Ad
Ad
Ad
Pages: [1]   Bottom of Page
Print
Author Topic: How to optimize for non optimal viewing light  (Read 5483 times)
ErikKaffehr
Sr. Member
****
Online Online

Posts: 7426


WWW
« on: July 29, 2008, 12:22:31 AM »
ReplyReply

Hi,

I'm profiling my screen (iMac) and Printer (Epson right now) with an Xrite ColorMunki and I'm quite happy. That is my colors match when I have very good viewing light. My question is how do you optimize prints for non optimal viewing light?

As far as I understand viewing light should be about 400 Lux (or more) and have D50 characteristics. This is much more bright than what you would have in a home.

The issue I seem to have is that I cannot see good separation in the dark areas. If I have enough light the print "glows up", almost like a transparency. With more subdued light it's just murky, dark and boring.

I would appreciate any views, suggestions or ideas.

Best regards
Erik
« Last Edit: July 29, 2008, 02:23:25 AM by ErikKaffehr » Logged

tho_mas
Sr. Member
****
Offline Offline

Posts: 1696


« Reply #1 on: July 29, 2008, 02:36:21 AM »
ReplyReply

Quote
My question is how do you optimize prints for non optimal viewing light?
"Optimizing" for non optimal viewing conditions will be tricky ;-)

Quote
As far as I understand viewing light should be about 400 Lux (or more) and have D50 characteristics. This is much more bright than what you would have in a home.
For viewing the prints in a light box!
Luminance in front of the display should be around 32lux.

If you don't have a light box to eye up prints you sould set everything in a way that is comfortable.
"Gamma" as you like, whitepoint 5500K-5800K, luminance 120cd/qm - 160cd/qm.
No direct lights on the display. That should be fine.
The problem with the iMac is the glossy panel. So contrast will always look to high and the darks to bold. And especially with a glossy panel you should dim down the lights in the room as much as possible.

http://www.creativepro.com/article/the-dar...akes-a-comeback
Logged
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6943


WWW
« Reply #2 on: July 29, 2008, 07:15:37 AM »
ReplyReply

Quote
Hi,

I'm profiling my screen (iMac) and Printer (Epson right now) with an Xrite ColorMunki and I'm quite happy. That is my colors match when I have very good viewing light. My question is how do you optimize prints for non optimal viewing light?

As far as I understand viewing light should be about 400 Lux (or more) and have D50 characteristics. This is much more bright than what you would have in a home.

The issue I seem to have is that I cannot see good separation in the dark areas. If I have enough light the print "glows up", almost like a transparency. With more subdued light it's just murky, dark and boring.

I would appreciate any views, suggestions or ideas.

Best regards
Erik


[a href=\"index.php?act=findpost&pid=211320\"][{POST_SNAPBACK}][/a]

Erik,

The main line of attack on this problem is to get the display to simulate your print viewing conditions as closely as possible, so that it "forces" you to adjust the image accordingly. Hence, if you have rather dim print viewing conditions, you should begin the process with a rather dim display. This will encourage you to pump-up the brightness and contrast enough to achieve the end result you seek. I find that with my display calibrated at 110cd, it's a quite reliable indicator of prints viewedabout five feet under direct Solux D50 illumination, which is quite bright. Hence in your situation I would try calibrating the display at about 90~95 cd and see whether that helps. YOu may also experiment with the display at set at white points within the range of 5000~6500 and see whether that helps, but as you get closer to 5000 the tones start to look excessively yellowish, which can be a put-off.

The next thing is the printing paper you are using. If you are printing on matte papers, the DMax is not as good as it would be printing on a gloss or luster paper, so tonal separation, particularly below the three-quarter mark gets squished, producing that muddy feel to the images. Hence you can improve your dark tone performance by using non-matte papers.

Finally, there is soft-proofing. You should make your final luminosity adjustments under soft-proof so that the display gives you a better impression of what the print will look like. Make sure you have the Simulate Paper White box checked, and implement Black Point Compensation.

Also, I assume you know to make sure that if you are letting Photoshop Manage Color using your paper/printer profile, you have Printer Color Management switched off, and you check that it really is OFF before you print.

