Pages: 1 [2] 3   Go Down

Author Topic: How many prints does it take?  (Read 9338 times)

langier

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1503
    • Celebrating Rural America, the Balkans and beyond
Re: How many prints does it take?
« Reply #20 on: June 26, 2011, 11:39:14 am »

On a good day when I'm in the grove and don't find nits to pick, what I see is what I get when I click "print" and then compare the print back to the monitor.

By using hardware monitor calibration and then very good print profiles, printer waste is usually caused by human error by using the wrong profile after changing papers, an errant spit of the ink jet nozzle, dropping the print on the floor, etc. With a good workflow, there should be little to no waste!
Logged
Larry Angier
ASMP, ACT, & many more! @sacred_icons
https://angier-fox.photoshelter.com

Sven W

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 514
Re: How many prints does it take?
« Reply #21 on: June 26, 2011, 04:24:01 pm »

I believe it was 642:1 - or somewhere around there. 


I think that's way to high.
The ISO recommendation for softproofing is a contrast ratio of not more than 1:250.
"...... a monitor contrast ratio of approximately 1:250. This is ideal for soft proofing because it
simulates a rich glossy print’s density range of approximately D2.40 (Dmax 2.50 - Dmin 0.10) or
8 EV (2.40 ÷ 0.3 = 8 ) which equals a contrast ratio of 1:256"

/Sven
Logged
Stockholm, Sweden

MHMG

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1285
Re: How many prints does it take?
« Reply #22 on: June 26, 2011, 07:41:43 pm »

I think that's way to high.
The ISO recommendation for softproofing is a contrast ratio of not more than 1:250.
"...... a monitor contrast ratio of approximately 1:250. This is ideal for soft proofing because it
simulates a rich glossy print’s density range of approximately D2.40 (Dmax 2.50 - Dmin 0.10) or
8 EV (2.40 ÷ 0.3 = 8 ) which equals a contrast ratio of 1:256"

/Sven



And a 2.40 dmax translates to an L* value =3.6, which is what you'd want your monitor to display with softproof viewing turned "on". It's not the native contrast ratio that the monitor needs to be able to reach for softproofing to work well.  A subtle fact about most monitor profiles is that they remap the measured blackpoint of the monitor (sort of like relcol with BPC), no matter what the black level is  in reality,so that L*=0.0 always gets assigned to RGB 0,0,0 in the display profile. (Note: ColorEyes display Pro does have an optional "absolute" rendering mode for monitor profiles which will map RGB 0,0,0 to the actual measured monitor blackpoint L* value , but that's an exception rather than the rule).

Thus, the best monitors for softproofing will require a black level which approaches L*=0, not 3.6, and then the monitor black will be increased ever so slightly above that black level to the 3.6 L* value when the softproof mode (in PS, for example) is turned on.  In practice, L* = 0 isn't attainable, but to get to an L*=1.0 level on your monitor, for example, you need a contrast ratio nearly 1000:1, and better monitors can achieve this level if you also take the precaution to shield stray light from hitting the screen.  Once you achieve it, and you then open a digital file with both RGB 255,255,255 whitepoint and RGB 0,0,0 blackpoint, the PS info tool will dutifully tell you that you are seeing L*100 for maximum image white areas and L*=0 for maximum image black areas, but your monitor is only getting down to approximately L*=1.0 relative to the monitor white in reality. Not bad at all but not perfect.

For poor monitors being used in lighting conditions that compound the already poor screen contrast problems, it's not unusual for them to reach only L*= 4 or 5 at native black level relative to monitor white.  Then when you turn on softproofing, the predicted shadow values get lifted on a relative basis above this L* error level, so the softproofed image ends up looking flatter in contrast than the actual print. This situation accounts in part for the "makes my image look ugly" complaints many people express about soft proofing.  Bottom line: you need a good monitor with high native contrast ratio to get the best softproof. It will have to have a native contrast ratio of 750:1 or better still.

cheers,
Mark
http://www.aardenburg-imaging.com
« Last Edit: June 26, 2011, 07:56:10 pm by MHMG »
Logged

gromit

  • Full Member
  • ***
  • Offline Offline
  • Posts: 133
Re: How many prints does it take?
« Reply #23 on: June 26, 2011, 09:18:09 pm »

Thus, the best monitors for softproofing will require a black level which approaches L*=0, not 3.6, and then the monitor black will be increased ever so slightly above that black level to the 3.6 L* value when the softproof mode (in PS, for example) is turned on.

