one consideration in the print is the type of light where the print will be displayed. Also, perhaps making the print, rather than the monitor, match your vision.

I agree that adjusting the print to match the light of it's final destination is important, if for example it will hang in a gallery with gallery lights. As I mentioned above, I evaluate my prints under the same controlled lighting as the monitor is calibrated under. I find that when selling prints, one can't possibly control how the buyer will display them. Consequently, I make them the best I can and control the things that I can, such as making it match (or at least very closely) the image I created on my monitor. Thank you for your reply!

 

[mrca]

I include an in house design consult with clients that warrant it. It gives me a chance to evaluate potential areas for wall prints, to photo those areas for projection sales that shows them the photos virtually on those walls encouraging larger prints and to understand the lighting in each area. Most folks don't have the finely trained eye to know the difference. I expect even as the maker of the image you have to hold the print beside the monitor to see some of the difference. I print on epson and epson legacy baryta comes out of my printer a near perfect match to the monitor and even has better tonal range than my former favorite, exhibition fiber paper.

 

[freixas]

I agree that adjusting the print to match the light of it's final destination is important, if for example it will hang in a gallery with gallery lights.

One can calibrate a printer for accurate color reproduction, but is there a way to "calibrate" a printer for lighting? I suspect you can't do it with ICC profiles.

It would be nice if Photoshop or whatever had a tool in which you could enter the lumens (or maybe you aim a camera at the wall and enter the meter reading) at the anticipated display location and have the tool simulate what a given image would look like so you could then adjust the levels. Does anyone know of any such tool or does everyone do it by trial-and-error (or maybe by experienced guessing)?

I've never worked with a ColorMunki Photo, but I'm guessing it has an internal light source for calibrating prints, so you wouldn't be able to calibrate the test prints at the anticipated print viewing location and expect to get a different profile.

 

[crimbfighter]

I include an in house design consult with clients that warrant it. It gives me a chance to evaluate potential areas for wall prints, to photo those areas for projection sales that shows them the photos virtually on those walls encouraging larger prints and to understand the lighting in each area. Most folks don't have the finely trained eye to know the difference. I expect even as the maker of the image you have to hold the print beside the monitor to see some of the difference. I print on epson and epson legacy baryta comes out of my printer a near perfect match to the monitor and even has better tonal range than my former favorite, exhibition fiber paper.

That is certainly more than I would be willing to go through for a print sale, but my income is also not reliant on my photography. I do see a great benefit, though, to your approach as it would ensure the best possible outcome when a print is hung. I have asked where a person intends to hang an image and viewed the print under similar lighting to make sure I was happy with it before sending it out. However, I think that unless it's in a gallery setting or being illuminated with gallery specific lighting, the tendency for people to change bulb types from incandescent to LED to CFL based on what's cheaper at Home Depot makes it virtually impossible to ensure it will be viewed under the same lighting indefinitely. Bottom line, I make sure it looks good under daylight and call it a day. Though I will make adjustments if I know it will be under a specific type of lighting. I did notice, though, the new X-Rite i1 Studio (the refreshed version of the X-Rite Photo) does include an option to measure different ambient light to adjust the ICC profile to be optimized for different lighting conditions. That might take all the guess work out of it for me! I also like your system of projecting a print on the wall to demonstrate the impact of a larger piece. That's a brilliant idea! I can't speak for Epson, but I definitely do not have that same experience with the Canon printers. Though I wonder if on the higher level professional printers they are more accurate from the factory.

I agree that adjusting the print to match the light of it's final destination is important, if for example it will hang in a gallery with gallery lights.

One can calibrate a printer for accurate color reproduction, but is there a way to "calibrate" a printer for lighting? I suspect you can't do it with ICC profiles.

It would be nice if Photoshop or whatever had a tool in which you could enter the lumens (or maybe you aim a camera at the wall and enter the meter reading) at the anticipated display location and have the tool simulate what a given image would look like so you could then adjust the levels. Does anyone know of any such tool or does everyone do it by trial-and-error (or maybe by experienced guessing)?

I've never worked with a ColorMunki Photo, but I'm guessing it has an internal light source for calibrating prints, so you wouldn't be able to calibrate the test prints at the anticipated print viewing location and expect to get a different profile.

Note what I mentioned above, the new X-Rite i1 Studio does seem to allow for the creation of different ICC profiles based on it's measurements of different ambient lighting conditions. That's a really attractive feature. I don't think that's a feature that's available in other consumer level calibration tools. Short of using a feature like this, I think it requires using something like the Canon Print Studio plug-in that generates thumbnail images of different color balances, brightness, and contrast and you pick the ones that best match your monitor and it inputs adjustments to the colors and contrast. You could view it under the different lighting conditions and make the best picks based on how that light effects it.

 

[crimbfighter]

I actually pulled the trigger tonight and bought the new X-Rite i1 Studio, which was the replacement for the Colormunki Photo. The differences are more so in the software, but they are having a sale this weekend of $100 off so it was time. I'll post a follow up with how it works out.