Which temperature light is best for evaluating prints?

<p>I am setting up a color management workflow using Photoshop for fine art photos that includes monitor and printer/paper calibration, ambient light control and a dedicated print viewing area.</p>

<p>I'm having trouble deciding which temperature lighting to use in a print viewing area for evaluating the color of final prints. Some experts say 5000K (kelvin) lighting is ideal, others tout 4700K (D50) as ideal, while others like John Paul Caponigro say prints are best evaluated at the temperature lighting they are to be displayed at -- like 3500K.</p>

<p>That makes sense to me, but how do you get photos to print exactly how they appear on the monitor when the monitor is calibrated at the recommended 6500K, and the prints are evaluated under 3500K lighting? Or am I comparing apples & oranges here? Could anyone help clarify this for me? Thanks.</p>

<p>It all depends on your definition of "exact".</p>

<p>Two points to consider:</p>

<p>1) Color temperature (degrees Kelvin) only compares the intensity of the red and blue components of the light. It totally ignores green. Fluorescent lights commonly have an excess of green. For that reason alone I would prefer to use incandescent lighting, which has the proper ratio of green to the other colors.</p>

<p>2) You will never get a print to __exactly__ match what you see on the screen because you're comparing two different environments. Your monitor emits red, blue, and green light. The prints are reflective, giving off cyan, magenta, and yellow. Even the mixture of those three subtractive colors will not give you a clean black, which is why good printers include a separate black cartridge (or more than one).</p>

<p>I think it's a good idea to duplicate the type of lighting under which the prints will be displayed if at all possible.</p>

<p>- Leigh</p>

<p>Unless you are viewing both the print and the monitor at the same time, the color temperature will make less difference than you might expect, because your eyes will adjust to whatever the white balance is. I would tend to agree with using the same lighting prints are to be displayed at, but I doubt that there is a single right answer.</p>

<p>If you did need to match prints under 3500K lighting to a monitor at 6500K, you could soft proof using absolute colorimetric rendering intent, though you would need to have a printer profile for 3500K. You could also calibrate the monitor to match the 3500K white balance, which might save time soft proofing.</p>

<p>Also, color temperature by itself is not enough to guarantee that the white balance matches. You need both color temperature and tint to match, or x and y which are not the same thing but cover the same space.</p>

<blockquote>

<p>You will never get a print to __exactly__ match what you see on the screen because you're comparing two different environments. Your monitor emits red, blue, and green light. The prints are reflective, giving off cyan, magenta, and yellow.</p>

</blockquote>

<p>I disagree with this. <em> </em> What you said about the colors is more or less true, but your eyes only send three signals to your brain (long, medium and short—roughly red, green and blue). If the monitor and print under controlled lighting do not match to the human eye, it is for some reason other than the different color spectra.</p>

<blockquote>

<p>I disagree with this. <em> </em> What you said about the colors is more or less true, but your eyes only send three signals to your brain (long, medium and short—roughly red, green and blue).</p>

</blockquote>

<p>Hi Joe,</p>

<p>My comment about color matching was specifically addressing the realm of computer monitors v. prints viewed in broad-spectrum light (preferably incandescent).</p>

<p>The monitor emits light in three very narrow wavelength bands, corresponding pretty well with human eye response.</p>

<p>By contrast, prints reflect light over a broad spectrum due to the nature of the subtractive printing technology. This light is filtered by the receptors in the eye, which have a broader response than the emitted wavelengths from the monitor.</p>

<p>Take for example yellow ink. It absorbs blue and reflects everything else. The definition of "everything else" varies with ink formulation, as much as we would like to think it's precise. Some of that reflected energy may cause spurious response in the eye's color receptors.</p>

<p>A fine point, to be sure, but the OP did use the word "exact". ;-)</p>

<p>- Leigh</p>

<p> </p>

<p>All of that is true except for the part at the end about spurious response. It is just response, and is only short, medium or long. Whether these three signals came from yellow ink, red and green phosphors, tungsten lights, rainbows, lasers or anything else makes no difference. There are three signals coming out of the eye, and they can be the same for the monitor as for the yellow ink.</p>

<p>I am not saying that the light will be the same, only that people with normal color vision will see the same color. The extra information about the exact color spectrum has been lost and does not make it to one’s brain. I think that even the three signals then get processed down to two, but I know less about the details of that.</p>

<p>As for exact, there are all sorts of reasons why almost nothing is exact, but the additive versus subtractive color mixing is not one of them.</p>

<p>What the subtractive mixing does mean is that if the lighting is changed, the colors of the print may change while the monitor will not.</p>