There are two fundamentally different approaches to monitor calibration. The first is where the monitor is calibrated to the paper (white point and contrast range), the second where the monitor is arbitrarily calibrated then soft-proofing (with Simulate Paper Color and Simulate Black Ink) is used to hopefully give an idea of what the printed results will be.

The former has the benefit that the image, once opened, is displayed pretty close to how it will appear on output. Soft-proofing is only required when it is suspected that some image colours may be out of gamut (rarer these days with current generation inksets). It's suited to a production environment where two dozen images need to be printed by lunchtime.

The latter is aimed at hobbyists who are happy to spend forever on a single image and apply (mandatory) edits to image contrast etc in a simulated environment for the paper chosen.

It's debatable which in turn produces the best results.
Logged

MHMG

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1285
Re: How many prints does it take?
« Reply #24 on: June 26, 2011, 09:40:45 pm »


The former has the benefit that the image, once opened, is displayed pretty close to how it will appear on output. Soft-proofing is only required when it is suspected that some image colours may be out of gamut (rarer these days with current generation inksets). It's suited to a production environment where two dozen images need to be printed by lunchtime.

I can see that workflow being used in a lab environment where all output is going to one type of media (e.g. a Lightjet printer printing exclusively on Fuji Crystal Archive paper), but the majority of printmakers I know (pros as well as amateurs) print on everything from matte to glossy to fabric materials. I don't see how one fixed monitor calibration can simulate those very different outcomes, and dedicating multiple monitors to multiple output devices doesn't seem particularly practical, IMHO, but to each his own.
Logged

Mike Guilbault

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1040
    • Mike Guilbault Photography
Re: How many prints does it take?
« Reply #25 on: June 26, 2011, 10:25:47 pm »

... and on top of that, I have to deliver digital image files to commercial clients for web as well as various print uses.  Do you use a different monitor profile for each 'delivery' method too? How would you distinguish them?
Logged
Mike Guilbault

gromit

  • Full Member
  • ***
  • Offline Offline
  • Posts: 133
Re: How many prints does it take?
« Reply #26 on: June 26, 2011, 11:10:26 pm »

I don't see how one fixed monitor calibration can simulate those very different outcomes, and dedicating multiple monitors to multiple output devices doesn't seem particularly practical, IMHO, but to each his own.

Most people haven't picked up on this yet, but modern monitors (at least the NEC PA series) can be re-calibrated/profiled on the fly. But in reality, if you're printing to a closely matched range of papers (which is what I do) it's rarely required. My reprint rate is probably about 1% ... it has to be because it's a business.

Anyway, the point of this is not to say that one way is better, just that there are two different approaches and which you choose will govern the calibration settings.
Logged

MHMG

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1285
Re: How many prints does it take?
« Reply #27 on: June 27, 2011, 09:14:24 am »

Most people haven't picked up on this yet, but modern monitors (at least the NEC PA series) can be re-calibrated/profiled on the fly. But in reality, if you're printing to a closely matched range of papers (which is what I do) it's rarely required. My reprint rate is probably about 1% ... it has to be because it's a business.

Anyway, the point of this is not to say that one way is better, just that there are two different approaches and which you choose will govern the calibration settings.

