Printing - compensation of dynamic range narrowing

[Guill]

Hello,
I'm planning to print some photographs I made (it'll be the first time I print photos, so I'm a beginner in this field).
I did some soft proofing in PS in order to see how the photographs will look like on the paper I'll use.
Some contrast and colors are lost.
My question is : do people increase the contrast and something else (saturation ?) in their images, in order to have a print which is closer in appearance to the image they were used to vizualize on the screen of their computer (at the end of its postprocessing and without soft proofing) ? Or should it be avoided to do anything for some theoretical reasons ?
Thank you in advance for your help,
you can make me suggestions, recommend me websites/pages to read,
best regards.

(NB my screen is correctly calibrated :EIZO ColorEdge CX241 + X-Rite i1 Display Pro
but I'm not completely sure of one setting I shall use : the white point - is it related to the paper+ink+printer I'll use ?
I read that for a white paper I should use D65-6500K, for a warmer paper : <6000K
or shall I use D50-5000K for any paper ?)

[pschlute]

There are many "printers" on this forum who will give you better advice than me, especially regarding contrast and saturation.

Two points I would make though:

First make sure your monitor is (ideally) correctly calibrated (XRite or Spyder for example), and especially is not set too bright. Otherwise you are likely to find your prints are too dark.

Secondly, printing usually requires more output sharpening than screen viewing. As a rule of thumb i double my USM sharpening settings when preparing an image for printing.

[Manfred M]

Welcome to CiC, Guill. Would you mind clicking the "My Profile" button on the top of the page and at least fill in your first name and the country where you live?

Peter is correct, there are a few of us that do a lot of print making and I am one of those people. While you have calibrated and profiled you computer screen (which is Adobe RGB compliant); what brightness level (in candela / square meter) was your screen output set to? I run mine at 80 cd/ square meter and that gives me a fairly good representation of what my print output will look like. The ambient light level at my work station is just below 40 lux, which is actually a tiny bit bright. I suspect your screen might be a touch too bright if you are noticing a significant difference between what your screen and what your print looks like.

There are a number of reasons for this; your screen is an RGB, transmitted light, additive colour process. Most screens have a 10-stop dynamic range. Your prints are going to be a CMYK, reflected light, subtractive colour process. A gloss paper is going to have a dynamic range of around 8-stops and matte papers are around 7.5 stops, so that accounts for part of the contrast issue you have identified. Your prints will never look 100% the same as the image you are viewing on screen for that reason.

Your calibration white point is your screen's white point and have nothing to do with the paper / ink set you are using; both sRGB and Adobe RGB are based on D65. This is what you should use when you calibrate / profile your screen.

Just about every print maker I know has said something along the lines of "I wish that they would make a screen that accurately reproduces what I see in my prints". Soft proofing is not a 100% solution by any means. It is an emulation only. I tend to use it when I prepare my prints to see how much of my material is out of gamut (OOG), based on the ICC profile of the paper I am using. I do test prints, rather than using what I see on my screen to judge my prints. Rendering intent will be either Relative Colorimetric or Perceptual; I will check both in soft proofing to see which on I like better. Make sure you turn Black Point Compensation on. Some people will tweak their images in soft proofing; this is usually more of an issue when printing on cotton rag fine art papers than on glossier papers like baryta ones.

Test prints is what I used to use when I worked in the colour "wet" darkroom and this is still the most accurate method of evaluating your prints. Ideally you will evaluate your prints in the light that you will be displaying them under.

I was lucky enough to learn printing from a couple of very well regarded fine art print makers with a lot of experience in the business. I have not seen a lot of material on the internet that really resonates with me. The best I have found are some YouTube videos that Mitch Boyer produced. You might want to have a look there to start. Personally, I found the biggest single factor that improved my print quality is "calming" the image through dodging and burning so that it work well in the lower contrast print environment.

[DanK]

Manfred is a more expert printer than I am, so I don't have much to add. Contrast and saturation will definitely look very different if you have the wrong ratio of monitor brightness to ambient lighting. This is partly trial and error, because your brain is trying to translate between emissive and reflective images. The only way to settle on values is to use test prints. I have found that what works for me is a slightly brigher monitor than Manfred's. I set mine to 100 or 110 cd/m^2.