<p>I don’t think the monitor even necessarily uses narrow color bands. I imagine that LCD monitors would use bands as wide as possible in order to be brighter.</p>

<p>Thanks Joe and Leigh.</p>

<p>In regard to my use of the word "exact" to describe how my final print compares to the image on my monitor, let me just say that my goal is to be able to produce a print that looks like the edited image on my monitor. Then I hope to become proficient at soft-proofing in order to save time and materials. I'm primarily a photographer/artist and rather challenged at the technical aspects of color management.</p>

<p>I'm working with fine art photographs that require attention to fine color gradations and tonality, but its not as critical as the exact color matching required in advertising layouts or product catalogs.</p>

<p>The lighting that I plan on using in my dedicated print viewing area are SoLux lamps which are available in 3500K, 4700K and 5000K. My work will be sold through local galleries that use lighting in the 3000K to 3500K range with occasional indirect daylight. I want my work to look as intended (from my monitor) in the final print that's on display at the gallery. How may I best achieve that relative to my monitor calibration and the quality light in which I evaluate my prints?</p>

<p> </p>

<blockquote>

<p>I want my work to look as intended (from my monitor) in the final print that's on display at the gallery. How may I best achieve that relative to my monitor calibration and the quality light in which I evaluate my prints?</p>

</blockquote>

<p>I recommend calibrating the monitor to a different white point. The best match may be at a higher color temperature than 3500K. Also, since any color temperature is a range of RGB values rather than a single color, you might have to fine tune it by setting x and y for the white point to get an exact match to the SoLux lamps.</p>

<p>I do not know how to create a printer profile at 3500K, since practical measuring devices will use built-in lamps. Perhaps the profiling software allows color temperature either as a setting or based on an ambient light reading?</p>

<p>Even with the regular printer profile (probably for a D50 white point) you will still want to soft proof to the printer profile to make sure that no colors are outside of the printer’s gamut.</p>

<p>Hi Richard,</p>

<p>Understand.</p>

<p>There are calibration systems available that set color profiles for both the monitor and printer. After setting these parameters the results __should__ match.</p>

<p>One such system is the Spyder3 Studio SR made by ColorVision, which I have. <br /> Here's the product link on Amazon.com: http://www.amazon.com/ColorVision-S3SSR100-SPYDER3-Studio-Sr/dp/B002N2Z332/ref=sr_1_8?ie=UTF8&s=electronics&qid=1278303078&sr=8-8</p>

<p>There are other similar systems from other manufacturers. I expect they all work similarly, although I've not worked with the others.</p>

<p>Even using one of these systems to set a starting point, you'll probably have to fine-tune the profiles to get what you want.</p>

<p>It's a non-trivial exercise, but probably well worth the effort.</p>

<p>Good luck.</p>

<p>- Leigh</p>

<p> </p>

<p>Thanks again. Leigh, good suggestion, which validates my choice of calibration equipment. I have a Spyder2Pro for monitor calibration and the Spyder3Print device for printer calibration.<br>

<br /> I noticed that the SpyderPro documentation has recently added a single statement as an alternative to calibrating the monitor at 6500K, "...some photographers may prefer selecting the 5800K option setting." That's all it says, but maybe this is in regard to my concerns.<br>

<br /> I can't find any further information on John Paul Caponigro's website concerning his well-publicized recommendation of calibrating final print output for the display environment of 3500K. I guess the details of that procedure are privy to those who take one of his fine art printing seminars.</p>

<p>Hi Richard,</p>

<p>I believe the Spyder3 Studio is the monitor and print equipment all packaged in an aluminum carry case. The SR is a plastic guide for the print sensor ("spectro"??), which I don't like.</p>

<p>I've not tried calibrating the monitor at other than 6500°K, so I'm not sure how to accomplish that. I expect there's a parameter entry field.</p>

<p>Perhaps a question to the folks at ColorVision would clarify the monitor temperature issue.</p>

<p>Good luck with the project. Sounds like you have everything you need.</p>

<p>- Leigh</p>

<p> </p>

<blockquote>

<p>I want my work to look as intended (from my monitor) in the final print that's on display at the gallery. How may I best achieve that relative to my monitor calibration and the quality light in which I evaluate my prints?</p>

</blockquote>

<p>simple, you start at the end. The one thing you can't (more often than not) control is the lighting conditions in a gallery. So you can calibrate and profile all you want but there you have it. The last exhibit I did was in fairly "dark" conditions. I never bothered to measure color temperatures but just printed somewhat lighter and that was that as long as you realise that these prints are hardly suitable for sale. Going all scientific about it can have its place but sometimes a more pragmatic approach will get you further with a lot less work.<br>

