Luminous Landscape Forum
Raw & Post Processing, Printing => Colour Management => Topic started by: MikeFletcher on July 18, 2010, 04:49:12 pm
-
Im curious to hear what everybody is profiling/calibrating their screens to, especially if youre using modern papers and pigment ink printing.
It has been my experience that with the inkjet papers we have today, some old golden rules just dont apply anymore, or never applied to us (only to cmyk proofing?)
There is no way i can match a print under D50 light (i still prefer daylight) with my Eizo CG calibrated to 5000k or better 5800k, 80cd/m2 oder 120cd/m2. Especially the low luminance settings arent convincing, you get a wrong impression of the white temperature and highlights compared to the paper print. Ugra suggests 160cm/m2 which works alot better. Ugra also suggests 5800k since its a better match to D50 light than calibrating your screen to D50 (it is!). However with the papers i use (EFF and Canson Baryta), 5800k is still too warm and i have to get all the way up to 6500k. That way my screens are extremly close to my prints under daylight and D50.
So i use 160cd/m2 and 6500k and you can hardly see any difference between the screen and print. 6000k is already obviously too warm, i even feel that 6500k tends to be on the warm side of neutral sometimes compared to the print but in the end it always works out. I wonder what works for you
Mike
ps: Screen and printer (epson 3880) profiles were done with an i1pro.
-
I assume you view your prints (gallery, walls, office) at these print screen match levels (160cd/m2 and 6500k), or else your match has no real basis for your "standard"!
Jpegman
Im curious to hear what everybody is profiling/calibrating their screens to, especially if youre using modern papers and pigment ink printing.
It has been my experience that with the inkjet papers we have today, some old golden rules just dont apply anymore, or never applied to us (only to cmyk proofing?)
There is no way i can match a print under D50 light (i still prefer daylight) with my Eizo CG calibrated to 5000k or better 5800k, 80cd/m2 oder 120cd/m2. Especially the low luminance settings arent convincing, you get a wrong impression of the white temperature and highlights compared to the paper print. Ugra suggests 160cm/m2 which works alot better. Ugra also suggests 5800k since its a better match to D50 light than calibrating your screen to D50 (it is!). However with the papers i use (EFF and Canson Baryta), 5800k is still too warm and i have to get all the way up to 6500k. That way my screens are extremly close to my prints under daylight and D50.
So i use 160cd/m2 and 6500k and you can hardly see any difference between the screen and print. 6000k is already obviously too warm, i even feel that 6500k tends to be on the warm side of neutral sometimes compared to the print but in the end it always works out. I wonder what works for you
Mike
ps: Screen and printer (epson 3880) profiles were done with an i1pro.
-
I assume you view your prints (gallery, walls, office) at these print screen match levels (160cd/m2 and 6500k), or else your match has no real basis for your "standard"!
Jpegman
I view them under D50 light (just normlicht) that has around 300-400 lux and throughout the day in all kinds of daylight (morning, evening, afternoon, cloudy, sunny etc). I base my judgement on these situations. The way the screen brightness affects the highlights obviously doesnt transfer to paper. A print looks fine at something close to 80cd/m2 but a screen at that luminance mutes the highlights and the whites. Besides that to really match the brightness of a print, you have to go high with screen brightness. Calibrate your screen to your papers whitepoint and try make the paper white dissapear infront of the screen's white.. you'll probably end up with a cd/m2 value of around 160-240 in a very dimmed room.
-
The correct value for white point and luminance are those that produce a visual match. Everyone’s is different (everyone is using different print viewing conditions next to the display).
(http://digitaldog.net/files/Print_to_Screen_Matching.jpg)
-
The correct value for white point and luminance are those that produce a visual match. Everyone’s is different (everyone is using different print viewing conditions next to the display).
This is exactly why i started this thread, i wasnt going to tell everyone my way is the way it works (sorry if i came off this way). Im curious what works for each one of you.
Besides that, i think we all have the same daylight: sunny, cloudy, sunset, sunrise, bluesky etc. For example under no daylight condition i could ever get to match anything to a D50 calibrated screen. Maybe if were living on the sun this would work?
I think your picture shows what i mentioned, the luminance of the tft still is too low to give a good match especially in the highlights and upper mid tones. But it could also be the effect of your camera. For me it's not only important to match the colors (that's the easy part!) but also to match the "impression" you get when watching the print... hope this doesnt come off too voodooish That would be the main reason i re-print. Colors are easy.. but too much saturation... too low contrast..., things that impact the emotions an image should portray. To translate these things well from screen to print it's important to match the luminance to get an overall picture of how the print will look. Or maybe i'm a crazy perfectionist
-
I think your picture shows what i mentioned, the luminance of the tft still is too low to give a good match especially in the highlights and upper mid tones.
