Color management : theory vs reality

[Guigui]

Hello,

I'd like to hear your thoughts on something that's been bothering me for a while. I'm still a beginner at digital printing so hopefully you'll correct me if I'm wrong.

From time to time, I print photos or artworks for other people (people who have never heard about color management).

Let's assume that I have a perfectly color managed workflow and that I get good matches between prints & display. Let's also assumre that the customer doesn't. Instead, what the customer sees is a dark looking image compared to their beloved uncalibrated display which has a native 400cd/m² luminance, high contrast, high saturation, high everything. Furthermore, they don't light prints well which results in prints looking even darker.

How do you explain to your customers that although their print looks like crap on their walls, it looks just right in a viewing booth ? Why should they care about that ?

Am I the only one to be tempted to significantly brighten every image before printing in order to satisfy the average customer ?

Thanks, and a happy new year to everyone around here.

[digitaldog]

How do you explain to your customers that although their print looks like crap on their walls, it looks just right in a viewing booth ? Why should they care about that ?

Go over to customers display with ugly preview and adjust the controls so it looks different (better, worse, doesn’t matter). Show them the RGB numbers didn’t change. Only when the RGB numbers preview correctly and make a matching print, do you have consistency and a correct way to look at what is nothing more than a huge pile of numbers. Numbers don’t show us what colors look like on their own. Color management ensures the numbers produce the results the numbers represent (in a defined color space). Color management is number management.

As for viewing prints, take them into room at night, show them print. Then turn out the lights. Now if you have some odd illuminant (like a black light or some ugly illuminant like metal halide), show them the print. Does it look the same? The illuminant plays a role (there’s no color without light). The viewing booth is a standard, controllable illuminant. It matches the display. That’s all we really need at this point.

[Ken Bennett]

How do you explain to your customers that although their print looks like crap on their walls, it looks just right in a viewing booth ? Why should they care about that ?

Speaking as your customer, I don't care at all how good the print looks in your viewing booth. I'm not going to display the print there -- I'm going to display it on the wall in my living room.

I often have to adjust my prints for the display conditions. If I were printing for others, I would definitely ask about those conditions and make a print that best matches them.

[digitaldog]

If it looks good in the booth,  unless you take it home and illuminate it with some really weird-ass illuminant, it will look fine.

Or you could pay the printer to measure that illuminant with say an EyeOne, build a custom profile for that one condition and it would look even better.

We've been making prints (and other reproductions) from transparencies, using a totally analog process for years (while viewing the transparency with defined or weird-ass illuminants).  This is nothing new expect we have to deal with big piles of numbers and numbers alone don’t define color let alone color appearance.

[Rhossydd]

Or you could pay the printer to measure that illuminant with say an EyeOne, build a custom profile for that one condition and it would look even better.

Interesting comment.
So, if you think it’s worth building profiles for specific lighting conditions, what’s wrong with editing profiles to brighten prints to suit most ordinary users not using high quality calibrated monitors who see their prints under lower than ideal lighting conditions, as Guigui suggests ?

Every photo forum I come across has one or more postings moaning about ‘dark prints’ these days.
Yes. I know the profiles being used are almost always “correct” by ICC standards and their monitors are too bright, but if the majority of ‘normal’ users are seeing problems with the “correct standards” isn’t there a hint that something is outdated about the assumptions that were built into the “standards” used ?
Now monitor technology has moved on from old dim CRTs to brighter flatscreens, maybe the consequences of that development need to be reconsidered ??

Paul

[digitaldog]

Interesting comment.
So, if you think it’s worth building profiles for specific lighting conditions, what’s wrong with editing profiles to brighten prints to suit most ordinary users not using high quality calibrated monitors who see their prints under lower than ideal lighting conditions, as Guigui suggests ?

I only recommend profile editing for masochists and those that can do this with OPM (Other People’s Media) while being paid by the hour.  

Profiles reflect printing and viewing conditions (they profile those conditions hence the name). When they do this poorly, editing is a major rabbit hole to transverse.

[feppe]

If it looks good in the booth,  unless you take it home and illuminate it with some really weird-ass illuminant, it will look fine.

I bet many of the new "eco" bulbs with spiked color curves, LEDs and whistles would fulfill the technical requirement for being weird-ass

[digitaldog]

I bet many of the new "eco" bulbs with spiked color curves, LEDs and whistles would fulfill the technical requirement for being weird-ass

Not necessarily all those.

[Eric Myrvaagnes]

I only recommend profile editing for masochists and those that can do this with OPM (Other People's Media) while being paid by the hour.

     


Great quote!

[MHMG]

Hello,

Am I the only one to be tempted to significantly brighten every image before printing in order to satisfy the average customer ?

Thanks, and a happy new year to everyone around here.

Well, adding to what Digitaldog said, you need great lighting to see great color.  While we can "tune" a print to be lighter and arguably look "a little better" under dim light, it will never look great under such a poor lighting situation, and it will absolutely look worse than a fully optimized print if you ever decide to use better lighting.  It is when reflection prints are well lit that the color reproduction in reflection prints becomes optimal.  BTW, "well lit" doesn't necessarily mean viewing booth illuminance levels. Lighting designers know full well that you can enable people to adapt well to lower light levels if you also spot light the print and keep surrounding conditions dimmer.  Most home display conditions use neither strong illumination levels nor spot lighting techniques, so prints almost never look great in typical home viewing environments, no matter how you skew the tone curve to compensate.  

