Open Photography Forums  
HOME FORUMS NEWS FAQ SEARCH

Go Back   Open Photography Forums > Color Management Discussion > CM Theory and Practice

CM Theory and Practice Profiles, color spaces, perception, science.

Reply
 
Thread Tools Display Modes
  #1  
Old June 16th, 2010, 09:04 AM
Cody White Cody White is offline
Member
 
Join Date: Oct 2008
Posts: 307
Default Monitor Profiling: Why does my color on my monitor look right to me but not others?

What do you know about it?

Here are my laptop specs:

AMD Sempron 3200+,
128MB ATI RADEON(R) XPRESS 200M w/Hypermemory(TM)
15.4" WXGA BrightView Widescreen (1280x800)





This thread follows Cody White's original thread here where corrected pictures appeared different to various viewers than the appearance the photographer described or intended. ADK

Last edited by Asher Kelman; June 16th, 2010 at 01:14 PM.
Reply With Quote
  #2  
Old June 16th, 2010, 11:00 AM
Asher Kelman Asher Kelman is offline
OPF Owner/Editor-in-Chief
 
Join Date: Apr 2006
Posts: 34,778
Default

Quote:
Originally Posted by Cody White View Post
what do you know about it?
here is my laptop specs. AMD Sempron 3200+,
128MB ATI RADEON(R) XPRESS 200M w/Hypermemory(TM)
15.4" WXGA BrightView Widescreen (1280x800)

Cody,

What your screen shows depends on how it is instructed to assign a code for each color. Each color in your picture has one such code. You screen has a profile that might be bright and fine for games, business graphics and text. However, what you are seeing is not necessarily what others would see on a profiled monitor.

So what is that profile? In your graphics card, (monitor or computer), there's a look up table. Here the software sends a code to the card and sends on a color mapped to look like the real color.

Unfortunately the conversion factors and the way colors are mapped to what the screen can deliver is likely to be particular to that screen! What's important is what is being shown on the screen. Here's where we can control the process. We simply add a simple device to closely sit of the face of the center of the screen. It's like a mouse. It measure light intensity and color. It's either a colorimeter or spectrophotometer. Software is loaded into the computer. Wear a grey shirt, remove any bright colors in the room, dim or turn off the lights and click on the "start" button. The screen will cycle through a series of screen colors and shades of grey.

As the result of this process, the lookup table, LUT, is now reset with numbers which now should assign accurate screen colors to the stream of numbers it receives.

Now some screens do not have electronics capable of delivering the exact same color each time given the same instruction 5 minutes later or tomorrow or next week or away from the center of the screen!

This is why there's a market for the more expensive screens that professionals and serious enthusiasts use. Apple displays in the desktop computers are pretty good. Dell used to be problem, as it's screens catered for gamers and they want flashy colors. Now I've heard they are much better.

The solution is to profile your monitor. Find someone who has a Spyder or Eye One or similar to get your monitor profiled. Do it every 3 months as electronics drift.

That's my feedback, hope it's a help!

Asher
__________________
Follow us on Twitter at @opfweb

Our purpose is getting to an impressive photograph. So we encourage browsing and then feedback. Consider a link to your galleries annotated, C&C welcomed. Images posted within OPF are assumed to be for Comment & Critique, unless otherwise designated.
Reply With Quote
  #3  
Old June 16th, 2010, 11:09 AM
John Angulat John Angulat is offline
Senior Member
 
Join Date: Oct 2008
Location: New York
Posts: 1,272
Default

Cody,
Asher brings up some very good points.
One key question is: does your display allow for individual adjustments to Red, Green and Blue?
If you cannot perform these adjustment, then I believe you cannot calibrate the display.
I'm not too familiar with the Xpress 200 chipset, but it is primarily geared for gaming.
That probably means colors are artificially boosted to enhance the gaming experience.
That, in turn, will cause your images to appear "off" when viewed on any other display (such as we see here when we view your images).
__________________
Warm regards,
John Angulat
www.lightsimage.com
Reply With Quote
  #4  
Old June 16th, 2010, 11:45 AM
Asher Kelman Asher Kelman is offline
OPF Owner/Editor-in-Chief
 
