My print is too dark - and I *have* calibrated and profiled

[Slobodan Blagojevic]

... If I lower the luminosity of the monitor, the shadow details dissapear there. While that sounds like an improvement, it isn't really, because it hides detail that is in the image....

In my humble opinion, the above statement is at the core of the problem. If you lower the luminosity of the monitor and the shadow details disappear, bring them back in post-processing.

[Rand47]

That's what I would like to achieve, yes. Just to make sure I haven't overlooked something simple (which, given how new I am at this is entirely possible), these are the steps I went through:

1) calibrate monitor using NEC's software
2) profile printer using Colormunki's software
3) take a picture into Photoshop
4) setup the softproof to match the printer profile created in step 2
5) swiching softproof on. Noticing that there is shadow detail
6) printing picture
7) noticing that there is (almost) no shadow detail on the print, regardless of whether I watch it under my (not daylight) halogen lights or outside in overcast shade (the only light available at this time of the year in this country. Rain optional)

I thought the starting point should be to make sure that the screen can show a full greyscale. In other words, if some of the dark steps blend together, the monitor is too dark. If some of the bright steps blend together, then the monitor is too bright. Is that incorrect?

OK, I may be showing a lack of knowledge here, but in step 4 above you mean "the paper specific profile" created, right?  Not a printer profile?

[fdisilvestro]

It is actually a paper/printer combination profile

[Rhossydd]

If I lower the luminosity of the monitor, the shadow details dissapear there. While that sounds like an improvement, it isn't really, because it hides detail that is in the image. What I would like to achieve is for the monitor to show me everything that is in the image, and only throw away detail when it comes to softproofing (if the problem is caused by the printer's inability to print as fine luminosity details). And even "real" daylight won't show me the shadow details.

This paragraph sums up a core problem.
Yes, you can match your monitor to your print, BUT you want to see things in the image you can't print, you expect soft proofing to deliver that.
I can understand that point of view, but right now the only way to do it is to lower monitor luminosity when soft proofing. Not impossible, but not convenient either.

Slobodan's post is probably most helpful " If you lower the luminosity of the monitor and the shadow details disappear, bring them back in post-processing." Get used to working with monitor images that closer to printable.

Maybe the problem here is that current soft proofing expects a higher level of luminosity matching between print and monitor than most people can achieve. Maybe it's better to think of soft proofing as just a colour matching tool, rather a perfect overall print match.

[Schewe]

Maybe the problem here is that current soft proofing expects a higher level of luminosity matching between print and monitor than most people can achieve. Maybe it's better to think of soft proofing as just a colour matching tool, rather a perfect overall print match.

Bullshyte...if you can't get a screen>print match using soft proofing, something is wrong in your chain or you don't have a clue how to soft proof. Look, I've been doing this successfully for years and with LR4 it's even easier. It ain't rocket science...it's a matter of having the right circumstances and set up. If you learn how to do it, it's really useful. If you cover your eyes and run off screaming, not so much.

If your print is too dark, you don't understand what you are doing and it would be useful to start at scratch and work from the ground up. Some of us have been doing this stuff for years and yes, it really can work. But if you want to stick you head in the sand and claim it doesn't work...then I suggest you walk off the deep end of the pier and ask for help. Likely you'll drown but maybe not, ya know? Maybe you can be saved (and learn how to do this stuff correctly). Or not...in which case, go ahead and drown...

[Simon Garrett]

In my humble opinion, the above statement is at the core of the problem. If you lower the luminosity of the monitor and the shadow details disappear, bring them back in post-processing.

In my (I hope equally) humble opinion, this is not the right way to go.  Get the monitor right, don't compensate for the error in post processing. 

[SunnyUK]

If your print is too dark, you don't understand what you are doing and it would be useful to start at scratch and work from the ground up. Some of us have been doing this stuff for years and yes, it really can work. But if you want to stick you head in the sand and claim it doesn't work...then I suggest you walk off the deep end of the pier and ask for help. Likely you'll drown but maybe not, ya know? Maybe you can be saved (and learn how to do this stuff correctly). Or not...in which case, go ahead and drown...

With all due respect... I have stated from the beginning that I am very new at this. I have also stated that I started by reading about the issues and watching the training video that YOU and Michael put together. When I realised that I was not getting the same results as you and other people (unsurprising, for I am a newbie), I came here asking for help.  So it grates to be told that I don't know what I'm doing and that  I'm sticking my head in the sand.

