What exactly does 'Simulate paper white' do?

[Chris Kern]

I keep the simulation option OFF when I decide which rendering intent per image I prefer.

Interesting.  Why?

[digitaldog]

Interesting.  Why?

easier to see.

[Chris Kern]

easier to see.

Ahhh, apparently yet another instance where I misremembered (and have been slavishly following my faulty recollection of) something I read in Jeff Schewe's book, The Digital Print.  The rationale is explained on p. 149 of the paper edition:

With this option selected, instead of using a [relative|perceptual] colorimetric, Lightroom uses an absolute colorimetric that shows the maximum black and the maximum white of your actual paper.

So I gather I should choose the rendering intent first to get the better color match, and only then—in Schewe's delicate phrase—‟make my image look like crap,” and tweak the contrast and clarity adjustments.  Is that right?

[digitaldog]

So I gather I should choose the rendering intent first to get the better color match, and only then—in Schewe's delicate phrase—‟make my image look like crap,” and tweak the contrast and clarity adjustments.  Is that right?

Yes. Without simulation is ideal for picking a rendering intent or editing outside of full screen mode. Then turn it on, that's your most 'accurate' view of what the print should look like, again under the assumption the preview table in the profile is good and sync's up with the output table.
And IF you turn your head away for the display when invoking the simulation, give your visual system a second to update, then look back, it doesn't look as much like crap as Jeff reports when watching the simulation update on screen. LR does a better job here in this respect! Dark bkgnd, gradual change of the simulation and those cool output specific edits applied to a virtual copy!

[Chris Kern]

Dark bkgnd, gradual change of the simulation and those cool output specific edits applied to a virtual copy!

Got it!  Thanks.

[Hening Bettermann]

>>How could I know the validity of my monitor calibration?
Andrew:
> It matches the soft proof.

I don't understand. My monitor matches the soft proof in the sense that the 2 or 3 papers I have compared so far look all the same in color, just differ in contrast. Nowhere do I see a color change anywhere near the one that TonyW has demonstrated. That would mean my monitor calibration is close to perfect? And Tony's is not? ??

[digitaldog]

>>How could I know the validity of my monitor calibration?
Andrew:
> It matches the soft proof.

I don't understand. My monitor matches the soft proof in the sense that the 2 or 3 papers I have compared so far look all the same in color, just differ in contrast. Nowhere do I see a color change anywhere near the one that TonyW has demonstrated. That would mean my monitor calibration is close to perfect? And Tony's is not? ??

Perhaps, without seeing both setups, it's just speculation at this point. AND contrast SHOULD also match to each paper, so many need multiple calibrations and profiles for each paper! Which is why something like a SpectraView is so nice. It allows you do update multiple calibrations and profiles on the fly based on what you're soft proofing.

[Doug Gray]

This is what I did back when I was first learning about printer profiles, monitor profiles and such.

Since ICC profiles are made to a D50 Illuminant, I profiled my monitor to D50. And for convenience in comparing colors, I set the monitor white point to 100 Nits (cd/m^2). Then, when reading the emissive color with an I1  Pro spectro, one should get a good match between the LAB color in Photoshop with the picker info tool and the spectro reading. This setup makes it easy to measure, not just "see," what happens when you soft proof.

When I put a white (255,255,255) patch on the monitor and measure it with a spectro I get pretty close to LAB 100,0,0. Then, if I select show paper white the luminance drops and the LAB reading drops as well to the paper white described in the paper profile.

If one has a "perfect" printer paper that reflects 100% of the light, and is therefore neutral by definition, then the LAB reflectance reading would be 100,0,0.  Show paper white would result in no change at all! No paper does this though some PTFE based proofing papers come close. Most papers like Epson Ultra Premium Glossy (similar to Costco's) are about 85% reflective with a lot of OBA and is LAB 95,1,-9 (M0) The closest I have to a "perfect" paper is a Baryta paper that is LAB 97,0,0 (M0) which is about 91% reflective. The Baryta paper shows less brightness reduction with "show paper white" and, of course, physically looks brighter next to the Costco.

OBAs are another factor. In the case of the OBA laden Costco glossy, using a profile made with M0 (uV but about 30% less than what true D50 contains) the LAB value measured on the screen drops to 95,1,-9. If I just read the reflectance LAB value directly off the paper then I also get LAB 95,1,-9.  That's a pretty significant blue shift.

But, I don't actually see that kind of blue shift when comparing Baryta with Costco Glossy. Unless I take the papers outside and view them in daylight. Then I get the blue shift seen when selecting "show paper white" with the Costco but no blue shift with the Baryta. They look vastly different.  The reason is simple. My proofing setup currently uses a good quality LED illuminant and has no significant uV. For that reason I make profiles using M2 (uV cut) unless making prints for displaying outside.

