My print is too dark - and I *have* calibrated and profiled

[Simon Garrett]

What responders don't seem to understand is there is another way that's not based on a technological fix, but on looking and learning and applying. It doesn't work for everyone; all I'm saying is that you don't need to reach for the Colour-Munki or Spyder to get great, repeatable results.

If you're suggesting that technology is somehow an alternative quick fix to "looking and learning and applying": well, I would say it's the opposite way round.  Photography is a technology-based creative art, and one needs some knowledge of the underlying technology to get the best out of it.  Without that knowledge one is just muddling through, it seems to me. 

I think all photographers would agree that to get the best out of the craft, a photographer needs to understand something of technologies such as aperture, shutter speed, focal length and so on, and to understand the implications of choices of those parameters.  Time has moved on, and photography is almost entirely digital.  That brings a whole new set of technologies an understanding of which enables better use of them. 

To some extent I agree with you second point: "you don't need to reach for the Colour-Munki or Spyder to get great, repeatable results".  I've just this last week given a talk to a camera club on colour management, and I know full well that most of them don't (and won't) use colour management.  As a result, a significant part of the talk was about the limitations of not using colour management, and the pitfalls to avoid when not using colour management.  However, IMHO one can get more repeatable, more consistent, faster results by "reaching for the Color-Munki or Spyder"!

[JoachimStrobel]

So I spend a long time trying to calibrate my monitor to the results I am getting from professional printing labs, mainly Whitewall in Germany. I illuminate the prints with a 50 watt tungsten bulb from 1-2 m distance, my set-up in my house for many rooms. Somme rooms have a wooden ceiling.
In winter, little extra light comes into the corner where the print hangs, now, in spring it is a bit more, therefore I waited with this post.
In spring, now, calibrating my screen to 5500k and 90cd gives a good match from screen to print, 6000K is still not bad. But everything above is not good. In winter 5000k is not low enought, but my Spyder 3 would not allow calibrating to 4000k. I tried it manually and it looked good.
I only have a simple LCD screen, in fact I tried a Dell and an Acer. They just so pass the SRGB triangle, it gets better with lower color temperature settings.
So at least I am done. As said before, I also have a lot of screens hanging at my walls. It is difficult to calibrate a TV screen to 5000K, 6500k seems to be "warm", some makers seem to set neutral to 9500K which people then find " brilliant". So I end up making two sets of outputs: One for TV-screens, adjusting my computer screen to 6500K and making pleasing adjustments on my photos for that. And one for print. I look at  prints mainly in wintertime, so I set the white point low.... Two outputs are fine, just as in old times: Slides were for projection and negatives for printing...

[Peter_DL]

So I end up making two sets of outputs: One for TV-screens, adjusting my computer screen to 6500K and making pleasing adjustments on my photos for that. And one for print. I look at  prints mainly in wintertime, so I set the white point low.... Two outputs are fine, just as in old times: Slides were for projection and negatives for printing...

Doesn't it also require two different luminance settings.
At least that's how I proceed: ca. 200 vs 75 cd/m2.

Regarding all the different white points which you mention, it is certainly a subject which can easily drive you crazy. In my experience the process of chromatic adaptation simply takes some time, longer than needed to support side by side comparisons. Hence the approach to calibrate the monitor's white point to a visual match e.g. between the display and the print next to it does not really make no sense for me. Others may *see* it differently.

With the Spyder Elite software it is possible to place the sRGB.icc profile in the file folder where the calibration targets are stored, so that the D65 CIE xy chromaticity values as well as the sRGB-TRC can be used as the target for calibration. Works well for me, I don't make it output-dependent.

Peter

--

[JoachimStrobel]

Thanks for the icc hint, I try that.
Sure, I forgot to mention the luminance problem. 80 to 90 cd/m2 is good for me for judging how a print may look. When considering photos for a screen display, then I can assume 140 to 200 cd/m2 as target. Again like in old times: slides vs prints.
For me it us about expectation management. The eye adapts to all. That is why one never "sees" a color taint in a dark room slight projection, but the very same photo hanging at a wall suddenly looks different. "Calibrating" my screen prepares me for what I will be getting. I had cases where the print looked much different than on screen - but I liked the print better. But this is a bit of a gamble...

Joachim

[Peter_DL]

"Calibrating" my screen prepares me for what I will be getting.

Hmm, I think the question is, if and how the different monitor and calibration settings really influence the way how we process and edit the images.