As for evaluating prints I'm not sure about other people but I evaluate mine in daylight conditions., Always.</p>

<p>I can relate Tom. But I have a unique opportunity in that I CAN control the lighting in a particular gallery that my photographs will be displayed -- with 3500K SoLux lighting.</p>

<p>Here is an interesting article just presented to me by Larry over at the Adobe Forum:<br /> http://www.solux.net/cgi-bin/tlistore/infopages/eyes-response.html</p>

<p>In this article it points out that when an individual is outdoors 6500K looks white, but when indoors 4700K looks white while 6500K looks too blue. Then in a dimmer museum-like lighting condition, 3500K looks white with 4700K looking blue -- all based on the function of the human eye. Interesting study, especially in regard to how your potential customers are perceiving the colors in your displayed work.</p>

<p>Of course, my question is how do you go about setting up a color management workflow based on lighting temperatures less than the "holy grail" calibration standard of 6500K and print evaluation under 5000K?</p>

<blockquote>

<p>My work will be sold through local galleries that use lighting in the 3000K to 3500K range</p>

</blockquote>

<p>So using a similar lighting you can preview what clients will view.</p>

 

<p>Go with Caponigro. This is an issue that I began to take a lot of notice of maybe ten years ago when I was working on a very large multi-site office decor project. The lab-produced prints always looked great in my office once I had cured the lab of printing everything too dark - typical of looking at stuff at arms length rather than at wall viewing distances. However none of the rooms, corridors and reception areas were consistently lit, and none had lights directly over where photographs were required. Whats more some rooms were internal, others variably sunny at different times of day. There was further complication in that the client didn't want to pay for reduced reflectivity glass except in the most public areas. The only thing I could do was to survey each room and assign a code of light, medium, and dark to each hanging spot and reach an understanding with my printer about what these descriprions actually meant for the printing process. Then I had to worry about whether the client's maintenance staff would hang the right picture in the right place, and sometimes they didn't.</p>

<p>What i learned from this is I guess two things. First is that the interreleationship between how you print and the lighting in the hanging place is indeed an issue that will affect how people appreciate your photography and is worth thinking about. Second is that whilst I agree that your work has to look great in a gallery, bear in mind that you can't guess how that relates to the eventual hanging place. You'll still have customers that just can't work out why the photograph on their wall doesn't look like it did in the gallery.</p>

<p>The right color of the viewing conditions are those that produce a match to the display next to the booth. Otherwise, if you are just viewing prints, your eye will adapt to the white. <br>

Kelvin is a range of colors, not an exact color. see: http://www.ppmag.com/reviews/200512_rodneycm.pdf<br>

A Solux rated at 5000K, a Fluorescent bulb rated the same will not produce a match. Fluorescent lights have a spiky spectrum (due to the nasty mercury) that can have issues with some papers containing high UV components (optical brighteners). <br>

I’ve seen many galleries examine what I consider to be the closest man made light to sunlight, Solux bulbs. The CCT 4700K are supposed to be best for proofing (D50), yet it appears too cool for a pleasing rendering of most artwork, CCT 3500K are much more pleasing. Which should you use? Kind of depends on the use. Most galleries end up with 3500K. This is why JP is suggesting this temp WITH Solux bulbs. <br>

A 3500 Solux used to view a print next to a display calibrated to CCT 6500K (better, D65) might produce a match. Depends on the software used to calibrate the display and the display itself. The values are far from actual absolutes! You can take the same instrument and use two different calibration packages, use the same white point target and you will not get the same results. I know, it should not work this way but it does. Again, the only actual D65 Lightsource is 93 million miles away. Maybe if your display, when heated to a molten pool of plastic and metal, acted like a theoretical black body radiator, it would produce something close to the definition of D65. </p>

"Again, the only actual D65 Lightsource is 93 million miles away."

 

And even that is not really true. The CIE standard illuminants are mathematical models. D65 was created

by interpolating spectral power distribution data from hundreds of different locations, not by measuring a

single source in a single place/time. There is no actual D65 source.

