Ad
Ad
Ad
Pages: [1]   Bottom of Page
Print
Author Topic: Color temperature - monitor versus viewing lamp  (Read 3023 times)
mkokic
Newbie
*
Offline Offline

Posts: 1


« on: January 07, 2009, 04:35:30 PM »
ReplyReply

Hi there,

I am having a photo book published and am getting some professional help with retouching the images.  The guy retouching has his monitor calibrated to 6500 Kelvin at gamma 2.2.  Now we are viewing the printed images (rgb printed on a high end epson) under a 5000 K light.  They pretty much match what we are seeing on monitor.  First question is why do they match given that the monitor is at 6500 k and the viewing light is at 5000 k?  Shouldn't the viewing light also be at 6500k?  

I am sending digital files to the offset printer who will convert the digital rgb images into cmyk.  Along with these digital images I am sending the prints we printed on the epson.   I want them to use the prints I sent them as a reference and I want them to send me back proofs of what the photos will look like when printed.  They view all proofs under a 5000 k light as well.  Their cmyk printed proofs have to match as close as possible the rgb printed proofs I sent them.  They send me one set of proofs, I review and comment, they make any necessary corrections and send me a last set of proofs for validation. After dealing with a lot of printers on many publications I find this the only way to ensure the offset printed images match what they originally looked like on monitor.  I will also be present when we go to press to have a last look in case any minor adjustments are required.  The press operator and I will be comparing what comes off of press with the cmyk proofs.  I am not sure this is the most efficient way to do this.  It's not a cheap way to get a book printed but am willing to spend a little more to get what I want.  But can anyone suggest a better method?  Thanks.
Logged
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6931


WWW
« Reply #1 on: January 07, 2009, 04:42:48 PM »
ReplyReply

Michael has a whole article on this website about dealing with this very issue. http://www.luminous-landscape.com/essays/100-books.shtml. By the way, I've seen both the original prints and the book reproductions. What he describes here really works.
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
MBehrens
Full Member
***
Offline Offline

Posts: 156


« Reply #2 on: January 07, 2009, 07:26:07 PM »
ReplyReply

An excellent article as usual from Michael. And it does mention very briefly that a D50 viewing lamp is used. However what is the rationale for a D65 display and a D50 viewing lamp?
Logged
walter.sk
Sr. Member
****
Offline Offline

Posts: 1328


« Reply #3 on: January 07, 2009, 07:56:37 PM »
ReplyReply

Quote from: MBehrens
An excellent article as usual from Michael. And it does mention very briefly that a D50 viewing lamp is used. However what is the rationale for a D65 display and a D50 viewing lamp?

As long as they are not in the same visual field, they eye adapts to each.  Despite all of the science and theory behind color management, as far as I can tell there is only one reason for the D65 display and the D50 print viewer:  It works better than other combinations in terms of being able to judge the accuracy of softproof to print.  If it didn't, I'm sure there would be other combinations used.
Logged
Hermie
Full Member
***
Offline Offline

Posts: 207


« Reply #4 on: January 08, 2009, 01:58:36 AM »
ReplyReply

Quote from: mkokic
Hi there,

I am having a photo book published and am getting some professional help with retouching the images.  The guy retouching has his monitor calibrated to 6500 Kelvin at gamma 2.2.  Now we are viewing the printed images (rgb printed on a high end epson) under a 5000 K light.  They pretty much match what we are seeing on monitor.  First question is why do they match given that the monitor is at 6500 k and the viewing light is at 5000 k?  Shouldn't the viewing light also be at 6500k?

When viewing scenes or hard-copy reproductions, it is generally assumed that one adapts almost completely to the color and luminance of the prevailing light source. This is likely not the case when soft-copy image displays are viewed. Differences in the degree of chromatic adaptation to hard-copy and soft-copy displays point to two types of chromatic-adaptation mechanisms: sensory and cognitive. Sensory mechanisms are those that act automatically in response to the stimulus, such as retinal gain control. Cognitive mechanisms are those that rely on observers' knowledge of scene content.

The cognitive mechanisms that are largely responsible for chromatic adaption are not as active when viewing a soft-copy. Consequently chromatic adaptation isn't full and a more bluish white point is required to match a D50 light booth to a display.

One should of course always use what works best, but generally a D65 display is somewhat too bluish compared to a D50 box. As Walter already pointed out, as long as both are not in the same field of view, it works pretty well.