[Doug Gray]

>>How could I know the validity of my monitor calibration?
Andrew:
> It matches the soft proof.

I don't understand. My monitor matches the soft proof in the sense that the 2 or 3 papers I have compared so far look all the same in color, just differ in contrast. Nowhere do I see a color change anywhere near the one that TonyW has demonstrated. That would mean my monitor calibration is close to perfect? And Tony's is not? ??

The change TonyW demonstrated is quite typical of what one sees with high OBA content paper. However, most people, including me, do not have hard proofing stations, or even normal room light, with significant levels of uV so we don't see the amount of change on the physical prints that "show paper white" produces. But you will see that kind of shift with higher levels of uV. A good way to see for yourself is just bring two sheets of blank paper, one with high OBAs the other without, outside. You will see exactly that kind of blue shift on the OBA paper.

Since I don't usually have prints illuminated with significant uV, even on my hard proof setup, I use profiles made with M2 (uV cut) and avoid OBA paper on critical prints.

[TonyW]

The ICC profile I chose was for a Canon TS8000 series (glossy type) an entry level A4 inkjet that I use for everyday printing duties.  This was not the worst looking simulation but just one to show enough difference after being squashed into a Gif.

As I think you are going to an outside source for printing perhaps this example may be of interest.  This time using the Master and Proof preview which somewhat lessens the shock looking side by side?

This time with a profile provided by a lab I have used on occassion for a Fuji Frontier glossy paper.  I decided I liked the Perceptual rendering best and with a small tweak or two (contrast and saturation green/yellow) should print nicely using this lab service.

[Hening Bettermann]

Thank you for your contributions. I will study them more closely when I'm back home in some days.

[Hening Bettermann]

So now I'm back home with my calibrated monitor.

Andrew,
I still don't understand in which way the soft proof should match the monitor profile. Everything I see is displayed through the monitor profile, right? Now, in PhotoLine, I switch the softproof on. Now PL chooses the paper profile. The image changes in the exspected way, it becomes less contrasty. I had exspected it also would change in color. You say BOTH the contrast and color should match the monitor profile. Then there would be no change, the master would be the same as the softproof, meaning the soft proof was moot?? Please enlighten me...

Doug Gray and TonyW:
My take away from your posts is that I have to choose a paper without OBAs, otherwise the soft proof is useless for me.

Thank you all for your contributions!

[digitaldog]

So now I'm back home with my calibrated monitor.
Andrew,
I still don't understand in which way the soft proof should match the monitor profile. Everything I see is displayed through the monitor profile, right?

Depends on the soft proof settings of course. I explained what RI is used depending on the selection. You can decide to see a paper white or ink black simulation depending on the settings.

[Robert Boire]

At the risk of reanimating this discussion, my own experience is that I get a better print to screen match if I do NOT select simulate paper white in LR when soft proofing. The loss of contrast on the actual print as compared to the screen simulation is never as dramatic.  I have never really understood why and the discussion above particularly wrt OBA's leaves me even more confused.

My situation... I use EPSON premium lustre or EFP which both have OBAs. I soft proof ( and print) using the generic EPSON profiles. I do not use a viewing booth. I use à Solux 4700k lamp (which I understand is really calibrated to simulate the D50 standard) and my monitor is calibrated to a D50 white point. I work in a darkened room and I am careful to ensure that the lamp does not spill onto the screen ( I use a lens hood on a NEC monitor using SpectraView).

Given that the Solux has a very low UV intensity and is presumably not exiting the OBAs, I would have thought their presence would have very little difference on the white point simulation? Or am I to assume that the printer profile in fact assumes that it is being observed in daylight with UV?

BTW, I recently tried checking simulate paper white with EPSON radiant white watercolour paper. In this case the soft proof IS a better match.

Thoughts anyone?

[digitaldog]

At the risk of reanimating this discussion, my own experience is that I get a better print to screen match if I do NOT select simulate paper white in LR when soft proofing.
Thoughts anyone?

Simulate has more to do with more than just OBA's! It's about the color of the white and it's 'density' compared to the white of the display. Same with black (it's not a black hole where the display attempts to emit as little as possible). It's about the contrast ratio of the two! You've got a print with maybe a 300:1 contrast ratio on a display that may have a ratio of 1000:1. Now maybe you're calibrating with software that allows control over black and thus the display contrast ratio but not everyone does. All of this works or fails depending on the quality of the profile tables as well; they have two. One for soft proofing, one for output. They should be in sync so to speak. Not all ICC profiles are created equally.

[Robert Boire]

Simulate has more to do with more than just OBA's!

I realize that. However I was under the impression that there was a suggestion that you agreed with (see Doug's input from sept 20 above and your response and the discussion of UV light) that OBAs somehow limited the accuracy of the simulation. That is the part I do not understand.