Join Date: Apr 2006
Posts: 34,778
Default

Quote:
Originally Posted by John Angulat View Post
Cody,
Asher brings up some very good points.
One key question is: does your display allow for individual adjustments to Red, Green and Blue?
If you cannot perform these adjustment, then I believe you cannot calibrate the display.
I'm not too familiar with the Xpress 200 chipset, but it is primarily geared for gaming.
That probably means colors are artificially boosted to enhance the gaming experience.
That, in turn, will cause your images to appear "off" when viewed on any other display (such as we see here when we view your images).
John,

You make a good point. LCD screens cannot alter the output of RGB, AFAIK! However, one can remap where one sends an instruction. So if a color in Adobe RGB is 130, 260, 174 then to show that, the corrected profile might instead call for 130, 274, 174 to map the wanted color to what where on the LCD screens native gamut that particular color is found. If that color does nor exsit in the colors available in the screens gamut, then it will give a color that perceptually, (hopefully, in the context of the actual picture), gives a color close to what's called for.

So unlike CRT monitors with separate electron guns for shooting at the screen phosphors, which can be controlled to some degree, LCD screens cannot give anything but the native colors they possess. It's which of these colors are chosen for a particular instruction for each dot on the screen that is altered in profiling a modern LCD screen.

So here we have brought in two terms
  • Calibrating: when the RGB outputs can be altered

  • Profiling: where we have a LUT, look up table, in the graphics card or monitor to reassign color designations to allow the screen to seem to show what the numbers would represent if the screen had unlimited gamut and perfection.

Asher
__________________
Follow us on Twitter at @opfweb

Our purpose is getting to an impressive photograph. So we encourage browsing and then feedback. Consider a link to your galleries annotated, C&C welcomed. Images posted within OPF are assumed to be for Comment & Critique, unless otherwise designated.
Reply With Quote
  #5  
Old June 16th, 2010, 11:50 AM
John Angulat John Angulat is offline
Senior Member
 
Join Date: Oct 2008
Location: New York
Posts: 1,272
Default

Ahhhhh....seems I'll gain an education along with Cody!
Question - if an LCD display, or it's graphics card or chipset lacks the ability to allow for individual RGB adjustments, then the display cannot be either calibrated nor profiled - correct?
__________________
Warm regards,
John Angulat
www.lightsimage.com
Reply With Quote
  #6  
Old June 16th, 2010, 12:22 PM
Asher Kelman Asher Kelman is offline
OPF Owner/Editor-in-Chief
 
Join Date: Apr 2006
Posts: 34,778
Default

Quote:
Originally Posted by John Angulat View Post
Ahhhhh....seems I'll gain an education along with Cody!
Question - if an LCD display, or it's graphics card or chipset lacks the ability to allow for individual RGB adjustments, then the display cannot be either calibrated nor profiled - correct?
John,

An LCD is profiled. What happens is that the LUT is re-populated so that the gamut of the colors in the file is remapped to the existing possible colors in the world of the LCD monitor. It's a matter of saying, O.K. this yellowish green in the gamut of the image file is best shown on this guys monitor by calling for a color a color with a different factory RGB designation perhaps but much closer to what we need to get the perception to be as close to the required color as possible.

The monitors we use have a small set of colors than the colors delivered in the RAW 16 BIT wide RGB color file. The number of colors and volume of the color world or "gamut" shrinks when we go down to 8 BIT and sRGB. So now there are less colors to have to find in the display or for the printer to print. That's why we should keep pictures in a wide RGB Gamut like Adobe RGB 91988) and 16 BIT until we are ready to deliver the file to a dumb drug store print tech or a simple printer.

When we bring any file into out computer, the colors have to be remapped in order to take account of the fact that some colors in the file cannot be discovered in the monitor, so the computer assigns the color that's the best for the job. That's perceptual remapping.

