Matching Print and monitor: the last little bit

[Tim Lookingbill]

Of course if one is working in the evening this great solution is not possible.  An inexpensive approach is to get a clamp-on Solux fixture and put it next to the monitor (https://www.solux.net/cgi-bin/tlistore/clampon.html ) this way you have a uniform light source that can be used to view the printed image and adjust the monitor's parameters until you get as close a match as possible.  Although viewing booths are perhaps best, you cannot easily get the monitor next to it and it is a more costly option.

I should've paraphrased by pointing out my advice was meant only as a method to check if the amount of light is affecting the look of the print. This LuLa B&W Master Printer article... http://www.luminous-landscape.com/tutorials/bw_master_print.shtml

...demonstrates a similar effect light reflectance plays on print viewing with regard to "luminosity" vs "luminance" that affects whether the print looks dim, dull or vibrant. "Dark" is a way too subjective and vague description.

Note: the picture of the egg's luminance vs luminosity which can be affected by the amount of light the print is viewed under. Color saturation reflectance acts the same way depending on the quality of light the best being full spectrum daylight or the Solux bulb.

This discussion... http://www.luminous-landscape.com/forum/index.php?topic=79948.msg644973#msg644973

...demonstrates how changing available room lighting either from a window or turning on a center room light makes a print look dark or vibrant. This is to show that it's best to change the light and not have to go to the trouble of changing an established calibrated and color managed editing environment.

Here's a demonstration of how light affects vibrance, contrast and luminosity using my own print off of a $50 Epson "All In One" on Epson Ultra Premium Glossy paper compared to the original file on top.

[Slobodan Blagojevic]

The problem with comparisons like that, Tim, is that they are, well... comparisons. In absence of something to compare it with, your "window blind light" would look the same as the other two, after our eyes (and brain) compensate for less light.

As someone above already stated, the best thing is to look at the print in the most likely light it is going to be displayed and see if you like it, not to use the screen as the ultimate criterion, but just a starting point. Yes, calibrate it properly; yes, soft proof; yes, use the good lighting to match it, etc., but ultimately, look at the print and see if you like it.

[Tim Lookingbill]

The problem with comparisons like that, Tim, is that they are, well... comparisons. In absence of something to compare it with, your "window blind light" would look the same as the other two, after our eyes (and brain) compensate for less light.

As someone above already stated, the best thing is to look at the print in the most likely light it is going to be displayed and see if you like it, not to use the screen as the ultimate criterion, but just a starting point. Yes, calibrate it properly; yes, soft proof; yes, use the good lighting to match it, etc., but ultimately, look at the print and see if you like it.

The "window blind light" image was to show the effects of changing levels and character (diffused or spot) of light has on a print compared to viewing on a calibrated transmissive display. The B&W Master Print "Luminance vs Luminosity" linked article was demonstrating the same effect except how it's rendered through edits on a real image viewed on a display.

It's a characteristic of light we tend to edit into our images (especially Raw) to make them appear 3 dimensional or "pop" on screen but can be lost on a print because we aren't viewing the print under the same level and character of light that got us to make the image "pop" on screen. This is basic digital darkroom editing to take advantage of the dynamics of a transmissive display. You don't want to dumb that process down by reducing the dynamics of the display.

And the OP isn't talking about "liking" the print results. They're wanting their print to match their screen edits. If I can get a match off a $50 printer of an image "I like on screen" then anyone else can do the same as long as the print viewing light matches.

You can always edit the image on the screen to look lighter to compensate for viewing the print in lower lighting situations but it will never have that 3D "pop" as it appears on screen because prints have a severely reduced dynamic range compared to a display and viewing the print in much lower light reduces those dynamics even further.

[dwswager]

Right but you expect that base adjustments (on what I presume you mean the RGB working space) is done based on what you see on-screen right? Backed up by numbers and/or Histograms to a degree. Further, you expect soft proof to look like print using a profile. There's no role here for the sliders. It's a kludge. The print is or isn't too dark or appears too dark for another reason and that needs to be addressed.

I was reading the Luminous Landscape article linked above in this thread.  To my mind, most prints are too dark because most monitors are too bright.

I have 2 question:

1. In the old days (I actually owned an original Epson Stylus Color printer circa 1994) it was customary to set the output levels of a file sent to a printer to bottom out around 5 and top out around 250-252.  Is that still standard practice or has Color Management handled that?

2. In the article you state that basically you want to calibrate your monitor to the same point as the print viewing locations (brightness, white point, etc.) and then profile it and build the ICC profile.  I understand why, but then when I soft proof using that same paper/ink as the monitor settings I really shouldn't see a significant difference?

What most people want CMM to be that it isn't is to be able to set the monitor to calibration X and profile it.  Then when given a good monitor profile and printer profile, they want the CMM engine to take RGB values R1, G1, B1  that look like X on the monitor and translate it to (simpifying) C1, Y1, M1, K1 that looks like X.  But, of course, a print can never look like the monitor due to the limitations of the ink and paper medium to translate to the higher standard of a monitor.   Hence we softproof to basically dumb down the monitor to ink/paper level as best as possible.

[digitaldog]

1. In the old days (I actually owned an original Epson Stylus Color printer circa 1994) it was customary to set the output levels of a file sent to a printer to bottom out around 5 and top out around 250-252.  Is that still standard practice or has Color Management handled that?