<p>For those who actually have a real live client; it is *THEIR* area where the print will be displayed.</p>

<p>They might not know a rats rear about color temperature. They might have an evil mix of lighting types. They might be an older person with cataracts with a yellow bias in their eyesight.</p>

<p>Remember if your master piece looks great under lab conditions; you have person who is going to display it in a dark casino bar; under Kitchen and Bath lamps; under cold LED lamps; it might have some colors the TV folks really hate too.</p>

<p>*PART* of pleasing a client is actually making print look good to them.</p>

<p>Some folks may not want you to change up their lighting in their store or home.</p>

<p>Thus here in printing our color checking booth has not only the common daylight refenece sources; but a zoo of oddball ones too; probably more than anybody else. Thus the lighting mixer can mix up the normal color reference; but mixes.</p>

<p>If a client wants the beer wrapper to look good too under beer cooler fluorescent lighing at 7/11; we have that too; besides the "sheltered" case of daylight.</p>

<p>We often get clients who have left others who get their panties all wrapped up in color gobble gook; but have zero experience in the real world of bastard light sources a client uses.</p>

<p>One has the *bulk* of retail using Fluorescents and vapor arc lamps; and their stores are often not even consistant either.</p>

<p>In advanced printing for clients prints are noften made/biased to look great under *their* bastard lighting; not some sheltered perfect daylight black body. Thus in advanced printing one *closes the loop* around the clients goals; ie their lighting types,. Thus in casinio work, retail, and corporate a survey is done on site to inspect the lighting. The same goes with hockey dasher boards; I use a different for each rink that has a different type of lighting.</p>

<p>In actual printing for a living on has a real live person to please. They want their product to look good under their lighting type. If their product looks better at XYZ rink with a bias; so be it. Folks who live in a bubble can crank out stuff that look great under lab conditions; and poor under a clients lighting; and thus loose that customer.</p>

<p>Your grandmother with cataracts might prefer her or granddaughterswedding images to be bluer; and with a lot more saturation; and your perfect print is considered poor.</p>

<p>Thanks, David, Andrew and Kelly, for bringing the lofty Theory of Color Management Workflow down to the realities of working (and selling) in the real world. </p>

<p>So, from all this I surmise that it's best to first "standardize" your color management workflow with calibrating one's monitor to a 6500K white point and 2.2 Gamma, while adjusting luminance to a level in harmony with your studio's ambient lighting. Next, start with evaluating your prints (made from a calibrated printer/ink/paper profile) under bright (D50) lighting for any printing imperfections and sharpness, but then try to match the lighting of the print's ultimate destination in your print viewing area for the final evaluation -- in my case, under 3500K SoLux lamps.</p>

<p>Hopefully, if everything is set up properly (the light intensity in the viewing area and the monitor luminance), the "white" of the monitor is close to the "white" of the photo paper in your viewing area. Perceptually, the eye will compensate for color temp variance if monitor and viewing area are well separated. Then the needed editing can be accurately made with minimal waste of time and materials. Am I even close to being on the right track here?</p>

<p>Kelly, (or anyone else) for all the various lighting/viewing scenarios for which you must produce prints, do you first produce the prints to a standard as I mention above and then edit with curves for a specific lighting environment (via presets?), or do you have custom ICC print profiles for the various lighting scenarios?</p>

<blockquote>

<p>The CIE standard illuminants are mathematical models. D65 was created by interpolating spectral power distribution data from hundreds of different locations, not by measuring a single source in a single place/time. There is no actual D65 source.</p>

</blockquote>

<p>True although I seem to recall (and have to check some notes) that it was far fewer than hundreds of <strong>locations</strong>*. I think its was only two or three. So yes, averaging several or hundreds of collected measurements may have been its origin, I’m not so sure I’d say emphatically that there is no actual D65 source. Its certainly a possibility and it certainly was generated thanks to the object 93 million miles away. </p>

<p>*Found it:</p>

<blockquote>

<p>The series of D-illuminants was adopted by the CIE in 1971 based on 622measurements from the early 1960s: 249 at Rochester, NY (Kodak); 274 at Enfield, England (Thorn Electrical Industries); and 99 at Ottawa, Canada (National Research Council). Each of these labs contributed spectralmeasurements taken with different kinds of instruments measuring atdifferent spectral intervals over slightly different ranges. The data were combined into a master set consisting of averages over 10 nm intervals from 330 to 700 nm from which the average and four characteristic vectors were calculated. The average and first two of these vectors account for most of the variance in the observed data and live on as the S0, S1, and S2 vectors used to calculate the D-illuminants in the CIE standard (see Wyszecki & Stiles, 2nd. Ed., page146). S0 is the mean, S1 provides a yellow-blue variation relating to cloud cover and inclusion/exclusion of direct sunlight, and S2 providesa pink-green variation which was thought at the time to derive fromvariations in atmospheric water vapor and haze.<br />All of this was reported by Judd, MacAdam and Wyszecki, J. Opt. Soc.Am., Vol. 54, p. 1031 (1964) and was incorporated without change intothe 1971 CIE standard except for the addition of the formula for illuminant chromaticities in terms of correlated color temperature due to Kelly at NBS (now NIST, Washington, D.C.).</p>