You'll probably get a better match with a display at a color temperature between 5600-6000K (5800 average), which is BTW current UGRA recommendation for matching display to hard-copy. But again, just use whatever works best for you.
Logged
digitaldog
Sr. Member
****
Offline Offline

Posts: 9092



WWW
« Reply #5 on: January 08, 2009, 09:35:55 AM »
ReplyReply

Quote from: mkokic
Hi there,

I am having a photo book published and am getting some professional help with retouching the images.  The guy retouching has his monitor calibrated to 6500 Kelvin at gamma 2.2.  Now we are viewing the printed images (rgb printed on a high end epson) under a 5000 K light.  They pretty much match what we are seeing on monitor.

That's what matters. The numbers you start with, or use isn't the issue, that the two very dissimilar spectra appear to produce a match is all that's important! So white point and luminance of both the viewing conditions for the print and the luminance and white point of the display need to be such that you get a match. The end values don't matter (YMMV).


Quote
I am sending digital files to the offset printer who will convert the digital rgb images into cmyk.  Along with these digital images I am sending the prints we printed on the epson.   I want them to use the prints I sent them as a reference and I want them to send me back proofs of what the photos will look like when printed.

UNLESS you're actually cross rendering the RGB to CMYK then back to RGB for the Epson, forget about it. They will not match! The gamut of the Epson is huge compared to CMYK press gamut.

Quote
The press operator and I will be comparing what comes off of press with the cmyk proofs.

That's what you'll be shooting for and contractually what the press will match to.
Logged

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

Posts: 87


« Reply #6 on: January 08, 2009, 11:42:56 AM »
ReplyReply

Quote from: mkokic
I am sending digital files to the offset printer who will convert the digital rgb images into cmyk.

If not for this project then certainly for the next you'd be best served by doing the RGB-CMYK conversions yourself. There's much more of an art to it than going Image>Mode>CMYK. You can decide on the best rendering intent for the conversion of each image and just as importantly, then edit the CMYK image to determine how the out of gamut colors will appear. There is no right or wrong, hard fast rule or formula for wrangling out of gamut information (it could be hue, saturation , brightness, or some combination of all three that's beyond the press capabilities), though there are preferred practices to diagnosing and addressing the issue. In short, the RGB>CMYK conversion involves very subjective decisions and passing them off to someone else will invariably lead to different results than you may have desired. It's not just "I can't have that saturated green or dark blue on press", it's what you choose to replace it with that can really make the difference on the printed page. Having recently finished a book where I had to convert almost 300 images individually, I can say that the effort was well worth it in the end. Sure, a fair number of the images looked OK with a straight conversion, but many others would have been disastrous without some individual adjustments.
Plus, by sending RGB files you're giving the printer lots of colors he has no chance of being able to print. Typically, this will not make him happy. And nothing will come close to looking like your RGB inkjet proofs. Typically, this will not make you happy. At the very least you may want to consider using the PS proof option when making your proofs so that your inkjet simulates the press behavior.
Logged

Tim Lookingbill
Sr. Member
****
Offline Offline

Posts: 1153



WWW
« Reply #7 on: January 08, 2009, 03:29:36 PM »
ReplyReply

About your concern about lights your prints will be viewed under, below is a shot of a CMYK and inkjet print of two standard color targets that you can easily find on the web. The CMYK (Ole No Moire) can be found on a Photoshop CD install and must have assigned Photoshop's sheetfed coated CMYK profile for accurate viewing since that print came off a Komori sheetfed press from CopyCraft out of Lubbock, Texas.

It is lit by a 4700K 50watt Solux clip-on desk lamp and photographed and processed in ACR 3.7. It is not converted to sRGB but assigned my i1Display iMac profile because converting to sRGB crushes the yellow and cyan in the CMYK target and kind of makes the grayscale on the right look a bit more red so you need to view this file in a color managed app especially if viewing on wide gamut monitor like NEC's and Eizo's.

As you'll see by comparison to the R=G=B grayramp on the far right there's not much difference between 6500K and 4700K so 5000K shouldn't be any problem. Now the 5000K the printer is using could be a fluorescent variety and would probably look more yellowish but from the green side of the spectra. There is yellow in green. I don't see this as a problem. Some fluorescents that claim 5000K do make grayscale look a tinge redder than the 6500K display while making skintones appear yellower. But the differences by eye are going to be as subtle as the differences seen in the posted image.

High quality looking images (PHOTOS) are going to affect the eye and one's judgement as to what looks GOOD over what looks accurate. In this case it's better to LOOK good than to feel accurate. This does not apply to matching a corporate logo color on press which is whole other level of headache you don't want to get into and shouldn't worry about.

[attachment=10810:SoluxInk..._LLforum.jpg].
« Last Edit: January 08, 2009, 03:43:37 PM by tlooknbill » Logged
Pages: [1]   Top of Page
Print
Jump to:  

Ad
Ad
Ad