I would of course be grateful for any suggestions.

In my humble opinion, the above statement is at the core of the problem. If you lower the luminosity of the monitor and the shadow details disappear, bring them back in post-processing.

Even though several people have suggested this, I don't understand it. I would like to understand. If there is detail (that this specific printer cannot print on this specific paper), would it not be wrong to match the monitor to the printer (rather than matching the softproof to the printer)?  If I do as suggested, and if I print on different papers and different printers, I will end up adjusting monitor luminosity for each print. If that is best practise, then I'll have to do it. It just sounds as if this omits the whole purpose of the softproof. Can you help me understand where my thinking is wrong, please?

But aside from that I'm surprised no one caught the difference between the blue to orange transition in the sky in his image on the display compared to the print which may point to an inaccurate printer profile.

SunnyUK, can you confirm the differences? Or is it that your photo you posted isn't that accurate. There's a big white gap on the display between the blue to orange gradation in the sky that isn't on the print.

SunnyUK, can you post a downsized for the web version of the original image of the Abbey so we can see if your display is showing shadow detail and that blue/orange gradation faithfully so we can rule out display calibration/profiling?

Good point, Tim. I'm at work at the moment, but will post an updated picture tonight and also confirm any differences in the graduations. Thank you.

That shouldn't happen.  If your monitor is correctly calibrated/profiled at a lower brightness setting, it should still be able to show the full tonal range (e.g at http://www.lagom.nl/lcd-test/black.php and http://www.lagom.nl/lcd-test/white.php).  I don't have any problem with the full tonal range at a brightness below 100cd/m2. 

Obviously the total dynamic range between black and white is narrower, as black stayed the same but you've reduced white.  But that should match the print better. 

You show a picture of your room lighting: what happens if you hold a piece of printer paper by the screen.  Does the paper look darker than the whites on the screen?  If so, try bringing up the room lights or lowering the screen brightness (and recalibrate). 

Nice picture of Whitby Abbey, by the way!

Thank you for those two test images. Very useful! I will check them on the monitor tonight and reply.

Thank you also for the compliment. I like the Abbey.

In my (I hope equally) humble opinion, this is not the right way to go.  Get the monitor right, don't compensate for the error in post processing. 

That's also what my logic tells me. But I'm keen to learn from the experienced people and get a "true" setup that gives all the benefits of the monitor, of the printer and of the softproofing software in Photoshop.

[bjanes]

The problem with the prints is most easily seen in the shadow detail (or rather, the lack of shadow detail on the print). I have also watched the print outdoor in "real" daylight, and the shadows are also way darker there than on-screen.

Thank you for the very interesting thread about daylight bulbs. I will have to look into that in more detail when/if I get a separate room for my photo work.

If I lower the luminosity of the monitor, the shadow details dissapear there. While that sounds like an improvement, it isn't really, because it hides detail that is in the image. What I would like to achieve is for the monitor to show me everything that is in the image, and only throw away detail when it comes to softproofing (if the problem is caused by the printer's inability to print as fine luminosity details). And even "real" daylight won't show me the shadow details.

You have calibrated your monitor, but that is only half the solution. What about the printer profile? Printers have more difficulty in reproducing shadow detail than the monitor, since the monitor has a greater dynamic range and a lower DMAX. That said, an accurate printer profile should enable softproofing the appearance of the shadows. Also, what rendering intent are you using, and do you have black point compensation (BPC) in the case of relative colorimetric?

Regards,

Bill

[SunnyUK]

You have calibrated your monitor, but that is only half the solution. What about the printer profile? Printers have more difficulty in reproducing shadow detail than the monitor, since the monitor has a greater dynamic range and a lower DMAX. That said, an accurate printer profile should enable softproofing the appearance of the shadows. Also, what rendering intent are you using, and do you have black point compensation (BPC) in the case of relative colorimetric?

Hi Bill,

I created a printer/paper profile using Colormunki Photo and use that for the softproofing. When printing, I use rendering intent Perceptual with Black Point Compensation ticked. I seem to remember reading somewhere that perceptual is good in 80% of the cases, and since I am not skilled enough to know what the exceptions are, I have been sticking to that.

[jeremypayne]

I created a printer/paper profile

If you are truly a beginner, profiling your printer is significant overkill in my opinion.

Start with an off-the-shelf paper and use the profile offered by the manufacturer.