So today, with the availability of wider gamut monitors, many photographers have upgraded. This now allows us to see more of the colors that are present and modern printers by HP, Epson and Canon, for example can print out for us.

Perceptual remapping in our inadequate sRGB like monitors is really an amazing feat! To be able to allow us to design photographs and adjust colors we cannot actually see is to me a miracle of engineering and science interfacing with knowledge of how we can be tricked into appreciating what is not really being shown! Still, the lest remapping the better.

So the bottom line here is as follows:
  • Include a standardized high quality reference card for color in each shoot for each lighting condition there. I use the WhiBal™ from Michael Tapes,




    it's IMHO, the very best bargain, it's waterproof, can be sanded to be cleaned if need be and will last for ever. One click with the eye dropped tool in Photoshop or other software, will get your colors balanced pretty well. Even better, perhaps, try the new color color card and automatic mapping software from from Xrite. It comes in a neat package called the Passport color Checker.


    For $99, your color will be in the ballpark, even though it might not look right on any non-profiled monitor.

  • Use RAW files so you can re-edit the best pictures with the best monitor

  • If your screen is not profiled, don't adjust colors on it, just crop and edit for local brightness but don't change the colors.

  • Profile your monitor and recheck every few months

  • Try to get a monitor that is known for accurate colors when profiled

I hope this helps to get folk started in bringing some way of preserving the true color of your files to the final destination, be it the printer or someone else's monitor on the WWW.

Asher
__________________
Follow us on Twitter at @opfweb

Our purpose is getting to an impressive photograph. So we encourage browsing and then feedback. Consider a link to your galleries annotated, C&C welcomed. Images posted within OPF are assumed to be for Comment & Critique, unless otherwise designated.
Reply With Quote
  #7  
Old June 16th, 2010, 02:37 PM
Doug Kerr Doug Kerr is offline
Senior Member
 
Join Date: May 2006
Location: Alamogordo, New Mexico, USA
Posts: 8,558
Default

Hi, John,

Quote:
Originally Posted by John Angulat View Post
Ahhhhh....seems I'll gain an education along with Cody!
Question - if an LCD display, or it's graphics card or chipset lacks the ability to allow for individual RGB adjustments, then the display cannot be either calibrated nor profiled - correct?
Not so on either front.

Let me first review the overall situation here.

We are concerned with two matters:

Calibration of the display chain

Here, we aspire to make the display chain (comprising the display driver, the graphics card, and the monitor itself) exhibit in its own right a standard color space, perhaps sRGB.

In reality, this objective is not usually precisely attainable (which is, in fact, why we need the next stage, profiling).

Profiling of the display chain

Here, we develop a precise description of the color space actually exhibited by our already calibrated display chain.

This will then be used by profile-aware applications to transform the colors in an image from the display space in which they (ostensibly) are recorded to the color space of our (calibrated) display space.
Actually, to be more precise, the application first (at least conceptually) transforms the color from the existing color space to a certain standard color space (based on the profile for the source color space) and then from that color space to the color space actually exhibited by our display chain (based on its profile).
Back to calibration

Now let me return to the calibration aspect of this. The color space exhibited by a display chain is affected by several things, notably:

• The actual current physical situation of the monitor.
• The settings of user controls on the monitor, including (as applicable) color temperature or white point preset, channel bias and or gain, and so forth.
• The population of the lookup tables (LUT) in the display card. These tables define, pointwise, a transfer function from the digital description of an image color to the digital input to the display mechanism itself (perhaps to the digital-to-analog converter that develops the monitor channel drive voltages).

The calibration process is done by a package that includes a colorimeter to measure the color emitted by the display screen and a software package that administers a test protocol, analyzes the result, and finally develops a suite of values for the LUT (and implants them in the LUT, and arranges for them to be re-implanted whenever the system restarts).

