Dynamic Range mapping, file to pirnt.

[eronald]

How many stops DR in a print? Does the fact that we try to compress a display image down into a print have anything to do with the "my print is too dark" syndrome?
Should one try to clip shadows/highlights before printing, or apply a brutal S curve?

Edmund

[digitaldog]

The most you'll see on a glossy print is approx 350:1.
Yes: that's why we have Simulation of paper white/ink black when soft proofing, no, nothing to do with “my prints look too dark.”

[eronald]

The most you'll see on a glossy print is approx 350:1.
Yes: that's why we have Simulation of paper white/ink black when soft proofing, no, nothing to do with “my prints look too dark.”

Andrew,

Let's stay away from glossy prints and all the specular issues, see below. What do you expect  from matte art paper?

And then, how do you think the simulation gets you out of the issues that you are shoehorning a lot of DR into mucho less DR? Perceptually you can simulate (actually I'm not so sure you can but we can go there later), but the DR still isn't there so the viewer sees a compressed image or one where you lop of the head and/or the feet.

Edmund

PS. I had a really interesting "glossy paper" experience the other day. I wanted to test my iSis XL after taking it out of storage, so I did a profile for a 15 year old glossy paper (Tetenal PG) which I had lying around in 13x19, and then a print of a Sargent painting to give a friend. Print was dark in my room in the evening. Histogram was ok etc etc. So as my whole print process is nailed down,  I assumed that if the file was ok, and it looked ok on the screen, then this was the best I would get. So I took the print to my friend, and we looked at it in the rather dark backroom of the restaurant, away from the windows and it was rather dark. Then when we walked out, the winter sun came out and I held the print vertically  at arm's length in front of us with the low winter sun behind us and, as you guessed the print was _glorious_. I held the print horizontally, lit by the cloudy sky,  and we looked down at it, and it was a bit dark again. Did you say specular? Which was the right view of the print? The one you would get in a "sales gallery" with the direct  spotlights shining on it, or the one you would get with it mounted on an indoors office wall with diffuse ambient illumination? I would assume the instrument was "seeing" the print rather more as directly lit than diffusely lit, and therefore constructed a profile for those viewing conditions.

PPS I apologize to fellow forum members having to watch this train wreck - my IQ is going down with age, and my knowledge of color was always a bit amateurish, I'm a blogger, not a color scientist, after all.

[digitaldog]

You have a Spectrophotometer; make your own measurements on whatever paper you desire. Conversion of spectral to density is a snap in the right software. You must have BabelColor CT&A. Then, when you have white vs. black measurements, you do the math and get the ratio.

[Kyle Jackson]

Let's stay away from glossy prints and all the specular issues, see below. What do you expect  from matte art paper?

And then, how do you think the simulation gets you out of the issues that you are shoehorning a lot of DR into mucho less DR?

Well matte is even worse range ofc. Black point is less dark. White point is typically duller without brighteners, etc.

Simulation doesn't get you out of anything, it simply gives an estimated preview of what's going to come out of the printer. It's still up to us to decide what adjustments we want to make to the image before printing it, whether it's for tone or colour gamut or contrast or whatever.

Colour mgmt doesn't replace the creative process. If the preview indicates you're going to lose a bunch of highlights and shadows and that's important to you, then you get to decide what changes to make to the image to suit your tastes. (Just like you do with exposure when capturing the image.)

[Czornyj]

So I took the print to my friend, and we looked at it in the rather dark backroom of the restaurant, away from the windows and it was rather dark. Then when we walked out, the winter sun came out and I held the print vertically  at arm's length in front of us with the low winter sun behind us and, as you guessed the print was _glorious_. I held the print horizontally, lit by the cloudy sky,  and we looked down at it, and it was a bit dark again. Did you say specular? Which was the right view of the print? The one you would get in a "sales gallery" with the direct  spotlights shining on it, or the one you would get with it mounted on an indoors office wall with diffuse ambient illumination? I would assume the instrument was "seeing" the print rather more as directly lit than diffusely lit, and therefore constructed a profile for those viewing conditions

Instrument doesn't "see" anything, as colors physically don't exist. Profiles are constructed for D50 500 lux viewing condition.