For the artist who wants to produce an optimum work of art, it begs the question.  Do you compromise your colors and tones for dim lighting in order to please the consumer who cares little whether the piece is well lit, or do you produce a work that "sings" when optimum display conditions are met?  I personally strive for the latter, but others may decide pleasing the uneducated consumer is more important.

cheers,

Mark

[pfigen]

I make a lot of fine art prints for all sorts of people who know absolutely nothing of profiles or calibration, and for the most part I never really have a problem. I open the file and if it seems within bounds on my Artisan (and yes I still have three in great shape) I go ahead and print it. If I come across an image that seems like it might have been adjusted on a too bright monitor, I'll probably call that client and get a better feeling for what they're looking for, or just make them a test print.

As far as profile editing goes, I don't think it's really necessary for this type of situation, unless perhaps, you have a ton of prints to do for the same client/output. Profile editing in and of itself is really not that hard - once you get your head around it and have a good profile editor. I've edited hundreds of profiles for minor bugs in ProfileMaker's bright yellow response, others to force white matching when proofing (arguably more difficult), and after using both the Gretag edit module and the Kodak editor, I have to say that the Gretag is the far superior tool. You'd think that you might want the familiar Photoshop adjustments that Kodak uses, but in reality, the tool in Gretage Edit are much better suited to this task. It is my go to app for editing.

[tongelsing]

I personally strive for the latter, but others may decide pleasing the uneducated consumer is more important.

Its good to remember that the clients are paying you and thanks to your clients you can work!

[Slobodan Blagojevic]

Its good to remember that the clients are paying you and thanks to your clients you can work!

No, no, it is the other way around! Your clients should thank you for the opportunity to part with their money for the benefit of such a smart and educated and creative artist, who knows so much more than they do and does not shy from shoving it down their throats. Your clients should thank you for taking them into a dark room and then blinding them with a metal halide light, so that they can better see all those totally awesome piles of RGB numbers (which they mistakenly thought, in their ignorant ways, is just a nice print they would like to see on their wall).

[MHMG]

Its good to remember that the clients are paying you and thanks to your clients you can work!

If you are printing work for paying clients, "the customer is always right" applies in spades. If they want a lighter print, or anything else for that mater, by all means do your best to make it for them.  I was simply trying to point out that the most beautiful reflection print quality will be compromised when you are being asked to print for optimal viewing under less than optimal lighting conditions. If said prints are later illuminated properly they will then become the inferior print.  A similar problem exists for printmakers using bright white paper that depend on high levels of OBAs for that look.  Many professional framing shops will talk the customer into conservation glass or UV filtering acrylic with good intentions and the customer will accept, but it defeats the bright white intended appearance of the image no matter how one illuminates it.  

When I used the word "uneducated" in my original post, I didn't mean to imply "stupid".  Explaining options to good clients and then letting them make an informed choice is certainly preferable.  If I, personally, was the paying customer, I would appreciate a dialogue with the print provider on image quality issues, lighting issues, and preservation issues. No doubt some customers will truly appreciate being informed of the choices while others will find it "too much information", and no doubt some printmakers won't want to spend the additional time in such a dialogue with the customer, but others will.

regards,

Mark

[Wayne Fox]

If it looks good in the booth,  unless you take it home and illuminate it with some really weird-ass illuminant, it will look fine.

I suppose if the booth is too bright you might have problems.  In this case the colors might look OK on the wall, but the print might  be dark for many locations.

[David Saffir]

Someone may have commented on this already - but:

My proofing setup includes both a viewing box, and a set of tracklights (Solux) using 3500 (or 3800) bulbs. This gives me a final check on the print, in a lighting environment that is used by many of my customers.

In fact, I make many local deliveries in the evening, so I almost always wind up looking a print under tungsten or halogen lighting in the client's office or home.

The reality is that they want it to look good in a display environment.

Now, if a file is provided to client and they will make additional prints on their own, a different scenario unfolds....

David Saffir

[Jonathan Wienke]

"It has been said that democracy is the worst form of government, except all the others that have been tried."
-Sir Winston Churchill

This same principle applies to color management. As annoying and limited and imperfect as it may be, it beats all of the alternatives hands-down. Applying standards to camera captures, monitor images, and prints may not always yield perfectly satisfactory results by default, but any alternative based solely on guesswork and manual tweaking is certain to be worse.

[Peter_DL]

Let's assume that I have a perfectly color managed workflow and that I get good matches between prints & display. Let's also assumre that the customer doesn't. Instead, what the customer sees is a dark looking image compared to their beloved uncalibrated display which has a native 400cd/m² luminance, high contrast, high saturation, high everything. Furthermore, they don't light prints well which results in prints looking even darker.
...
Am I the only one to be tempted to significantly brighten every image before printing in order to satisfy the average customer ?

IF an image really looks good at 400 cd/m2, let’s say on a profiled but not calibrated monitor,
it was most likely not processed an edited enough to look still good
a.) on a calibrated screen at 100 cd/m2 or so,
and b.) at further reduced luminance w/Softproof > Simulate Paperwhite, or on print.
a.) and b.) will appear dark and dull.

In other words, we edit images to meet a certain output dynamic range & white luminance.
While this can be a trial and error thing at times, it’s not really a CM issue – IMO.

Peter

--

[Shirley Bracken]

Good thread, thanks everyone.

[Guigui]

Indeed, thanks everyone for the input. Since I'm really no expert on the subject I don't have any comment for the moment. Still thinking over/trying to figure out everything that has been said so far.