This rig can usually do this regardless of the current physical state of the monitor and the current state of the user controls. (They merely influence the nature of the "subject of the exercise" as it shows up for the calibration process.)

However, often the process works best if the user controls are set (perhaps arbitrarily) to certain values. Thus, if our monitor has separate RGB controls (perhaps both bias and gain for each channel, or maybe only one of those), we may be instructed to set them to some arbitrary values before proceeding with the calibration protocol itself.

Or, in some cases, the calibration suite will lead us through a preliminary exercise to set these to the most advantageous place.

But in any case, after that is done (or if it cannot be done), the calibration suite takes the monitor, as it now lays, and seeks to develop a suite of LUT values so that the color space exhibited by the display chain when the LUT is repopulated will be as close as possible to the target color space (again, often sRGB).

Back to profiling

Then, all that having happened, the suite moves on to the profiling phase, developing a precise description of the display chain as the whole thing now lays (with the LUT repopulated).

Neat, wot!

Best regards,

Doug
Reply With Quote
  #8  
Old June 16th, 2010, 03:30 PM
Cody White Cody White is offline
Member
 
Join Date: Oct 2008
Posts: 307
Default

how does one adjust color on a WXGA flat screen monitor?
Reply With Quote
  #9  
Old June 16th, 2010, 03:37 PM
Asher Kelman Asher Kelman is offline
OPF Owner/Editor-in-Chief
 
Join Date: Apr 2006
Posts: 34,778
Default

Cody,

It's just an LCD so you don't adjust the color of the monitor, it's color is fixed. However you can maneuver it to showing the right colors! That's what profiling is! It remaps the colors into the capabilities of your screen. The end result seems like you are successfully adjusting the colors of the screen but actually it's just a matter of giving the screen a new order for each color point required to build your picture.

It's so simple. Buy or borrow a profiling device. Load the software. Plug in the device to your USB port, hang the puck on the center of the monitor, turn off the lights, click the button and the rest is automatic. now your monitor is profiled as best as can be for that monitor at that time.

It may need to be redone every week or month or 6 months. That's why folk actually buy the device.

Asher
__________________
Follow us on Twitter at @opfweb

Our purpose is getting to an impressive photograph. So we encourage browsing and then feedback. Consider a link to your galleries annotated, C&C welcomed. Images posted within OPF are assumed to be for Comment & Critique, unless otherwise designated.
Reply With Quote
  #10  
Old June 16th, 2010, 03:49 PM
Doug Kerr Doug Kerr is offline
Senior Member
 
Join Date: May 2006
Location: Alamogordo, New Mexico, USA
Posts: 8,558
Default

Hi, Asher,

Quote:
Originally Posted by Asher Kelman View Post
It's just an LCD so you don't adjust the color of the monitor, it's color is fixed.
The color shown by an LCD monitor, for any given set of inputs to the monitor interface, is not fixed. It for example varies with the user controls for such things as white point, channel bias and gain (if provided), and so forth, as well as by the current life cycle state of the back lights, etc.

The color shown by the monitor for any given color input to the display driver (the beginning of the display chain) is not also not fixed, but varies with all the things mentioned above as further affected by the values in the LUT.

Quote:
It's so simple. Buy or borrow a profiling device. Load the software. Plug in the device to your USB port, hang the puck on the center of the monitor, turn off the lights, click the button and the rest is automatic. now your monitor is profiled as best as can be for that monitor at that time.
Most profiling "systems" also do calibration. If you have that kind, do you advocate turning that part off, and leaving the LUT populated to the factory load? (They normally offer that option.)

There is no disadvantage to that, so long as all the viewing in which you are interested in "proper" rendition is done with the benefit of a profile.

The advantage of doing the calibration is that we get "the best available result" when viewing through a chain that does not take advantage of a profile (such as with many image viewers, browsers without profile capability, and so forth).