</blockquote>

<blockquote>

<p>The series of D-illuminants was adopted by the CIE in 1971 based on 622measurements from the early 1960s: 249 at Rochester, NY (Kodak); 274 at Enfield, England (Thorn Electrical Industries); and 99 at Ottawa, Canada (National Research Council). Each of these labs contributed spectralmeasurements taken with different kinds of instruments measuring atdifferent spectral intervals over slightly different ranges. </p>

 

</blockquote>

<p>Well at least we know someone was gainfully employed in all of this color science. Too bad their data still can't be understood or implemented in creating and reproducing images to the degree we no longer get color temp questions like this.</p>

<p>To put it bluntly, Richard, there really is no standard to go by visually, except what the color scientist use in a controlled lab. I haven't seen artwork come out of a lab, so it's hard to say if even they can implement what they calculate mathematically or theoretically.</p>

<p>You create the standard with the source of your image and last time I checked there's no way to look at or edit an image digitally without a <strong>DISPLAY</strong> which is now the de-facto source of your image. Might as well stick with a coolish 6500K looking color temp because everybody else has set there displays to that. Yours and everybody else's eyes whether viewing in a gallery or at a desktop will correct for any mismatch because of the adaptive nature of the eyes. </p>

<p>No one in a gallery is going to see or be comparing and judging the quality of your image to how it appears on your display. No one is going to know the difference if a warmish 3500K light is going to make that vibrant cyan blue sky look any less breath taking than what you see on your display. </p>

<p>Make your images look great on your display and be done with it, because everyone else who sees your images in a gallery is going to see it for the first time and not see the differences the lighting has on them.</p>

<p>Don't work too hard. Let the color management technology do the work.</p>

 

<p>Check ISO 3664</p>

<p>Thanks. ISO 3664 (2009) looks like it could be interesting.</p>

<p>Anyone know of a source for free copy of it? The ISO guys charge $130 for a PDF copy.</p>

<p>Yes, ISO 3664 makes prints look great. That math says so.</p>

<p>I view my prints and everything else in my studio under a variety of lights that include (1) 50 watt Solux 4700K D50 task lamp, (3) 15 watt GE 5000K Sunshine tubes, (4) 100 watt 5500K HD Ottlight CFL's, (2) 40 watt 4100K GE SP41's in my kitchen, (3) 25 watt 2800K vanity mirror incandescent globes with three CFL equivalents in my bathroom.</p>

<p>The lights that make my prints look the worst are the globes, both CFL and incandescent.</p>

<p>In ascending order from tolerable to print to screen match are the SP41's, the Ottlights, GE Sunshine and Solux with the Solux giving the best match and the GE Sunshines a very close second.</p>

<p>Every one of these lights give a different color cast that defy the Kelvin numbers. Solux actually matches VERY CLOSE to the light of a direct sunbeam, not noon outdoor daylight with the sky to contaminate the cast.</p>

<p>Thanks Tim, very helpful. Glad I have decided to go with SoLux lamps.</p>

<p>However, rather than trying to match the lighting of my print viewing area to my monitor, my main concern is how to match my monitor's color calibration to my color balanced prints as viewed under 3500K SoLux lighting -- the lighting under which my prints will be displayed for sale. </p>

<p>The interaction of subtle hues and tonality upon a particular print paper play an important part in the composition of my color fine art photographs. Unless these are all intact and visible, the composition can fail. This is often the case when images are carefully edited on a monitor with standard calibration (6500K white point, 2.2 gamma, etc.) and then viewed under moderate intensity 3500K lighting. </p>

<p>My goal is to tweak my color management workflow so that editing and producing a "successful" print is less of a trial & error experience. I don't know enough about the voodoo of monitor color calibration to deviate much from a "standard" calibration in order to get my monitor to emulate the color balance necessary for the display/viewing environment of my prints. Any advice?</p>

 

<p>Maybe I'm way wrong, but I always thuoght the first job of the viewing station is to have the print match the monitor, one of the main purposes of calibration and setting white balance. We learned that whatever tempurature your monitor is calculated to, that's what you viewing station should be. We happened to use 6000k at school with printer profiles to match. The viewing stations were hooded to create the viewing environment as were the computer monitors. All and all a very expensive set up, but they had the equipment funds.</p>