Soft Proof with Profile that is Slightly Off

[jethrodesign]

HI, we have a bit of a unique case use that I'm looking for some tips or advice on. Trying to get a better soft-proof match to final output from a specific vendor.

We are outputting large panels with historic text & images onto 3M vinyl with a matte laminate. For this project, all photos are saved as grayscale, then colorized in InDesign in CMYK (using foreground/background color swatches). Some of the colors are very light, almost 'pastel' colors, that shift a great deal visually with a small percentage movement in value.

To properly set colors, we asked our print vendor to provide a profile for us based on the device that will be outputting, on the 3M vinyl we'll be using, and with the matte laminate applied.  They sent us an initial profile that just skewed colors a bit more than I had expected, so they made a completely new profile, making sure to laminate the swatch patch before measuring and profiling. So I have to assume this profile is as good as we're going to get from them, and 'should' be accurate if done correctly, right?!?

Unfortunately, the color test we ran using that profile (InDesign doc assigned & retained when making PDF/X) was slightly off from what we're seeing on screen when soft-proofing. My monitor is a new NEC PA242W, that has been properly profiled using i1, so it should be good.

Compared to screen, the proof is about 2-4% light on magenta in midtones, and about 2-5% heavy in yellow overall. Apart from that, it looks pretty good (levels & detail look good). I tested by pulling 2-4% magenta, and adding a bit of yellow, to our existing swatches. This caused the colors on screen to match pretty closely. Unfortunately, a 2-4% magenta shift on the light colors were using makes a pretty dramatic difference in how the color appears (e.g., 'brownish' becomes 'greenish', 'blueish' becomes 'tealish', etc.).

- So to get accurate output, I've had to add about 2-4% magenta, and reduce a bit of yellow, to our colors in InDesign. I know this is not ideal, but it 'should' give us the actual colors we were shooting for.

But now our on screen preview is obviously off and a bit hard to work with.

- Apart from continuing to ask our vendor for new profiles, which may make no difference, anything I can do to at least allow my soft-proofing to be a bit more accurate to reality??

- I was thinking that if I could do a minor tweak to a copy of their profile (pull 2-4% magenta), I could use that just for soft-proofing. But the thread I found on editing profiles seems to indicate it's only possible with very old, outdated software (that I prob couldn't get and/or run).

- Any other tips or ideas????

THANKS!!!

[TheDigitalDog]

jethrodesign  wrote They sent us an initial profile that just skewed colors a bit more than I had expected, so they made a completely new profile, making sure to laminate the swatch patch before measuring and profiling. So I have to assume this profile is as good as we're going to get from them, and 'should' be accurate if done correctly, right?!?

Not necessarily! And ICC profiles have two tables: one that affects the output, the other that affects the soft proof. Further, soft proofing is only as effective as how you've calibrated our display and view a proof next to that display. See:

Why are my prints too dark?

A video update to a written piece on subject from 2013

In this 24 minute video, I'll cover:

Are your prints really too dark?

Display calibration and WYSIWYG

Proper print viewing conditions

Trouble shooting to get a match

Avoiding kludges that don't solve the problem

High resolution: http://digitaldog.net/files/Why_are_my_prints_too_dark.mp4

Low resolution: https://youtu.be/iS6sjZmxjY4

[jethrodesign]

Hi, thanks for the reply and the great video tutorials. I'll have to go through them more in-depth when I have a moment.

My monitor has been calibrated (using SpectraView II and the i1 Display) to a target of D50 / 90 cd/m2. We've found that this provides the most consistent overall match for most offset printing we do with shops that are G7/GRACoL calibrated. I also keep a more standard sRGB profile setup for doing web work.

And while, unfortunately, we do not have a viewing booth setup here (these proofs would be too large anyway), we do use very high quality daylight-balanced CFL bulbs with a high CRI value for our lighting. We also have large frosted front windows that get very good non-direct daylight (SoCal) most of the time that we can walk over to (so kind-of a giant lightbox).

So I'm hopeful that our viewing conditions, while maybe not 100% accurate, should be pretty darned close for what we're doing (as is generally the case). And I doubt that creating a monitor profile with a drastically different white point would compensate for the primarily magenta shift we're seeing, though I of course may be wrong. As mentioned, the overall levels DO look pretty accurate. It's just a specific hue shift.

- So you're saying that the profile they sent us may have tables that incorrectly affect the display on screen??

- If so, is this something they would have had to handle when creating the output profile??

- Any way we can modify it every-so-slightly to get better on-screen soft proofing???

Thanks again for any input here!!!

[TheDigitalDog]

jethrodesign  wrote

- So you're saying that the profile they sent us may have tables that incorrectly affect the display on screen??