No, that's another kludge!

2. In the article you state that basically you want to calibrate your monitor to the same point as the print viewing locations (brightness, white point, etc.) and then profile it and build the ICC profile.  I understand why, but then when I soft proof using that same paper/ink as the monitor settings I really shouldn't see a significant difference?

Soft proof view should match the print as closely as the technology you embrace allows, simple as that.

What most people want CMM to be that it isn't is to be able to set the monitor to calibration X and profile it. 

There is a big difference between calibration and profiling.

Then when given a good monitor profile and printer profile, they want the CMM engine to take RGB values R1, G1, B1  that look like X on the monitor and translate it to (simpifying) C1, Y1, M1, K1 that looks like X.

  And if they follow the article and use the proper tools, that's what they see (Soft proof view should match the print as closely as the technology you embrace allows). There is calibration and there is profiling and one doesn't happen until the first is accounted for.

But, of course, a print can never look like the monitor due to the limitations of the ink and paper medium to translate to the higher standard of a monitor. 

Nor could a Polaroid and a piece of transparency film. But the closer you get, the more predictable the final process becomes. A display and a print can look pretty darn close. Such that what you see and accept on the display is what you get and accept on the print. That's all color management is (number management). The Polaroid and the film were analog processes, the match or mismatch was what it was. We don't suffer those limitations thanks to digital color management even if the final results are not prefect. They are almost always better, closer to a match with color management than without!

[dwswager]

No, that's another kludge!
Soft proof view should match the print as closely as the technology you embrace allows, simple as that.
 There is a big difference between calibration and profiling.
  And if they follow the article and use the proper tools, that's what they see (Soft proof view should match the print as closely as the technology you embrace allows). There is calibration and there is profiling and one doesn't happen until the first is accounted for.
 Nor could a Polaroid and a piece of transparency film. But the closer you get, the more predictable the final process becomes. A display and a print can look pretty darn close. Such that what you see and accept on the display is what you get and accept on the print. That's all color management is (number management). The Polaroid and the film were analog processes, the match or mismatch was what it was. We don't suffer those limitations thanks to digital color management even if the final results are not prefect. They are almost always better, closer to a match with color management than without!

I final and important question.

I just profiled my display.  The print viewing light was around 112.  I set it to a D55 white point and profiled it.  Then I made a Print of the PrinterEvaluationImage_V002_ProPhoto.tif file using Epson Hot Press Bright on the Epson R2400.  Looking at the monitor and the print, there is a pretty good match.  Brightest areas and darkest areas on the monitor are not quite as bright and dark, and the reds in the strawberrys aren't quite as deep as the monitor, but overall, I'm extremely satisfied.

Problem is that when I turn soft proofing on selecting the "Simulate Paper Color" the colors get muted and the light and dark areas of the monitor come down well under what the actual print looks like.  Basically, I get a better match in normal space than with soft proofing on.  Is that normal?

[digitaldog]

Problem is that when I turn soft proofing on selecting the "Simulate Paper Color" the colors get muted and the light and dark areas of the monitor come down well under what the actual print looks like.  Basically, I get a better match in normal space than with soft proofing on.  Is that normal?

That is normal and it should better match the print.
Got control over contrast ratio? If not, the profile has to do a lot more heavy lifting and hopefully well. Either way, if you have a device with perhaps a 1200:1 contrast ratio and you want to view a print simulation that is 300:1 contrast ratio, something is going to look different when you toggle the two previews.

[dwswager]

That is normal and it should better match the print.
Got control over contrast ratio? If not, the profile has to do a lot more heavy lifting and hopefully well. Either way, if you have a device with perhaps a 1200:1 contrast ratio and you want to view a print simulation that is 300:1 contrast ratio, something is going to look different when you toggle the two previews.

Thanks Dog!

BTW, for an amateur like me, who doesn't get to sit and do this on a consistent basis, some times that the print is dark, it really is the RGB values in the image.  The best thing I usually do is to throw up some sort of reference image (monitor calibration or printer test image) next to my image.  It is amazing in that what looked like a great image is really darker than first thought.  When I do this, there is usually a WB tweak, overall brightness and even a shadow contrast change in the cards.  I find the darkest areas of the image tend to loose their contrast and merge closer in tone to each other.

[Tim Lookingbill]

Thanks Dog!

BTW, for an amateur like me, who doesn't get to sit and do this on a consistent basis, some times that the print is dark, it really is the RGB values in the image.  The best thing I usually do is to throw up some sort of reference image (monitor calibration or printer test image) next to my image.  It is amazing in that what looked like a great image is really darker than first thought.  When I do this, there is usually a WB tweak, overall brightness and even a shadow contrast change in the cards.  I find the darkest areas of the image tend to loose their contrast and merge closer in tone to each other.

I've experienced the same more than I want to admit. I can't tell you how many images I've had to re-edit because I unintentionally made them too dark due to the amount of time I've spent tone mapping a dim, dark and flat looking Raw viewed on a transmissive display causing my eyes to adapt to seeing shadows and contrast as too flat and foggy.

Here's an example I demonstrated in a thread I started about how to gauge the actual density of black from what was seen in the original scene to how it should look viewed on a transmissive display...

http://forum.luminous-landscape.com/index.php?topic=89689.msg731263#msg731263