If the tft is too low, raise it (or raise/lower the viewing booth illuminant).
-
I love this conversation, and I love coming back to it again and again as we seem to do, because even after reading all the responses over and over again, I never can quite figure out exactly what the answer is — or even exactly what y'all are saying the answer is.
Here, for example, the illustration says "Booth dimmed 50% to match LCD" and "LCD with proper luminance calibration ... to match booth." So which is it? Do we change the luminance of the monitor to match the print (in a given viewing light), or do we change the print viewing/assessment light to match the monitor?
The part that buffaloes me particularly is that I can never make sense of the notion of changing the light on the print in the assessment context to match the monitor. It seems to me that, except perhaps in galleries or in our own controlled environments, prints can rarely be expected to be viewed by our customers in precisely the same light that we choose to evaluate them in. That suggests to me in turn that we should choose print assessment lighting environments that come as close as possible to our reasonable predictions of the range of lighting that we expect the print ultimately to find a home in, or at least that come as close as possible to letting us make a good judgment of whether the print is going to hold up and be successful in that range of environments.
Now it strikes me that for the most part this ought to be a reasonably standard color temp and lighting level. I think of the example of magazine covers. They're viewed in everything from dark waiting rooms to harsh airport newstand lighting — but how often do you find yourself looking at a magazine cover and thinking "They printed that too dark" or "Wow, Newsweek is really looking washed out and reddish this week." This tells me that there must be a brightness and color temperature sweet spot (or at least a very narrow range) out there for print assessment lighting that is simply right[/b]. Not right for you, or right for me, but just right, period.
And so I conclude that we don't — ever — change our print viewing light to match our monitors. We do change it to get to that sweet spot, so that we can have some level of confidence that the judgment we make about the print in that light is going to hold up when the print arrives at, and basks in the light of, its final destination. And, having found that sweet spot for viewing the print, then we adjust the luminance of the monitor[/b] to match the print.
Throw into this mix the fact that our subjective judgment of the image on the monitor is influenced by the ambient light in our respective work environments, and you're back to having a wide variation in "correct" monitor luminance settings, but a very narrow range of "correct" lighting for assessing prints.
No?
Nill
-
Here, for example, the illustration says "Booth dimmed 50% to match LCD" and "LCD with proper luminance calibration ... to match booth." So which is it? Do we change the luminance of the monitor to match the print (in a given viewing light), or do we change the print viewing/assessment light to match the monitor?
Either or both. Doesn't matter as long as the dimming on the booth doesn't affect its color (not the case on the GTI box so dim away).
It seems to me that, except perhaps in galleries or in our own controlled environments, prints can rarely be expected to be viewed by our customers in precisely the same light that we choose to evaluate them in.
Doesn't matter. Outside the screen to print reality (a screen and viewing booth next to each other), your eyes will adapt to other conditions. The key here is screen to print matching (WYSIWYG).
Now it strikes me that for the most part this ought to be a reasonably standard color temp and lighting level.
Nope. For one, all displays are different (produce differing white points). Everyone has differing viewing conditions next to the display. How can there be “standard”? You can purchase a dozen illuminants that claim the are D50 or CCT6500K and they are all different, all producing differing color of white.
I think of the example of magazine covers. They're viewed in everything from dark waiting rooms to harsh airport newstand lighting — but how often do you find yourself looking at a magazine cover and thinking "They printed that too dark" or "Wow, Newsweek is really looking washed out and reddish this week."
You don’t unless they really did (rare) and because your eyes adapt to the condition, you accept what you now see is being OK. And since the display is 100 miles away, in your office, its impossible to say it matches that display but then it doesn’t matter. Its 100 miles away!
And so I conclude that we don't — ever — change our print viewing light to match our monitors.
Go for it. You got a match? If so fine. If not, you have to adjust one or both of those items to produce a match.
Throw into this mix the fact that our subjective judgment of the image on the monitor is influenced by the ambient light in our respective work environments, and you're back to having a wide variation in "correct" monitor luminance settings, but a very narrow range of "correct" lighting for assessing prints.
That is WHY we control the ambient light conditions.