However, I would add one thing: this depends a great deal on the paper you are using. in general, coated papers have a wider dymanic range than matte papers. I find that when I softproof for the baryta papers I use, I rarely make more than a small adjustment to contrast. However, when I use matte papers, the contrast and saturation appear considerably less. This is one reason why I exhibit almost entirely on baryta and similar coated papers, even though matte and "fine art" papers are the style now.

Yes, you can correct to some degree with soft proofing. However, as we say here, "there is no free lunch". Increasing contrast in softproofing doesn't give the paper a bigger dynamic range. It simply moves the larger changes in luminosity to the mid-tones from the extremes. The cost of doing this is that you can sometimes lose detail in the highlights or shadows.

[Guill]

@ pschlute - Thank you for your answer, yes my monitor is calibrated (but maybe not with the best values for printing (much more for producing images for internet...))
Best regards

[Guill]

@ Manfred
Hello Manfred,
(you're correct) I checked my photographs at 120 cd/m², but I was planning to decrease brightness level to 90 or 80 cd/m². I didn't check the photographs with this level because I was not sure of the white point I shall use for the new calibration for printing. Thanks for you recommendation on this point.
On the one hand I read the website of a good photographer who was talking about 6500K or a little bit less, depending on the paper used, on the other hand the website of my print maker recommends to use 5000K... quite different if not poles apart...
I'll have a look to Mitch Boyer videos
Best regards

[Guill]

@ Dan
Thank you for your answer. The paper I'll use (through my (online) print maker (for the types of print I want to have)) will be brillant, with Dmax=2,4 (which means a contrast of 250), 220g, it's "RC Brillant for pigmentary inks", and its ICC profile is provided by my print maker. I was planning to order some test prints. But before starting I wanted to be sure of the parameters I shall use to vizualize my photographs on my screen (and to calibrate it).
Thank you for the comment on the repercussion of contrast increasing. I was contemplating to change it in order to have dark areas as dark as on my screen (but I must first start to adjust it at 80 or 90 cd/m² as recommended by Manfred in order to see if I still see a substantial difference when doing softproofing).
Best regards

[Manfred M]

@ Manfred
Hello Manfred,
(you're correct) I checked my photographs at 120 cd/m², but I was planning to decrease brightness level to 90 or 80 cd/m². I didn't check the photographs with this level because I was not sure of the white point I shall use for the new calibration for printing. Thanks for you recommendation on this point.
On the one hand I read the website of a good photographer who was talking about 6500K or a little bit less, depending on the paper used, on the other hand the website of my print maker recommends to use 5000K... quite different if not poles apart...
I'll have a look to Mitch Boyer videos
Best regards

This part of photography is often more of a black art than a science. Many experienced print makers have their own secret approaches and even though they are technically incorrect, they still produce brilliant images. I know one excellent print maker who uses an un-calibrated computer screen and still creates stunning, accurate prints. His secret is test prints; he solely relies on his viewing of the test prints to make his adjustments.

As for the D65 (6500K) and the D50 (5000K) argument, the answer I have given you is the technically correct one. If you read the sRGB and AdobeRGB specs, both state that they are based on the D65 illuminant. ProPhoto RGB, on the other hand uses the D50 illuminant.

From a historical standpoint, the D50 standard was used in the offset press printing industry. Before the widespread use of the web to deliver content, photographers prepared copy for pre-press work and their work was reviewed using calibrated viewing boxes (MacBeth was the biggest maker of these). Some of the old timers still have an affinity for D50 lighting. Epson had a special version of both the 3880 and P800 printers that were specifically targeted for making pre-press proofs using the CMYK colour model; the ICC profiles provided by the specific printing houses would be for the specific presses, papers and inks that they were using for the production run.

Most other industries gravitated to the D65 standard, so far as I know; paints, fabrics, etc. They also did their colour evaluations using the MacBeth type light boxes. I suspect that at the time that the sRGB (1996) and Adobe RGB (1998) came out the CRT type displays had a slight bias to the bluer end of the spectrum, so this would have been a logical position.

As for your print maker, I would suggest he has found a formula that he is comfortable with, even though it isn't 100% "to spec" and as long as you are happy with the results he gives you, that is great.

Ultimately, the question becomes, what lighting conditions are you viewing the proofs and final prints under? This, in my view, is probably going to have an even greater impact on how you view your work.