There is for sure a significant influence from the monitor's luminance. A low setting of <100 cd/m2 for print will provoke some stronger processing, in particular stronger tonal adjustments, compared to let's say 150-200 cd/m2 (or maybe even higher) for digital presentation.
Recommended reading: http://www.lumita.com/site_media/work/whitepapers/files/pscs3_rendering_image.pdf

Actually, in consequence, it may lead to a 2 step image editing approach with the different outputs in mind, a kind of making the "slides" first and then the prints.

Regarding the monitor's white point, I think there is some tolerance here - let's say somewhere in the range from 6500 to 5500K CCT, along the daylight locus or the Planckian locus – where it does not really matter, provided that there is sufficient time for chromatic adaptation.

Some time ago in the past I had changed from a somewhat warmer monitor white point to D65 now, and when I now revisit old images I do not feel the need to re-edit the white balance.

Peter

--

[JoachimStrobel]

Reflecting on the many post that I meanwhile wrote about this subject - it is frustration.This wonderful digital world, it has not managed to produce a switch like: this print will be for a wooden room with a 50 watt bulb and hence be around 4000k and please make me a print that is pleasing "Viewing type = mother" so replace take my current setting which are for a 6500K screen at 200cd/m2 and make it at nice print.
I can hardly believe that we write the year 2016 and all I have is a slider in Lightroom that can be used to make prints somehow brighter for prints. There is a disconnect somehow between the 42 MB sensor and the real world. Or I just wait for 4K panels replacing prints, I guess that will happen first.

[FrankG]

What I find interesting is that we can all set our calibration targets slightly differently to one another - for example, White Point & Gamma are pretty much universally accepted targets but the Luminance seems more variable depending....

Once we have the WP, Gamma & Luminance and our room lighting, and our print viewing booth, all balanced and sorted out we can arrive at a good display to print match.
Well & good, but what happens if we send our files to an outside lab for printing? If their Luminance is set differently then the display on my monitor may not match their print output.

Case in point - I have a calibrated display set to WP 6500K, Gamma 2.2, Luminance 115 cd/2. This works well, in my setup, when printing to a 3880 using manufacturers paper/ink profiles.
I just sent a few files to a lab and the prints are a tad too dark (under a Pantone Color Viewing booth) does that indicate that I have set my display luminance too high ?

Because I see the image lighter/brighter on my display compared to the slightly darker print output from the lab, instead of 115 cd/2 I could try setting it to 110 or 105 ? Dimming the display a little will make me brighten the images more.
Alternately, I could create a correction curve , or use the Screen mode duplicate layer method mentioned above. And apply that to all the lab print files.

Soft proofing with the lab's paper profile does not preview the slight darkening experienced.
Any thoughts? I'm just trying to tweak my calibration to match the lab output.
Thanks
Frank

[digitaldog]

Once we have the WP, Gamma & Luminance and our room lighting, and our print viewing booth, all balanced and sorted out we can arrive at a good display to print match.
Well & good, but what happens if we send our files to an outside lab for printing? If their Luminance is set differently then the display on my monitor may not match their print output.

The goal is WYSIWYG so you'd still be expected to place the lab's print next to your display and view the soft proof (expecting a visual match).

Case in point - I have a calibrated display set to WP 6500K, Gamma 2.2, Luminance 115 cd/2. This works well, in my setup, when printing to a 3880 using manufacturers paper/ink profiles.
I just sent a few files to a lab and the prints are a tad too dark (under a Pantone Color Viewing booth) does that indicate that I have set my display luminance too high ?

So one could suggest the lab's printer isn't ideally setup; it's printing dark. Or the profile they supplied isn't so good or it's not really being used. Or the device has drifted from what the profile once defined. Or, the contrast ratio differs which you could account for in your calibration. Or some of each. The ideal way to work IMHO is to use a display system that allows you to calibrate to differing output which is what my SpectraView does. Think about just your 3880 and the profiles that are for it that are presumably good. A matt paper has a vastly different contrast ratio and probably paper white than a glossy paper. You'd calibrate for both and switch the calibration and profile for each use.

Alternately, I could create a correction curve , or use the Screen mode duplicate layer method mentioned above. And apply that to all the lab print files.

You could but that's a non color management hack. It might come to that. But it's not ideal.
If you take a reference image who's RGB values are sound, the output should reflect that. Now there are differences in color gamut, rendering intents, profile quality and so on. Some we can account for, some we can't.

Soft proofing with the lab's paper profile does not preview the slight darkening experienced.

Which is why I suspect either the profile isn't so good or suffers some of the above possible issues or you need to better target the calibration of the display for that output.

[FrankG]

Thanks for all that Andrew. So many variables. And so it could vary even further if I switched from one lab to another. Or from inkjet to c-type continuous tone. Different calibrations for different labs/output is probably a good idea but beyond my old tools and setup - good to know for the future though (I'm going to upgrade)