Well, at least on the Mac, one can also use the display preferences panel to switch with relative ease between various pre-built monitor calibration settings since the profiles embed vgct tags. But that's hardly any easier (and less accurate) than a workflow using a competent softproofing method such as readily available in modern version of PS. So, I"m not sure what production efficiency or accuracy you believe is being gained by recalibrating your monitor on the fly for different media. In fact, most display calibrators take a few minutes at best to complete the task and longer at worst.  As for staying on closely matched media and just getting to "know" how your monitor colors are finally rendered to your chosen media, or force-tuning the knobs to make the monitor "look like the print", both of these approaches are what everyone used before ICC profiles changed the color workflow game.  Most advanced amateurs and professional photographers that I know have moved on nowadays to full ICC-enabled workflows, although there are undoubtedly some "old timers" in the graphic arts industry that still rely on closed-loop CMYK color management methods. If the old ways work better for you, then by all means stick with them, but many of us here on LULA will take exception to your prior statement that ICC profile-enabled workflows are primarily "aimed at hobbyists who are happy to spend forever on a single image and apply (mandatory) edits to image contrast etc in a simulated environment for the paper chosen".
Logged

gromit

  • Full Member
  • ***
  • Offline Offline
  • Posts: 133
Re: How many prints does it take?
« Reply #28 on: June 27, 2011, 10:04:27 am »

If the old ways work better for you, then by all means stick with them, but many of us here on LULA will take exception to your prior statement that ICC profile-enabled workflows are primarily "aimed at hobbyists who are happy to spend forever on a single image and apply (mandatory) edits to image contrast etc in a simulated environment for the paper chosen".

I used the two extremes to emphasize the differences, not literally. There are efficiency considerations that those working on honing just a few images will be less concerned about. If you're editing an image on a monitor at higher contrast then remapping it for the print, it's a two-step process and as a consequence you'll be less productive. Again, this may not bother many. (Unfortunately the native tools in Photoshop aren't well adapted to this mapping. If you're using a composite curve to reinstate the lost contrast you also have to counter the side effects on colour ... but this is different discussion.)

I'm not sure how you figure that the monitor-based calibration approach isn't an ICC workflow. If I understand your advice on a specific contrast setting to make soft-proofing work properly it's no less prescriptive than using the actual value for the paper. I also think you make too much of the differences between papers as the printer profiles themselves (if built uniformly) will map an image value with an L* of 50 to a constant value on paper. So if you're printing to matte or glossy, the general appearance will be much the same. What does change is that the shadows will be more compressed on matte than glossy which is likely the trade-off you'd make yourself if adapting an image to a paper with less absolute contrast.

What is important here though is for people to make up their own mind which approach they're using, then calibrate accordingly. You can't mix a monitor contrast target from one and use the alternate approach. Which is the main point I'm trying to make.
Logged

Mike Guilbault

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1040
    • Mike Guilbault Photography
Re: How many prints does it take?
« Reply #29 on: June 27, 2011, 10:08:57 am »

So once I run the calibration on my monitor (I use the Eye1 Display 2 on an Eizo CE210W with the Eizo ColorNavigator software), how would I 're-calibrate' it for use with different workflows.  I need one for my lab, one for straight digital work (ie. commercial work that will be used on the web) and one for printing to my 4900.
Logged
Mike Guilbault

MHMG

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1285
Re: How many prints does it take?
« Reply #30 on: June 27, 2011, 11:06:22 am »


I also think you make too much of the differences between papers as the printer profiles themselves (if built uniformly) will map an image value with an L* of 50 to a constant value on paper. So if you're printing to matte or glossy, the general appearance will be much the same.

No. That is incorrect. Make a few prints with excellent profiles on various papers ranging from gloss to matte to plain uncoated paper and compare L=50 input to the final printed output values. This test will definitely not yield "constant" mid tone (L=50) results unless using absolute absolute rendering intent or to a lesser extent relative rendering intent without BPC, definitely not with perceptual mapping intents. However, Absolute and Relcol fail miserably for most digital images when printed to matte papers due to severe clipping of the shadow values and out-of-gamut colors among other issues. Hence the need for relcol with BPC and/or perceptual mapping, and hence the variability in mid-point mapping results.


What is important here though is for people to make up their own mind which approach they're using, then calibrate accordingly. You can't mix a monitor contrast target from one and use the alternate approach. Which is the main point I'm trying to make.