[Guill]

Hello Manfred,
thank you again for your detailed answer.

As I'm a beginner in this field I may ask myself too much questions.
Once I'll have found a combination/set of adjustments for calibration of my screen for printing that are reassuring me (most of them) I'll control again one of my photographs and then quickly send it to the print maker to quickly have a print to examine, in order to see if I'm satisfied or not with the result.

But I'll have no special lighting equipment to control them. (Moreover currently I'm in a temporary accommodation, with old tungsten light bulbs... old yellowish wallpaper/tapestry... not optimal).
Do you have some special lighting equipment to control your prints ? I saw it's easy to find some in D50, but those in D65 seem to be rare and expensive. I don't know if it really matters to have exactly a D65 lighting, just like on my screen.

I started to watch some Mitch Boyer videos. They are very instructive for a beginner in printing. Fortunately I was able to make a lot of links with what I know in other fields (like post-processing).
If you've other names or pages to suggest me : don't hesitate, they'll be welcome. Two decades ago I was reading Norman Koren's website a lot. I should have a look again, it was very detailled on a lot of topics, but I didn't read the parts about printing. On the other hand, in the meantime equipments have evolved. I'll see if informations are still useful.

Thank you again,
best regards

[Manfred M]

Hello Manfred,
thank you again for your detailed answer.

As I'm a beginner in this field I may ask myself too much questions.
Once I'll have found a combination/set of adjustments for calibration of my screen for printing that are reassuring me (most of them) I'll control again one of my photographs and then quickly send it to the print maker to quickly have a print to examine, in order to see if I'm satisfied or not with the result.

But I'll have no special lighting equipment to control them. (Moreover currently I'm in a temporary accommodation, with old tungsten light bulbs... old yellowish wallpaper/tapestry... not optimal).
Do you have some special lighting equipment to control your prints ? I saw it's easy to find some in D50, but those in D65 seem to be rare and expensive. I don't know if it really matters to have exactly a D65 lighting, just like on my screen.

I started to watch some Mitch Boyer videos. They are very instructive for a beginner in printing. Fortunately I was able to make a lot of links with what I know in other fields (like post-processing).
If you've other names or pages to suggest me : don't hesitate, they'll be welcome. Two decades ago I was reading Norman Koren's website a lot. I should have a look again, it was very detailled on a lot of topics, but I didn't read the parts about printing. On the other hand, in the meantime equipments have evolved. I'll see if informations are still useful.

Thank you again,
best regards

The ideal light to evaluate your prints is to view them under the conditions where they will be displayed. In practice, this is often not possible or practical.

Getting light sources that are full spectrum and have adjustable output would be ideal and the 5000K and 6500K are somewhat moot. Those two colour temperatures reflect mid-day. outdoor lighting conditions that we would rarely view finished prints under. A lot of high end, world class museums and galleries use lights from the American specialty lighting supplier, Solux.

http://www.solux.net/cgi-bin/tlistor...ges/index.html

I have used Solux tungsten bulbs and found that they give excellent results, but they have one drawback; colour temperature drops when the bulbs are dimmed. They work great at full line voltage.

My problem is that I want to see the prints that come off my printer under two different sets of illumination; prints that are shown in galleries, exhibitions and competitions are frequently blasted with light levels that are in the order of 200 - 250 lux. Images shown in homes are more likely to be illuminated at 100 - 150 lux, so I want a light source right above my printer so that I can read the prints as they come off.

Technically a high CRI dimmable LED light (CRI>95) will do the job, but I need to do some testing. I have a specialty lighting supplier in town who can get me what need. Unfortunately, having a high CRI does not guarantee success as the definition is a bit loose and some fairly poor lights can have high CRI values. I also want a nice, soft and even light distribution with no hot spots. While I was working on this a year ago, COVID-19 has gummed up the works and I am unlikely to finish this little project before later in this year.

[DanK]

High CRI alone is definitely not enough, IMHO. The index was developed in the fluorescent era and lacks some reds. (What I have been told, but can't verify, is that the fluorescent bulb manufacturers wanted that red range excluded because they had no phosphors that produced it well.) You should look for bulbs that have high values of both CRI and R9. In rare cases, you can find bulb specifications that include a fidelity index and a gamut index. In the US, many LEDs with high CRI values have R9 values as low as 50.