I don't know the details of ColorMunki's display profiling process, but I presume/hope it lets you select the luminance of the display and the white point before it reads the patches. Another aspect of ColorMunki over which you have no control is that for generating a printer profile it reads only 100 patches. There is a question in some peoples' minds about whether this is sufficient to produce a sufficiently refined paper/printer profile. I don't know the answer to that first-hand because it is not possible to buy this product on a trial period basis, but there is much discussion about it. To help sort out that one, I would suggest comparing prints made under otherwise identical settings using your ColorMunki paper/printer profile and the canned profile supplied by the paper/printer manufacturer, to see which profile gives you better results in general and better deep tone separation in particular.
Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
digitaldog
Sr. Member
****
Offline Offline

Posts: 9110



WWW
« Reply #3 on: July 29, 2008, 07:44:35 AM »
ReplyReply

Quote
Hi,

I'm profiling my screen (iMac) and Printer (Epson right now) with an Xrite ColorMunki and I'm quite happy. That is my colors match when I have very good viewing light. My question is how do you optimize prints for non optimal viewing light?

You can't (not if its not optimal and the profile(s) are based on that usage).

You can build printer profiles with some packages like ProfileMaker Pro such that the exact lighting under which the prints are viewed (not automatically D50) is used. You can alter the luminance and white point of the display target calibration for that viewing condition. But if you mix up or alter these, then the quality of the print to screen matching is going to be off. But its likely you'll be viewing prints nowhere near that display.
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
tho_mas
Sr. Member
****
Offline Offline

Posts: 1696


« Reply #4 on: July 29, 2008, 08:31:41 AM »
ReplyReply

Quote
Make sure you have the Simulate Paper White box checked
This is a point stated quite often but does not work. Measurement devices measure the paper white blueish (and greyish) if the paper contains optical brighteners. And nearly all the photographic papers do. And the measured buleish white is stored as whitepoint in the profile (according to the maths of color management).
But if you adjust the display to a paper equivalent whitepoint (and contrast) there is no need to simulate paper white as the display already is set to the right whitepoint.
BPC helps to estimate the paper contrast and should be activated.
« Last Edit: July 29, 2008, 08:37:23 AM by tho_mas » Logged
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6943


WWW
« Reply #5 on: July 29, 2008, 08:46:04 AM »
ReplyReply

Quote
This is a point stated quite often but does not work. Measurement devices measure the paper white blueish (and greyish) if the paper contains optical brighteners. And nearly all the photographic papers do. And the measured buleish white is stored as whitepoint in the profile (according to the maths of color management).
But if you adjust the display to a paper equivalent whitepoint (and contrast) there is no need to simulate paper white as the display already is set to the right whitepoint.
BPC helps to estimate the paper contrast and should be activated.
[a href=\"index.php?act=findpost&pid=211389\"][{POST_SNAPBACK}][/a]

Well, actually, based on my experience and testing, it DOES work. What you are proposing may also work, if you know what whitepoint to select for the display.
Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
tho_mas
Sr. Member
****
Offline Offline

Posts: 1696


« Reply #6 on: July 29, 2008, 09:13:05 AM »
ReplyReply

Quote
What you are proposing may also work, if you know what whitepoint to select for the display.
I don't know it. I see it. I always edit the whitepoint manually so that it's visual equivalent to the paper white in the D50 light box and to ambient lights. Theoretically that would indicate a whitepoint of 5000K. But it's definitely to warm. In my case whitepoint is matching visually when the display is set around 5500K-5600K.
Logged
ThePhotoDude
Full Member
***
Offline Offline

Posts: 160



WWW
« Reply #7 on: July 29, 2008, 09:31:09 AM »
ReplyReply

All the techy stuff above is all very well, but also consider a phsycological (sp?) angle.

I also think you shouldn't concern yourself TOO much with non-optimal viewing light. Taking a step back from all the technical aspects above - have a look at your customers own perspective. They are using non-optimal and mixed lighting (at home say), so who says they are THAT concerned with colour rendition.

How are mass produced posters and prints done?

They will be hung in a massive array of different environments, offices with flouro lighting to little old ladies homes with 40W incandescent bulbs.

The prints WONT look the same in all this different lighting, (if you could compare them side by side) but viewing them one at a time they won't look that different. Why? Because our brain has an enormous capability to adapt to different white points and compensate.

As long as you get your colors correct for normal daylight, they'll look 'alright' for most consumers.
And if you don't know their lighting, then it's the best you can do.
« Last Edit: July 29, 2008, 09:33:32 AM by ThePhotoDude » Logged
digitaldog
Sr. Member
****
Offline Offline

Posts: 9110



WWW
« Reply #8 on: July 29, 2008, 10:29:46 AM »
ReplyReply

Quote
How are mass produced posters and prints done?
The prints WONT look the same in all this different lighting,

Agreed. The area you want total control is by the display so you can evaluate screen to referenced print viewing conditions. If the ultimate goal is consistent, proper and with the right tools, one can even say accurate viewing, you have to dial in a reference environment that defines the display, the viewing booth or conditions in this surround, ambient light levels and spectral properties etc. The better and more consistent this defined reference viewing environment, the better of course and you can duplicate these conditions and provide matching evaluations of color over larger distances.