I understand, but I'm also trying to guess what percentage of pros/advanced amateurs still (if ever) advocate this monitor contrast-altering approach to a digital workflow besides yourself. At the risk of going even more OT, perhaps some others who do use this method can weigh in on the discussion. Anyway, most pros I know rely on monitors calibrated to an internationally recognized display device standard so that prior to any discussion of printing options or softproofing simulations, at least the encoded digital file data shared between customer and client will have a reasonable chance at looking alike to all concerned. All major image colorspaces in use today (sRGB, aRGB, ProphotoRGB, etc) map RGB 0,0,0 to L=0.0.  If you want to view these image files with proper image contrast on your display you can't view them on a monitor tuned to have a blackpoint value matching typical matte surface media black levels (L* minimum > 15).
Logged

gromit

  • Full Member
  • ***
  • Offline Offline
  • Posts: 133
Re: How many prints does it take?
« Reply #31 on: June 27, 2011, 06:21:11 pm »

If you want to view these image files with proper image contrast on your display you can't view them on a monitor tuned to have a blackpoint value matching typical matte surface media black levels (L* minimum > 15).

In practice, I find a monitor calibrated to maximum paper contrast is sufficient for both matte and glossy, namely a value around 250:1 (100cd/m2 / 0.4cd/m2). The shadows will compress with matte but this is expected (and desirable) behaviour. (Except for extremely low key images popular with art school students where matte probably isn't a contender in the first place.)

As for the placement of L* 50 levels on matte and glossy, I know for a fact that this is what I get for B&W as I linearize the output of both to produce exactly this result. I'll have to check again what the results are for colour.

Since you seem to be unaware of it, you may want to take a look at NEC's MultiProfiler software and the possibilities this opens up.
« Last Edit: June 27, 2011, 07:10:53 pm by gromit »
Logged

gromit

  • Full Member
  • ***
  • Offline Offline
  • Posts: 133
Re: How many prints does it take?
« Reply #32 on: June 27, 2011, 06:54:33 pm »

So once I run the calibration on my monitor (I use the Eye1 Display 2 on an Eizo CE210W with the Eizo ColorNavigator software), how would I 're-calibrate' it for use with different workflows.  I need one for my lab, one for straight digital work (ie. commercial work that will be used on the web) and one for printing to my 4900.

See my post above for a possible starting point for calibration. Eizo monitors (apart for maybe the CG245W/275W which I haven't used) don't do on the fly recalibration so you'll have to opt instead for a median whitepoint ... or use soft-proofing tools to simulate this.
Logged

Mike Guilbault

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1040
    • Mike Guilbault Photography
Re: How many prints does it take?
« Reply #33 on: June 27, 2011, 10:56:10 pm »

That's what I figured... all the more reason we need soft-proofing in LR!
Logged
Mike Guilbault

gromit

  • Full Member
  • ***
  • Offline Offline
  • Posts: 133
Re: How many prints does it take?
« Reply #34 on: June 28, 2011, 12:54:16 am »

... all the more reason we need soft-proofing in LR!

Or a different monitor like NEC's PA series with MultiProfiler which can soft-proof internally (load a paper profile, set the rendering and other options and off you go). You then flip between a monitor calibration not too far from the output and another that satisfies Mark's requirements for higher contrast for soft-proofing. Dunno why I haven't tried this before!
Logged

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
Re: How many prints does it take?
« Reply #35 on: June 28, 2011, 07:58:47 am »

...............

 Anyway, most pros I know rely on monitors calibrated to an internationally recognized display device standard so that prior to any discussion of printing options or softproofing simulations, at least the encoded digital file data shared between customer and client will have a reasonable chance at looking alike to all concerned. All major image colorspaces in use today (sRGB, aRGB, ProphotoRGB, etc) map RGB 0,0,0 to L=0.0.  If you want to view these image files with proper image contrast on your display you can't view them on a monitor tuned to have a blackpoint value matching typical matte surface media black levels (L* minimum > 15).