BTW why do you refer to two profiles? Isn't the paper/ink profile used for both the soft proof (of the paper) and the output (on the paper)?

[Doug Gray]

I realize that. However I was under the impression that there was a suggestion that you agreed with (see Doug's input from sept 20 above and your response and the discussion of UV light) that OBAs somehow limited the accuracy of the simulation. That is the part I do not understand.

BTW why do you refer to two profiles? Isn't the paper/ink profile used for both the soft proof (of the paper) and the output (on the paper)?

There are a number of problems associated with soft proofing papers with high OBA content. The prepress industry has standardized on D50 levels of uV but viewing stations with the required uV level are not cheap and it requires that profiles be made using M1, which produces spectral values that include the fluorescent component from D50 uV levels.

Most photographers use profiles made with M0, which, in my experience, has about 2/3s the uV of M1. And current viewing booths very often have even lower uV levels, sometimes almost no uV.  Soft proofing with M0 profiles and high OBA paper will produce a significant tint shift. One effective way to handle this is to have different monitor white points, chromaticity and luminance, that are adjusted to make viewing booth paper white have the same apparent luminance and "white." Then, not selecting show paper white but selecting show paper blacks gives the best results. This is most easily done when your setup lets you switch monitor colorspace settings for whatever paper you are using. It makes M0 profiles work well without controlled viewing booth uV.  I believe this is what Andrew does in his videos. It's probably the best solution for photographers

However, if your viewing booth is compliant with prepress ISO viewing booth standards (2009 version, IIRC) and you use M1 profiles then the best results are using show paper white. Then the changes in tint as well as luminance (the amount white paper actually reflects) will match closely without switching monitor settings.

The approach I use for critical color work is to simply avoid OBA papers. When you do that then show paper white works great (look away when you do it because you have adapted to a higher luminance). And this works with downloaded profiles or ones made where you are limited to M0.

If OBAs are unavoidable and your prints will be displayed in a place you have access to and has consist lighting but unknown uV content then XRite's OBC approach is excellent. You compare neutral printed patches with spectrally neutral references under the lighting used for print display to create a special profile used for printing.

All of which is why I avoid high OBA papers for critical work.

[digitaldog]

I realize that. However I was under the impression that there was a suggestion that you agreed with (see Doug's input from sept 20 above and your response and the discussion of UV light) that OBAs somehow limited the accuracy of the simulation. That is the part I do not understand.

BTW why do you refer to two profiles? Isn't the paper/ink profile used for both the soft proof (of the paper) and the output (on the paper)?

OBA's produce a number of issues with and without soft proofing and much depends on the illuminant, the OBA content etc. It's just a factor that is uncontrollable.

OBA's are just bad news!
http://digitaldog.net/files/24TroubleWithFWAs.pdf

Why do I prefer what two profiles? One profile has two tables; one affecting the actual converted values, the other the soft proof. They can be out of sync or just not "accurate' (they don't provide a good output or simulation or both).

[Robert Boire]

Thanks Doug and Andrew for educating me about OBAs. I admit I had no idea what M0 and M1 are about but this post http://fieryforums.efi.com/showthread.php/5449-M0-M1-M2-measurement gave me a pretty good idea. I understand that OBAs are bad news and the accuracy of the soft proof depends a lot on the profile etc

But at the risk of flogging a dead horse, I come back to my original point. I have no idea if my profile for EFP, which I downloaded from the Pixel Genius site was calibrated with M0 or M1 (maybe Andrew knows), but I do know that with my Solux lamp I do not see (perhaps I am not looking hard enough) any significant color shift, blue or otherwise on the print as compared to the soft proof: I just see an exagerated reduction in contrast when simulating paper white in the soft proof. On the other hand the soft proof without simulating paper white is really quite good, both for color and contrast.  So is this just serendipitous.. or is there some technical reason?

[Doug Gray]

BTW why do you refer to two profiles? Isn't the paper/ink profile used for both the soft proof (of the paper) and the output (on the paper)?

I realized neither Andrew or I answered this question.

Andrew wasn't talking about two profiles, but the two tables in a profile, one of which is used for rendering color to the printer, the other takes the RGB values from that rendering to generate the actual colors printed. For instance printing white (Lab=(100,0,0)) results in no ink but the second profile table can determine the actual color of paper white which is often something like Lab=(95,-2,-1) or so which is a dimmer, and slightly more bluish color. When you select show paper white, the (95,-2,-1) values are what gets sent to your monitor after conversion to whatever monitor colorspace you have. When you don't, the values are scaled to paper white and the "proof" return values will be (100,0,0) or monitor white. That's why you see the drop in luminance selecting show paper white.

Unless you have a hard proof viewing station right next to your monitor with the luminance set so that a perfect paper white has the same brightness and chromaticity as the monitor you are better off not selecting show paper white. And this is without the confounding factors of OBAs which just makes life even more interesting.