-
...Doesn't matter. Outside the screen to print reality (a screen and viewing booth next to each other), your eyes will adapt to other conditions. The key here is screen to print matching (WYSIWYG).
...
If that is true, then why is the most common complaint heard on printing forums "My prints come out looking too dark?" It seems like the first problem the average Joe or Jane getting their feet wet in color management and trying to get a decent print match bump up against is the "prints too dark" problem. And in most cases, it's not just "My prints are darker than my monitor," but "My prints are too dark." And for those folks, isn't the most typical correct solution not "turn up your room lighting" (much less "stare at it a while longer and your eyes will adjust"), but "turn down your monitor luminance... it comes from the factory set to look good in office lighting or on the shelf at Best Buy, not in your den or home studio."
...and:
That is WHY we control the ambient light conditions.
Hmmm... easy to say, and perhaps easy to do if you work in a room without windows... ;-)
Nill
-
If that is true, then why is the most common complaint heard on printing forums "My prints come out looking too dark?"
Because out of the box, LCDs are way too bright! And people don’t account for the display target luminance they shoot for assuming they even do calibrate the displays, FOR the print viewing conditions. The fix for “my prints are too dark” is to RAISE the print viewing conditions. There is only so low a new LCD can be driven. If you’re unlucky enough to have an older iMac, you can’t even get close to dimming down the display.
And in most cases, it's not just "My prints are darker than my monitor," but "My prints are too dark."
You have to ask them. Are the prints dark no matter where you view them? 99 times out of 100, they say no (because it IS possible to make a dark print). 99 times out of 100, they tell me the prints look OK in the kitchen or outside. Bingo, the display is too bright (or the viewing booth is too dim), or both. The fix is easy.
-
Hmmm... easy to say, and perhaps easy to do if you work in a room without windows... ;-)
There is a relatively new invention you should look into for this issue. Curtains (or room darkening blinds).
-
I understand. Some of us have not only windows but also wives. ;-)
But what about the frequent "my prints are too dark" conundrum? Do you not agree that it's most often due to a too-bright monitor?
And do you not agree that this is a problem of prints falling outside[/i] the range of what our eyes will reasonably adjust to and consider "right?"
And if there's not a fairly narrow range of "right," then how come some of the magazine covers on the newstand shelf don't look brighter or darker than the ones sitting right next to them?
Nill
p.s. Whoops, wrote all that before I read the earlier of your two replies, in which it appears you are agreeing with me at least in part.
-
But what about the frequent "my prints are too dark" conundrum? Do you not agree that it's most often due to a too-bright monitor?
Or too dim a print viewing station.
And do you not agree that this is a problem of prints falling outside[/i] the range of what our eyes will reasonably adjust to and consider "right?"
No. There’s simply a mismatch between print and display luminance.
-
The fix for “my prints are too dark” is to RAISE the print viewing conditions.
This confuses a lot of people and it's easy to mistake what is going on. It is possible to make a print too dark and if you raise the lighting conditions it is not going to make that existing print better. The key to understanding this is to really understand the difference between brightness and lightness. Lightness is defined as brightness relative to the brightness of white. Consider printing a grey square on a white piece of paper. When you look at it in your office it looks grey. Take it outside it will still look grey despite the fact that it is reflecting several orders of magnitude more light. In other words it is much brighter, but the lightness has stayed more or less the same because the relationship between white and grey has remained constant. Another good example is looking at a white wall in a dim room. It looks white. But if you take a slide projector and project and image of a black text on a white background, the text will look black—yet it's pretty much exactly the same brightness as the white wall was before you turned on the projector. You change what is considered white and now the previously white wall looks black. Prints won't change that much going from one lighting environment to another even though the absolute measure of light reflecting from the print may change dramatically—direct sun to dark gallery—because we want to match the relative values like lightness. If you are being picky or have very special needs you will need to understand that luminance levels do change our perception of color but it's much more subtle than we are usually talking about when trying to get decent prints.
So taking your dark print and viewing it under brighter conditions isn't going to help much. But changing the viewing conditions of the whole working environment will change the relationship between monitor's lightness and the print's. Your monitor is going to look dimmer in a brighter environment (It's also going to get more glare reducing contrast, which is why hoods are helpful). It's also going to appear dimmer when switching your gaze between it and a brighter viewing booth. So if you can't target a low enough luminance with you monitor then raising the light levels of the booth and environment will be a good option.