This is really the bottom-line on this issue. The whole purpose of colour management is to have - to the extent feasible - consistent rendering across output devices whether it is a scanner, monitor or a printer, so that as long as each one of them is colour-managed to reproduce file numbers correctly, one gets the same viewing experience. As you start customizing your colour management policies to specific devices you start removing yourself from the basic paradigm. It shouldn't be necessary to buy a bespoke display model to be able to implement a "multi-lingual" colour management solution, and indeed it isn't, if you do soft-proofing. And yes it is a pain that soft-proofing is not yet available in LR. It will be one of these days. Until then, if you insist on printing from LR, you need to use workarounds, one of which is bending your display to mimic your printer. It's not necessarily a "wrong" thing to do, but one needs to recognize its limitations and then decide whether or not those limitations are important to your other requirements. If they are not, just do it; if they are, softproofing in Photoshop and round-tripping into LR for printing is another option - not one I would use although I have both applications. I simply print from Photoshop with softproofing. I believe in keeping life easy and reliable.
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

gromit

  • Full Member
  • ***
  • Offline Offline
  • Posts: 133
Re: How many prints does it take?
« Reply #36 on: June 28, 2011, 08:15:44 am »

Until then, if you insist on printing from LR, you need to use workarounds, one of which is bending your display to mimic your printer. It's not necessarily a "wrong" thing to do, but one needs to recognize its limitations and then decide whether or not those limitations are important to your other requirements.

I tried it with my PA271W. It works very well and you can just toggle back and forth with your default calibration. I'm curious what "limitations" you came across ...
Logged

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
Re: How many prints does it take?
« Reply #37 on: June 28, 2011, 08:52:42 am »

The only limitation I had in mind was that stemming from not using the universal approach to colour management as Mark McCormick-Goodhart described it in his last post. In the final analysis, it will be ideal when LR comes with soft-proofing for all those who wish to print from that application. Until then, if you wish to preserve a file that is not purposed to a specific output device, extra steps are necessary regardless of the workflow - via the display or via Photoshop. There is no magic bullet to this.
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

gromit

  • Full Member
  • ***
  • Offline Offline
  • Posts: 133
Re: How many prints does it take?
« Reply #38 on: June 28, 2011, 09:36:40 am »

The only limitation I had in mind was that stemming from not using the universal approach to colour management as Mark McCormick-Goodhart described it in his last post.

It's actually no different to soft-proofing with Photoshop except that instead of Photoshop converting to the printer space, then back to the monitor space and sending the resultant values to the monitor, a new monitor space (and corresponding internal LUT) is constructed that combines these two steps. In other words, it's no less colour managed. I spent a few hours this afternoon comparing the results with those from Photoshop (same paper profile, rendering, settings etc) as was impressed. And, as I said, you can just toggle between this and the original calibration. Well worth a try!
Logged

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
Re: How many prints does it take?
« Reply #39 on: June 28, 2011, 10:06:08 am »

It's actually no different to soft-proofing with Photoshop except that instead of Photoshop converting to the printer space, then back to the monitor space and sending the resultant values to the monitor, a new monitor space (and corresponding internal LUT) is constructed that combines these two steps. In other words, it's no less colour managed. I spent a few hours this afternoon comparing the results with those from Photoshop (same paper profile, rendering, settings etc) as was impressed. And, as I said, you can just toggle between this and the original calibration. Well worth a try!

OK - you build the soft-proof condition in the display instead of in Photoshop - that part I see no issue with. Now, when you receive the image in the "profiled for paper X" setting of your display, just like for softproofing with PS, you may find that you need an adjustment layer or two for tweaking the image to your satisfaction. So you build them, adjust as needed, and print. Looks in principle the same as triggering a softproof condition in Photoshop and doing the same thing. As long as the output-device-specific "tweaks" are retained as pieces of removable math floating above the image (which is what an Adjustment Layer does), it seems like 6 of 1 and half a dozen of the other.
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."
Pages: 1 [2] 3   Go Up