I'm not saying that. I'm saying that is always a possibility and that you need to calibrate the display for a visual match of the soft proof as outlined in the video.

[Test Screen Name]

Have you tested your "soft proofing" setup with some other profiled printer? If another printer is similarly wrong, you need to revisit your setup. There are other factors you didn't mention, like viewing conditions. And, given the undoubted expense, and the limitations of comparing RGB and CMYK views of the same colour, have you considered a hard proofing solution?

[jethrodesign]

Hi. Thanks for the replies. I did mention in my last post (a couple above) about my viewing conditions and that my soft-proofing is usually pretty good when working with a vendor that is actually color-managed well (not all are!), as we do majority offset press jobs.

I'm not saying that. I'm saying that is always a possibility and that you need to calibrate the display for a visual match of the soft proof as outlined in the video.

So it's looking like I need to try creating a new monitor profile, choosing different target values, to try to get my monitor to better match these proofs, correct?? The video talks mostly about correcting for level differences, which in my case are not the problem (levels match pretty well). So I'm left with just white point, correct??

I'm just thinking that adjusting color temperature for white point is going to shift things warmer (more yellow) or cooler (more blue). But I need to just reduce magenta in the midtones, which I don't think those moves would do.

- Any tips on how I would specify my target when creating the profile to specifically reduce magenta?!?

In SpectraView, I did try editing my current monitor profile, then clicking to edit White Point. I see the 'Visual Match' area where I could adjust RGB values, but it's grayed out so I can't do anything. Would boosting green here theoretically reduce magenta for me if I could adjust those things???

Thanks again!!!

[TheDigitalDog]

jethrodesign  wrote

So it's looking like I need to try creating a new monitor profile, choosing different target values, to try to get my monitor to better match these proofs, correct?? The video talks mostly about correcting for level differences, which in my case are not the problem (levels match pretty well). So I'm left with just white point, correct??

I'm just thinking that adjusting color temperature for white point is going to shift things warmer (more yellow) or cooler (more blue). But I need to just reduce magenta in the midtones, which I don't think those moves would do.

New calibration and resulting profile. A calibration that better targets the match to the proof. Which is why those with displays like SpectraView, which can utilize multiple calibrations and profile (on the fly) are in better shape to 'tune' the calibration for this visual match to one specific output.

Calibration of the Magenta/Green axis is the 'opposite' of White Balance which is Yellow/Blue. Not all software products provide this level of control. SpectraView does! In custom white point, you have control over x/y axis where y is green.

[jethrodesign]

Calibration of the Magenta/Green axis is the 'opposite' of White Balance which is Yellow/Blue. Not all software products provide this level of control. SpectraView does! In custom white point, you have control over x/y axis where y is green.

OK, that makes sense. So, as editing the x/y coordinates is totally new to me, could you give a very rough 'guestimate' of what I'd need to do to cause a roughly 2-4% reduction in magenta?!? I know you can't give anything close to concrete, but even knowing whether to increase or decrease it, and by .5 or 50, would get me going in the right direction.

Thanks again for the continued help here!

[TheDigitalDog]

You're going to have to adjust using trial and error.

[jethrodesign]

You're going to have to adjust using trial and error.

OK, thanks. And since you just happen to be familiar with SpectraView, and it's still pretty new to me, do I just select one of the target settings in the drop-down, click to edit, then click to edit white point?!? Then make changes to the 'y' value (to increase green, should I go up or down?!?). And lastly, once done, calibrate the monitor to this tweaked setting??

I appreciate any insight here, as I've only ever done a more straight-forward calibration.

Thanks!

[TheDigitalDog]

IF the green/magenta is the issue, Go into the Custom White Point dialog and edit the Y value OR you can try using the sliders below it but I find it's a less subtle and controllable adjustment with those sliders.

[jethrodesign]

Thanks again for the help! I've been playing with it a bit, and reducing the green slider in the 'Visual Match' area allowed me to get the hue much closer after a few tries. Unfortunately, the calibrating takes a loooong time, so hard to make quick minor tweaks.

So now the hue is much closer, but the colors on screen are just overall more saturated. This affects how some of the light quarter-tones appear.

I tried setting the Contrast Ratio all the way down to 150:1, but not sure how much that's doing. Would setting a custom Gamma value allow for better contrast adjustment in the midtones, allowing it to appear less 'contrasty' or 'saturated'?? I'd have no idea where to go on that one...

THANKS!

[TheDigitalDog]

You can speed up the calibration process, just for this rough work, by going into preferences and setting Calibration and Profile steps to the lowest value (8), then up it when you've found the magic settings that work.

Contrast Ratio:

X-Rite i1Display Pro Advanced Features | Contrast Ratio with Coloratti Andrew Rodney - X-Rite Photo ...

Leave the gamma setting alone, it will play no role really.