[eronald]

Instrument doesn't "see" anything, as colors physically don't exist. Profiles are constructed for D50 500 lux viewing condition.

I meant instrument geometry

[eronald]

You have a Spectrophotometer; make your own measurements on whatever paper you desire. Conversion of spectral to density is a snap in the right software. You must have BabelColor CT&A. Then, when you have white vs. black measurements, you do the math and get the ratio.

Andrew,

Thank you for these very precise instructions. Following them, here is what I arrived at:

I used Babelcolor CT&A and an i1Pro2  to measure the black on Epson 8550  testprints on Epson Premium Glossy, on a luster and a matte coated PermaJet paper which were supplied to me for profiling.
I got D= 2.4,  D=2.2 and D=1.8 roughly as density, and 0.05 roughly for white.

Also, from  the formulas D= log10 Rem, with Rem the remission (ratio), I deduced one can convert to photo stops by S= (Ln(10/Ln(2))*D= 3.32*D.
Whence I get roughly 8 stops for the glossy, 7.5 for the luster, and  6 stops for the matte, the latter a considerable difference.

I then used a rather mediocre but very adjustable Dyson lamp to evaluate the look of the prints, setting it to a warm light.
- When the light is set to high intensity, the prints look very similar and the matte print has a bit of the depth of matte paper.
- When the light is set to a lower intensity in this fairly dim living room, the matte print looks a bit dark and low contrast  to my eyes. No fun there. Especially the shadow colors go bad.

that was today's experiment.

I assume the numbers look different for high quality pigment printers and good paper, neither which I have available just now, but my take is still that print preparation really needs to the final print DR into account, and that cries of "my print is too dark" can follow from excessive DR compression.

Moral of the story, Photoshop will display what's in the file in the monitor gamut,  and with profiles will print wonderfully, but they cannot magically create contrast in print, it's up to the user to tailor the file to the print medium.

I vaguely remember that in the days of photo paper prints, one bought "hard" or "soft" paper in order to assist with said adjustment.

Edmund


 

[digitaldog]

Moral of the story, Photoshop will display what's in the file in the monitor gamut,  and with profiles will print wonderfully, but they cannot magically create contrast in print, it's up to the user to tailor the file to the print medium.

The moral of the story is that neither profiles, Photoshop, nor anything else can predict the illumination of a print. Prints that are not too dark appear too dark with insufficient illumination. Prints illuminated with far less light than a display emits can appear darker (too dark). Contrast (contrast ratio) and dark prints are totally separate entities.

[eronald]

The moral of the story is that neither profiles, Photoshop, nor anything else can predict the illumination of a print. Prints that are not too dark appear too dark with insufficient illumination. Prints illuminated with far less light than a display emits can appear darker (too dark). Contrast (contrast ratio) and dark prints are totally separate entities.

Who knows, maybe in the future smart glasses will afford better previews ... somewhere over the rainbow there is a land ...
I feel like I've gone on a really interesting sidequest, but the princess is still locked up in the dungeon

Edmund

[Czornyj]

I meant instrument geometry

I think you're overthinking it. Open the chart in PS, assign the profile, and match monitor white brightness to the print white brightness, and you'll clearly see if the instrument geometry is the problem (which is highly doubtful)

Or if you want to make profile brighter, use CIECAM02 gamut mapping in ArgyllCMS, or CoPrA / MYIROtools profiler that allows you to modify mapping lightness

[eronald]

Instrument doesn't "see" anything, as colors physically don't exist. Profiles are constructed for D50 500 lux viewing condition.

Czorny,

Belatedly, thank you very much for this link.

Edmund

[kers]

...Profiles are constructed for D50 500 lux viewing condition.

thanks for this info