In addition, I consider it essential that color remain consistent as the bulbs are dimmed, as I use much less than full brightness when I edit.

I settled on these: https://soraa.com/products/51-Soraa-VIVID-BR30. They are CRI=95 and R9=95. They run about US$17 (14 Euro) each. They maintain their spectral distribution well when dimmed. However, I believe they are not available outside of the US. It might be helpful to download the spec sheet from that link and show it to some French vendors to see whether they have anything comparable.

It makes a big difference. I recently noticed that the reds in a print on one wall looked quite far off. I wondered if I had made a printing error, so I took the print off the wall and carried it to a room with brighter light. The room where I checked it had Soraa bulbs, and the print looked fine. The room where the print looked bad had cheap LEDs with CRI=80 and no R9 value in their public specifications.

[xpatUSA]

The ideal light to evaluate your prints is to view them under the conditions where they will be displayed. In practice, this is often not possible or practical. <>


My problem is that I want to see the prints that come off my printer under two different sets of illumination; prints that are shown in galleries, exhibitions and competitions are frequently blasted with light levels that are in the order of 200 - 250 lux. Images shown in homes are more likely to be illuminated at 100 - 150 lux, so I want a light source right above my printer so that I can read the prints as they come off.

...

Would it be practicable to:

a) have a number of lamps that can be variously switched off and on to vary the incident lux or,

b) have a set of diffusers that can vary the incident lux or,

c) vary the distance of the lamp from the print or,

d) mount an aperture of some sort betwixt the lamp and the print?

As to diffusers, one assumes they wouldn't affect the radiant spectrum much and of course won't catch fire, LOL. How about stacking tracing paper?

Also, instead of diffusing, how about spreading the light (e.g through a lens) so as to increase the solid angle of the beam thereby reducing the incident lux.

If the only levels of concern are in the order of 200 - 250 lux or in the order of 100 - 150 lux, two lamps of appropriate lumens would do it - or, if they are sorta point sources, the same lamps at different distances per c) above.

Please pardon the technical response - which I realize "most people" may not get ...

[xpatUSA]

High CRI alone is definitely not enough, IMHO. The index was developed in the fluorescent era and lacks some reds. (What I have been told, but can't verify, is that the fluorescent bulb manufacturers wanted that red range excluded because they had no phosphors that produced it well.) You should look for bulbs that have high values of both CRI and R9. In rare cases, you can find bulb specifications that include a fidelity index and a gamut index. In the US, many LEDs with high CRI values have R9 values as low as 50.

I have a vague memory of reading that some manufacturers tuned phosphors in their lamps so as to get a good CRI score ...

[Manfred M]

Would it be practicable to:

a) have a number of lamps that can be variously switched off and on to vary the incident lux or,

b) have a set of diffusers that can vary the incident lux or,

c) vary the distance of the lamp from the print or,

d) mount an aperture of some sort betwixt the lamp and the print?

As to diffusers, one assumes they wouldn't affect the radiant spectrum much and of course won't catch fire, LOL. How about stacking tracing paper?

Also, instead of diffusing, how about spreading the light (e.g through a lens) so as to increase the solid angle of the beam thereby reducing the incident lux.

If the only levels of concern are in the order of 200 - 250 lux or in the order of 100 - 150 lux, two lamps of appropriate lumens would do it - or, if they are sorta point sources, the same lamps at different distances per c) above.

Please pardon the technical response - which I realize "most people" may not get ...

Ted - I had considered some of those options as well and in fact I use a set of fairly bright tungsten lights where I have two separate distances marked that give me 200 lux and 150 lux. It works but ins a PITA, so I am looking for a somewhat more flexible and versatile solution, so that I can just dial in dial in several values. I find that I sometimes can use the same print in exhibitions / print competitions and at other times, I have to make two different prints.

[Manfred M]

I settled on these: https://soraa.com/products/51-Soraa-VIVID-BR30. They are CRI=95 and R9=95. They run about US$17 (14 Euro) each. They maintain their spectral distribution well when dimmed.

I had looked to see if I could find these lights in Canada the last time you mentioned them, but there were no distributors here and shipping them in from the USA was proving to be an expensive challenge.

Soraa appears to have been bought out by Ecosense, who have an office in the west end of Ottawa, so I will see if they deal with consumers and if not, who their local dealers are.