It's confusing, but Andrew is right on. You are dealing with two light sources when comparing prints and monitors—you just need to make them match visually—your eyes and brain do the rest. When matching colors we are matching relative values. We don't need to match brightness, but rather lightness and since lightness is a measure relative to white, we need to make sure our whites match as closely as possible. This can involve changing the luminance of the monitor or changing the luminance of the print viewing area. If you transport that print to another viewing environment your eyes will adjust for the white and luminance and any changes in the print should be subtle and within most people's tolerances, just like magazines the dentist's foyer.
-
It is possible to make a print too dark and if you raise the lighting conditions it is not going to make that existing print better.
Yes, if the print is too dark (just about everywhere), raising the lighting conditions will not help. That’s why you have to ask the user, “are the prints too dark everywhere?” and most of the time they tell you no. If it is too dark everywhere, you provide them a good reference image that has known qualities, not continue testing output using their images. They output the reference image and if its still too dark, we have a totally different problem to fix. But most of the time, users tell me that’s not their problem. Its just usually a huge disconnect between the display and whatever odd-ass lighting they are using next to the display.
-
Now that's a step I have neglected. Where might we find such reference images (preferably free or something I already have... I vaguely recall perhaps seeing one in the C1 install...)?
Nill
-
Now that's a step I have neglected. Where might we find such reference images (preferably free or something I already have... I vaguely recall perhaps seeing one in the C1 install...)?
http://www.digitaldog.net/files/Printer_Test_file.jpg.zip (http://www.digitaldog.net/files/Printer_Test_file.jpg.zip)
http://homepage.mac.com/billatkinson/FileSharing2.html (http://homepage.mac.com/billatkinson/FileSharing2.html)
-
Thanks!
Nill
-
The correct value for white point and luminance are those that produce a visual match. Everyone’s is different (everyone is using different print viewing conditions next to the display).
(http://digitaldog.net/files/Print_to_Screen_Matching.jpg)
I'm curious as to whether you find the colors on your wide gamut LCD or standard gamut LCD more closely match what you see on your CRT.
I have to say that the color look awfully different to me between each display type even if calibrate them all with the same probe.
I think the fact that some miss and some hit sRGB primaries and even moreso the different spectral characters of the native primaries of each make things look differently the eye.
At least if i compare wide gamut to wide gamut and standard to standard things look reasonably, if not exactingly similar. But it seems like wide gamuts and standard gamuts and perhaps CRTs and probably oled, not too mention scenes in the real world, etc. all have a rather different balance and apparent look to the eye. It doesn't seem so good if you want to edit your images once and then have them look the same on any display.
Maybe a 1nm spectrophotometer would make everything look the same? Or maybe it still wouldn't quite....
-
My solution to this problem is somewhat pragmatic, but seems to work for me. First, I profiled my monitor using my Spyder 2 puck. Then I measured the luminance of the monitor at one foot distance using a white screen (Windows Explorer set full-screen on an empty folder). I measured it using my trusty Weston Master V meter on the low-light setting. The final step was to adjust the print viewing area light to give the same reading off a sheet of white printing paper at the same one foot distance.
The monitor and my prints seem to match really well.
John
-
I'm curious as to whether you find the colors on your wide gamut LCD or standard gamut LCD more closely match what you see on your CRT.
I have to say that the color look awfully different to me between each display type even if calibrate them all with the same probe.
I think the fact that some miss and some hit sRGB primaries and even moreso the different spectral characters of the native primaries of each make things look differently the eye.
At least if i compare wide gamut to wide gamut and standard to standard things look reasonably, if not exactingly similar. But it seems like wide gamuts and standard gamuts and perhaps CRTs and probably oled, not too mention scenes in the real world, etc. all have a rather different balance and apparent look to the eye. It doesn't seem so good if you want to edit your images once and then have them look the same on any display.
Maybe a 1nm spectrophotometer would make everything look the same? Or maybe it still wouldn't quite....
see here... first PDF on the top of the list ("Color Matching Between sRGB Monitors and Wide Color Gamut Monitors"):
http://www.eizo.com/global/support/wp/index.html (http://www.eizo.com/global/support/wp/index.html)
-
Where might we find such reference images (preferably free or something I already have... I vaguely recall perhaps seeing one in the C1 install...)?
you can create one by yourself based on the Gretag Color Checker Color patches.
For monitor calibration - i.e. to get a visual match of luminance and white point - I am finding it easier to start with "white" only, resp. with white and b/w shades. Color patches or "real world" color images may affect your perception of "white".
In the second step it's fine to compare to colors...