Print an unedited test file.  Review in your normal fashion.   Let us know how you find that print.

I think you'll find that unless you are using exotic papers (without supplied profiles for your printer) that you don't really need to profile your printer.  I've never found the need to profile a printer and get consistent and predictable output from several different HP and Epson printers.

[fdisilvestro]

If there is detail (that this specific printer cannot print on this specific paper), would it not be wrong to match the monitor to the printer (rather than matching the softproof to the printer)?  If I do as suggested, and if I print on different papers and different printers, I will end up adjusting monitor luminosity for each print. If that is best practise, then I'll have to do it. It just sounds as if this omits the whole purpose of the softproof. Can you help me understand where my thinking is wrong, please?

Yes, it has to be in "Softproof" mode when matching the monitor to the print, but this might require lowering the brightness anyway. If you do it correctly, it should work well for other printer/paper combinations without adjusting the monitor, provided you use the same viewing conditions.

[Slobodan Blagojevic]

In my (I hope equally) humble opinion, this is not the right way to go.  Get the monitor right, don't compensate for the error in post processing.  

Ha! The battle of two equally humble opinionators!

In all seriousness, I thought that lowering luminance is (a part of) "getting the monitor right," no? There should be no "error" to compensate for if the monitor is properly calibrated, which would include a proper luminance. This, of course, should be done only once (during calibration/profiling), not for every paper/print combination.

Now, about that "blue to orange transition in the sky." It seems that is yet another sign that the monitor is too bright. Bring the luminance down, and you might start seeing some (or all) of that orange back. It might be simply overblown by exposure (though not in camera, but by "overexposure" of the monitor).

If you ever shot film, especially transparencies, you will remember that how much shadow detail you can see depended on the light source behind the slide film. If weak light, your shadows would certainly look blocked. Keep increasing the light and you will start to see details in the shadows. Do it too much and you will blow highlights, just like in-camera exposure.

Monitor is just like a light source for transparencies, i.e., it emits light, vs. prints that reflect it.

It might be just an impression, but just by looking at your supplied photo of the setup, one gets the impression that the monitor is too bright and too blue (relative to the print). In other words, either the monitor's white point is off, or your room lighting is too warm.

[digitaldog]

This, of course, should be done only once (during calibration/profiling), not for every paper/print combination.

Not for us SpectraView users (or those with similar capabilities). We can build a suite of calibrations including contrast ratio's for each paper we hope to soft proof for, load them on the fly within the software and it swaps in the correct ICC profile for that calibration. It's the beauty of these smart display systems.

[SunnyUK]

Not for us SpectraView users (or those with similar capabilities). We can build a suite of calibrations including contrast ratio's for each paper we hope to soft proof for, load them on the fly within the software and it swaps in the correct ICC profile for that calibration. It's the beauty of these smart display systems.

Now think I understand why I didn't understand your previous suggestions. You are talking about using the SpectraView display modes in order softproof, instead of using Photoshop. Is that correct?

[digitaldog]

Now think I understand why I didn't understand your previous suggestions. You are talking about using the SpectraView display modes in order softproof, instead of using Photoshop. Is that correct?

No, I'm not talking about using the SpectraView soft proof functionality. We're still working within Photoshop and Lightroom using their soft proofing.

I'm saying if you need to calibrate to match a print, you have the option of multiple targets and calibrations thanks to that system. Without, you need to stick to one OR recalibrate for each use. 

[walter.sk]

20121120-_TP50991 by sunnyUK, on Flickr

The problem with the prints is most easily seen in the shadow detail (or rather, the lack of shadow detail on the print). I have also watched the print outdoor in "real" daylight, and the shadows are also way darker there than on-screen.

If I lower the luminosity of the monitor, the shadow details dissapear there. While that sounds like an improvement, it isn't really, because it hides detail that is in the image. What I would like to achieve is for the monitor to show me everything that is in the image, and only throw away detail when it comes to softproofing...

Some more thoughts:  While it is difficult to tell from your picture of the monitor, it seems to me that the shadows are blocked in your image, and that you are under the impression that the monitor is supposed to show you the "correct exposure" of your image.  If my image looked as blocked in the shadows on my monitor as yours does I would try to open up the shadows in Photoshop or LR, or whatever you are using.