You can also somewhat optimize the output if you know where it will be, such as measuring the illuminant under which the prints will be viewed and building the output profiles with that in mind.
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
ErikKaffehr
Sr. Member
****
Online Online

Posts: 7426


WWW
« Reply #9 on: July 29, 2008, 11:29:10 AM »
ReplyReply

Hi,

Thanks a lot all contributors. My question was probably not as clear as I intended it to be. The major issue for me is not the color of the light, but the intensity. What I can see is that my pictures are much better if well illuminated then under normal lighting.

Thanks for a lot of useful input anyway!

Best regards
Erik


Quote
Agreed. The area you want total control is by the display so you can evaluate screen to referenced print viewing conditions. If the ultimate goal is consistent, proper and with the right tools, one can even say accurate viewing, you have to dial in a reference environment that defines the display, the viewing booth or conditions in this surround, ambient light levels and spectral properties etc. The better and more consistent this defined reference viewing environment, the better of course and you can duplicate these conditions and provide matching evaluations of color over larger distances.

You can also somewhat optimize the output if you know where it will be, such as measuring the illuminant under which the prints will be viewed and building the output profiles with that in mind.
[a href=\"index.php?act=findpost&pid=211436\"][{POST_SNAPBACK}][/a]
Logged

ErikKaffehr
Sr. Member
****
Online Online

Posts: 7426


WWW
« Reply #10 on: July 29, 2008, 11:38:14 PM »
ReplyReply

Thanks a lot,

I'm already doing most of that. The reason I asked is to some extent that I try to help a friend with color management and also that I guess that I'm not alone with these issues. I generally have no problems when the pictures are well illuminated (like a halogen desk lamp). My color profiles seem to be close to Epson profiles, I cannot really see a difference between papers and profiles.

One test I do for reference is to reproduce a synthetic Color Checker and compare with a real color checker. In sunlight they are quite close and the gray scale is very close to original, except the white square which is whiter on the CC than my paper.

Both me and my friends use an iMac, I set my screen brightness to minimum. Instead of very low.

Got quite a lot of feedback from the forum, I need to go reconsider my viewing environment.

Best regards
Erik


Quote
Erik,

The main line of attack on this problem is to get the display to simulate your print viewing conditions as closely as possible, so that it "forces" you to adjust the image accordingly. Hence, if you have rather dim print viewing conditions, you should begin the process with a rather dim display. This will encourage you to pump-up the brightness and contrast enough to achieve the end result you seek. I find that with my display calibrated at 110cd, it's a quite reliable indicator of prints viewedabout five feet under direct Solux D50 illumination, which is quite bright. Hence in your situation I would try calibrating the display at about 90~95 cd and see whether that helps. YOu may also experiment with the display at set at white points within the range of 5000~6500 and see whether that helps, but as you get closer to 5000 the tones start to look excessively yellowish, which can be a put-off.

The next thing is the printing paper you are using. If you are printing on matte papers, the DMax is not as good as it would be printing on a gloss or luster paper, so tonal separation, particularly below the three-quarter mark gets squished, producing that muddy feel to the images. Hence you can improve your dark tone performance by using non-matte papers.

Finally, there is soft-proofing. You should make your final luminosity adjustments under soft-proof so that the display gives you a better impression of what the print will look like. Make sure you have the Simulate Paper White box checked, and implement Black Point Compensation.

Also, I assume you know to make sure that if you are letting Photoshop Manage Color using your paper/printer profile, you have Printer Color Management switched off, and you check that it really is OFF before you print.

I don't know the details of ColorMunki's display profiling process, but I presume/hope it lets you select the luminance of the display and the white point before it reads the patches. Another aspect of ColorMunki over which you have no control is that for generating a printer profile it reads only 100 patches. There is a question in some peoples' minds about whether this is sufficient to produce a sufficiently refined paper/printer profile. I don't know the answer to that first-hand because it is not possible to buy this product on a trial period basis, but there is much discussion about it. To help sort out that one, I would suggest comparing prints made under otherwise identical settings using your ColorMunki paper/printer profile and the canned profile supplied by the paper/printer manufacturer, to see which profile gives you better results in general and better deep tone separation in particular.
[a href=\"index.php?act=findpost&pid=211372\"][{POST_SNAPBACK}][/a]
Logged

Pages: [1]   Top of Page
Print
Jump to:  

Ad
Ad
Ad