Of course, there's nothing wrong with other reference images... just tought to show a slightly different approach.
Attached a screenshot of my custom made reference images.
[attachment=23241:calchart_bw.jpg] [attachment=23240:calchart.jpg]
-
I beg to differ on the usefulness of test images or colour charts or whatever. The best comparitive device of all is a fine photograph with a long tonal range.
Pick out one of your own photographs which has an excellent black, some just-visible shadow detail, an extended tonal range through the mid-tones, and some delicate highlight detail (clouds are good). Choose a picture which you know well and have printed many times in the past. Print it with a wide (3/4 inch) border so that you have paper white as a constant reference.
Then when you have to set up a new monitor or printer or paper you are dealing with an old friend, and one which you are so familiar with that any weak link in the chain will be self-evident.
John
-
I beg to differ on the usefulness of test images or colour charts or whatever. The best comparitive device of all is a fine photograph with a long tonal range.
Pick out one of your own photographs which has an excellent black, some just-visible shadow detail, an extended tonal range through the mid-tones, and some delicate highlight detail (clouds are good). Choose a picture which you know well and have printed many times in the past. Print it with a wide (3/4 inch) border so that you have paper white as a constant reference.
Then when you have to set up a new monitor or printer or paper you are dealing with an old friend, and one which you are so familiar with that any weak link in the chain will be self-evident.
I wouldn't say one way is "better" than the other. Simply find out what works best for your personally... as long as you follow the basic principles outlined by "digitaldog" above. Working with a well known print/image is also great - in particular with a white border as reference (which works fine for print related editing in any case... IMO).
Me I also adjust the 6 color-setting in the monitor hardware to match the color reproduction of my favorite paper (in a second profile). So for me the color patches work very well.
-
Keith Cooper has a link to a wide variety of test images here (http://www.northlight-images.co.uk/article_pages/test_images.html). I like to use his B&W test print and the one Jack Flesher created from some of the Bill Atikinson images (this one is good as there is a large black square in one corner for measuring Dmax) as a start and then move to two of my standard prints (one for B&W and one for color) to make a final judgment on any new paper/ink setting combination.
Alan
-
This whole question of screen to print match has confused me for some years. Like Nill, I have believed you have to match the monitor to the viewing conditions where the print will be displayed. I think now I am starting to see the light (or should that be brightness?)
But based on the above comments, can I conclude that it doesn't matter under what light you view the print in your "booth" (temp and luminosity), as long as it matches your display, the print will look the same no matter where it is displayed (museum, gallery, kitchen or toilet)??
If so, why is it necessary to spend so much money on "booth lighting"? As long as I get a "match" can't I just use a 40W incandescent bulb in a rampant dolphin holder with peacock blue shade (as long as it it shielded from the display?)??
Thanks for any illumination!
-
But based on the above comments, can I conclude that it doesn't matter under what light you view the print in your "booth" (temp and luminosity), as long as it matches your display, the print will look the same no matter where it is displayed (museum, gallery, kitchen or toilet)??
If so, why is it necessary to spend so much money on "booth lighting"? As long as I get a "match" can't I just use a 40W incandescent bulb in a rampant dolphin holder with peacock blue shade (as long as it it shielded from the display?)??
Steve
There is a very important consideration with your booth (or room) lighting, and that is one of print metamerism. Any old bulb won't necessarily do. I use tungsten lighting in my studio because I prefer its natural warmth, but it does not enable me to easily judge very subtle hue shifts. I work solely in B/W, and the Epson printers in ABW mode can and do wander a bit in absolute hue as the ink cartridges get low. Under tungsten it is very hard for me to see these shifts (usually towards blue-green), however daylight will reveal them quite mercilessly. As I print mostly in the evenings, it is only too easy to find out next day that the print metamerises quite badly under daylight. So now I have a 6500 kelvin light in another room where I do my print trimming, and a calibrated grayscale card to check the print against. Once you see a shift under the 6500 lamp, it is easy enough to compensate in the ABW mode by (say) adding +1 red or whatever. I would imagine the same thing would apply to colour printing, too.
John
-
Thanks for any illumination!
http://www.fogra.org/products-en/download/Extra18web.pdf (http://www.fogra.org/products-en/download/Extra18web.pdf)
http://en.wikipedia.org/wiki/Metamerism_%28color%29 (http://en.wikipedia.org/wiki/Metamerism_%28color%29)
http://www.just-normlicht.de/uk/shop/00000116.html (http://www.just-normlicht.de/uk/shop/00000116.html)
-
But based on the above comments, can I conclude that it doesn't matter under what light you view the print in your "booth" (temp and luminosity), as long as it matches your display, the print will look the same no matter where it is displayed (museum, gallery, kitchen or toilet)??