You say you would like to have the "monitor to show me everything that is in the image..."
Well, with your excellent monitor and and calibration tool, I would assume it *does*.  I suspect if you re-calibrated your monitor anywhere from 115 or 120 rather than the 140 for the luminance, reopen the RAW file and get it to look right, and then print it you would see whether the darkness of the print is moving toward the correct appearance.

A second thought:  Buying a task lamp, especially one like a Solux with controlled temperature that you can position relative to your print viewing area such that the luminosity of the paper resembles that of the monitor, would give you a consistent viewing light that would allow for judging the brightness of the print.

The third point:  It does not appear that, when you look at the softproofed version of your image you are making any further adjustments of it compared to the original image in order to recapture as much as possible the look of the original.  This would entail adding contrast, shifting color balance, brightening, boosting blacks, etc, that compensate for the shifts from the original image to the softproofed version.  

[Schewe]

I seem to remember reading somewhere that perceptual is good in 80% of the cases, and since I am not skilled enough to know what the exceptions are, I have been sticking to that.

Actually, you have that wrong...RelCol is often the best rendering intent because the keeps colors that are in gamut the same. Percep rending adjusts all colors so the out of gamut colors are in gamut–which can have a negative impact on in-gamut colors.

The ONLY way to KNOW which rendering intent will be best for a given images is to look at both and see which one does the best job. Blindly selecting Percep is a poor choice.

[bjanes]

I created a printer/paper profile using Colormunki Photo and use that for the softproofing. When printing, I use rendering intent Perceptual with Black Point Compensation ticked. I seem to remember reading somewhere that perceptual is good in 80% of the cases, and since I am not skilled enough to know what the exceptions are, I have been sticking to that.

BPC is not usually used with perceptual rendering, since that rendering attempts to match the source black to the output black as explained in this Adobe paper on BPC (see section 6.2). Nonetheless, Adobe states that BPC is still available for use with perceptual rendering in the case of malformed profiles (not suggesting that your profiles are malformed .

Two good but rather old articles on rendering intents are here and here. The important thing to remember is that the color management system (CMS) does not look at what colors are actually in your image. Perceptual rendering compresses the color gamut whether or not compression is needed for out of gamut colors. If your image contains no out of gamut color, this compression may be undesirable. The situation may change with "smart CMSs" with ver 4 profiles, but ver 4 is reportedly not ready for prime time and I have not seen a good explanation on how or when to use them. Perhaps Jeff or the Digitaldog can chime in and help us out here.

In past times when we were printing to devices with relatively narrow color gamuts, many authors recommended perceptual rendering for general use. However with current wide gamut ink jet printers, it is less likely that an image will contain colors that are out of gamut for the printer and I think that relative colorimetric is preferred most of the time as Jeff Schewe indicated. Soft proofing using both intents can be helpful to see which works best. However, soft proofing can be limited by the gamut of the monitor, which obviously can not proof colors that are out of its gamut and within the gamut of the printer. Nonetheless, it works most of the time.

Regards,

Bill

[SunnyUK]

Firstly, thank you very much for all the helpful input and explanations.

Secondly, I need to eat a large portion of humble pie. I finally re-calibrated my monitor to 120 cd/m2, edited the picture so the deep shadows got unblocked, and printed. To my untrained eye, the print is now very close to the Photoshop soft-proof. I consider the setup "good enough" for me at this stage.

Of course this was what you kind people were saying right from the beginning.

Thank you also for the links to further reading. I've read it all, and will keep it handy for when I start to get in doubt about profiling again.

I'm going to buy a good daylight balanced lamp, maybe one of the fancy Solux ones.

I haven't quite gotten my head around the use of multiple monitor profiles that can/should be changed on the go. I know my Spectraview monitor can do such stuff. I'll need to read up on that some more to get my head around it.

So thank you again for all your help. You've all been very kind.

[Kanvas Keepsakes]

 this is my first post to this forum - hope this helps :



  1.  using Photoshop create a duplicate layer of your finished photo

  2.  make sure this is the top layer

  3.  set the blend mode of this duplicate layer to "Screen"

  4.  set the opacity of this layer to ~ 25-30%

  5.  print and be amazed


(this tip is credited to Matt Kloskowski from NAPP)

Brilliant!  Such a simple trick yet works so well.  I was just posting about this same issue with my prints coming out a little darker.  I just got a Colormunki Photo yesterday and calibrated.  Things look great but with these prints I have here for this photographer that likes to take real dark images, this trick works great.  Thanks!