[DanK]

I have a vague memory of reading that some manufacturers tuned phosphors in their lamps so as to get a good CRI score .

Of course. It's called Campbell's Law, after Donald Campbell, one of greats in the field of program evaluation:

"The more any quantitative social indicator is used for social decision making, the more subject it will be to corruption pressures and the more apt it will be to distort and corrupt the social processes it is intended to monitor."

There are many, many examples, from vehicle emissions standards, CPU speed measurement, educational testing, policing, airline on-time statistics, and on and on.

[Guill]

Hello,

I made some new soft-proofing tests (with my screen now adjusted to 6500 K 80 cd/m² - much better setting, thank you for the help).

With "RC Brillant for pigmentary inks" paper I see much more differences (with the original photograph) than with "Baryta Hahnemühle 315g" or with "Platine Fibre Infinity Canson 310g" :

• much bluer and darker with the first paper (and a bit of contrast lost too),
• almost the same colors and contrast with the second and slighty darker,
• almost the same colors and contrast with the third and just slighty darker (less than with Hahnemühle).

I will check with some other papers. Unfortunately I may not be able to avoid the first paper, so I may have to counterbalance/offset a little bit with this one. I should order two prints in this case : the original and the counterbalanced/offset, and then judge which one is the most satisfying.

Thank you for your detailled answers about lighting too. As I've no knowledge in this field too : it gives me some points of reference.

Best regards

[Manfred M]

Hello,

I made some new soft-proofing tests (with my screen now adjusted to 6500 K 80 cd/m² - much better setting, thank you for the help).

With "RC Brillant for pigmentary inks" paper I see much more differences (with the original photograph) than with "Baryta Hahnemühle 315g" or with "Platine Fibre Infinity Canson 310g" :

• much bluer and darker with the first paper (and a bit of contrast lost too),
• almost the same colors and contrast with the second and slighty darker,
• almost the same colors and contrast with the third and just slighty darker (less than with Hahnemühle).

I will check with some other papers. Unfortunately I may not be able to avoid the first paper, so I may have to counterbalance/offset a little bit with this one. I should order two prints in this case : the original and the counterbalanced/offset, and then judge which one is the most satisfying.

Thank you for your detailled answers about lighting too. As I've no knowledge in this field too : it gives me some points of reference.

Best regards

All three of your papers appear to be "luster" papers and they are all going to give you decent contrast. I don't know the RC Brilliant product, I use a couple of Hahnemühle baryta papers (Photo Silk Baryta 310 gsm and Photo Gloss Baryta 320 gsm). I also have the Canson Infinity Platine Fiber Rag 310 gsm, but have not printed on it since the summer. You will find quite a difference if you use any of the matte cotton rag papers; beautiful papers, but they do print a lot flatter. I use Epson Hot Press Natural for portraiture.

What I would suggest is that you get an Adobe RGB print target from the internet, if you don't already have one and do a test print using both the Relative Colorimetric and the Perceptual rendering intents and keep them for reference. I tend to do this and have a binder of reference prints to refer to.

Here is a source of a print target: https://www.digitalmasters.com.au/Ca..._Adobe_RGB.jpg

[DanK]

Guillame,

I'm confused by your last post. I don't understand what you are comparing. Are you comparing the original image on screen to the softproofed image on screen before you make any adjustments? And are you doing this with ICC profiles specific to your printer and each of the papers?

I've tested a large number of coated papers using a standard test print, http://www.jirvana.com/printer_tests...Image_V002.zip, and the manufacturer's ICC profiles for my printer. Apart from the fact that the base stock varies on the cold-warm continuum, most of the papers were quite similar in terms of color. (The Hahnemuehla and Canson papers you mentioned are among those I tested.) I can recall only one that was substantially different, which I suspect was a problem with the ICC profile. Apart from that one paper, the differences were in some cases visible, but they weren't extreme. And consistent with that, the softproof images, without adjustment, were not dramatically different from the original.

So this leads me to guess (only guess) that one of two things is true: the "RC Brilliant for pigment inks" is an odd paper, or the ICC profile isn't correct for the combination of that paper and your particular printer.

However, I may be misunderstading what you did.

[LenR]

Some interesting points of view in this tutorial...................

http://digitaldog.net/files/Why_are_...s_too_dark.mp4