Not completely. Not all man made illuminants are created equally. I use Fluorescent boxes like the GTI seen above because I work with others who use similar booths. The dimmer is nice as it allows control without affecting color. Fluorescent lights can produce issues when viewing some printed materials with high OBA content. They have a spiky spectrum. Solux are preferable but the downside is control and heat. You can’t dim them without affecting the color. But they produce beautiful results.
-
Steve
There is a very important consideration with your booth (or room) lighting, and that is one of print metamerism. Any old bulb won't necessarily do. I use tungsten lighting in my studio because I prefer its natural warmth, but it does not enable me to easily judge very subtle hue shifts. I work solely in B/W, and the Epson printers in ABW mode can and do wander a bit in absolute hue as the ink cartridges get low. Under tungsten it is very hard for me to see these shifts (usually towards blue-green), however daylight will reveal them quite mercilessly. As I print mostly in the evenings, it is only too easy to find out next day that the print metamerises quite badly under daylight. So now I have a 6500 kelvin light in another room where I do my print trimming, and a calibrated grayscale card to check the print against. Once you see a shift under the 6500 lamp, it is easy enough to compensate in the ABW mode by (say) adding +1 red or whatever. I would imagine the same thing would apply to colour printing, too.
John
Excellent points, but the term metamerism is not appropriate for the change in color with different illuminants. The terms metameric failure or metameric mismatch are more appropriate. The ink set has a lot to do with metameric failure, and early pigment inks for inkjet printers were notoriously prone to metameric failure, but newer inks have reduced the problem. Rather than calibrating for various viewing conditions, one could use inks exhibiting less metameric failure, but information concerning metameric failure is hard to come by.
http://photoshopnews.com/2005/04/20/metame...-friend-or-foe/ (http://photoshopnews.com/2005/04/20/metamerism-%E2%80%93-friend-or-foe/)
http://www.macworld.com/article/54216/2006/12/inktype.html (http://www.macworld.com/article/54216/2006/12/inktype.html)
-
see here... first PDF on the top of the list ("Color Matching Between sRGB Monitors and Wide Color Gamut Monitors"):
http://www.eizo.com/global/support/wp/index.html (http://www.eizo.com/global/support/wp/index.html)
Well if I use the NEC metamerism toggle it brings the look closer to look of the sRGB displays but things still don't look quite the same (perhaps this does the same things as the 2 vs 10 degree options).
Also using the metamerism toggle makes quite a few patches on a color checker chart look farther off from what they look on a real-life color checker chart.
In fact displaying a CCC chart on a wide gamut monitor with no metamerism toggle actually looks to my eye closer like a closer match to a real-life CCC than when it's displayed on one of the sRGB displays.
Maybe it's the sRGB LCD that are the ones that actually look off?
(OTOH if many proof photos and especially movies/tv on sRGB LCD then maybe the toggle needs to be
on to better match third party content.)
I wish I hadn't dumped off my Mitsubishi Diamondscan CRT yet. I wonder whether an sRGB LCD or wide gamut one look closer to it.
Reading some other articles it sounds like the differences differ person to person too. The only way to get matches might be to use full spectral matching at 1nm instead of the tristimulus model.
-
Excellent points, but the term metamerism is not appropriate for the change in color with different illuminants. The terms metameric failure or metameric mismatch are more appropriate. The ink set has a lot to do with metameric failure, and early pigment inks for inkjet printers were notoriously prone to metameric failure, but newer inks have reduced the problem. Rather than calibrating for various viewing conditions, one could use inks exhibiting less metameric failure, but information concerning metameric failure is hard to come by.
Yes, you are absolutely right. I was using the term "metamerism" far too loosely, as you point out. Reading the links you gave, I think the effect I am describing would be more correctly called Gray Balance Failure, perhaps. Of course, the way I work the one thing I am not doing is matching the colour or hue of the print to the screen, only the luminance values. The printer provides the warmth or coolness or (as near as dammit) neutrality in the ABW mode, according to the tone I set in the print driver and the white point of the printing paper.
I have to say that I have virtually zero experience with colour printing on my Epson. Whenever I do try it, in the past I have got into all sorts of trouble
John