Best regards,

Doug
Reply With Quote
  #11  
Old June 16th, 2010, 04:53 PM
John Angulat John Angulat is offline
Senior Member
 
Join Date: Oct 2008
Location: New York
Posts: 1,272
Default

Asher, Doug -
As always...a learning experience!
Thank you both for the detailed explanations.

Asher, you're spot-on with the necessity to use a reference card at the beginning of any shoot (or more often if lighting changes).
I try to do so, but admittedly not 100% of the time.
My question regarding the the ability to profile if you lack RGB controls on your monitor stemmed from my own efforts at profiling.
I use the Eye-One Display colorimeter and my assumption one could not profile w/o RGB controls came from using the "advanced" choice at the beginning of the process.
As you probably know, after setting your brightness the next step in "advanced" are individual color adjustments.
Doug, thank you for clarifying this for me:
"This rig can usually do this regardless of the current physical state of the monitor and the current state of the user controls. (They merely influence the nature of the "subject of the exercise" as it shows up for the calibration process.)
However, often the process works best if the user controls are set (perhaps arbitrarily) to certain values. Thus, if our monitor has separate RGB controls (perhaps both bias and gain for each channel, or maybe only one of those), we may be instructed to set them to some arbitrary values before proceeding with the calibration protocol itself.
Or, in some cases, the calibration suite will lead us through a preliminary exercise to set these to the most advantageous place."

I did not realize one could skip this rather tortuous step by choosing "automatic" and the software would still generate a credible profile. I just tried it and am now cursing myself for being so stupid for so long!

Thanks again and Cody, I hope this has helped you understand the process a bit better.
Always a learning experience here at OPF!
__________________
Warm regards,
John Angulat
www.lightsimage.com
Reply With Quote
  #12  
Old June 16th, 2010, 06:23 PM
Doug Kerr Doug Kerr is offline
Senior Member
 
Join Date: May 2006
Location: Alamogordo, New Mexico, USA
Posts: 8,558
Default

Hi, John,

Quote:
Originally Posted by John Angulat View Post
I did not realize one could skip this rather tortuous step by choosing "automatic" and the software would still generate a credible profile.
And before that, it does a credible job of calibration (and the profile is developed for the calibrated display chain).

Please try to grasp the whole chain of events here.

If we ignore the existence of the calibration process (perhaps due to folklore that there is no such thing or whatever), we can get into some incorrect and confusing thoughts.

There are also some hazards. For example, suppose we somehow accidentally disable the startup software that reloads the LUT every time we reboot.

Now, the display system will not be calibrated, but profile-aware applications will be using a profile that describes the display chain's calibrated behavior.

Thus we will have "incorrect" color display results. And we'll have no idea what the problem might be.

By rteh way, you may be able to see the LUT being reloaded during startup. You may, at a certain point during startup, see the color of the desktop change. That is when the LUT is reloaded.

By the way, the LUT data is stored in a proprietary "caboose" in the profile file, called the vcgt tag, from “video card gamma table” (“gamma table” referring to the LUTs). The startup software knows (from a registry entry) what profile file has been nominated for use with our display. The startup software fetches the LUT data from that file's vcgt tag and implants it into the LUT.

Best regards,

Doug
Reply With Quote
  #13  
Old June 16th, 2010, 06:49 PM
Asher Kelman Asher Kelman is offline
OPF Owner/Editor-in-Chief
 
Join Date: Apr 2006
Posts: 34,778
Default

Quote:
Originally Posted by Doug Kerr View Post
Hi, John,



And before that, it does a credible job of calibration (and the profile is developed for the calibrated display chain).
Doug,

The "calibration" does not I think apply to his laptop apart from setting WP. It's not an Eizo, or NEC SpectraView with 12, 14 or 16 BIT calculations! So how important/useful it is to not let the profiling software do everything automatically, I don't know. Frankly, I'd keep it simple and just use the Eye One spectrophotometer to profile it at regular intervals. No brainware needed. I really am not convinced that one would likely improve on that with the ordinary LCD monitors in use.

Maybe you could try that out.

Asher
__________________
Follow us on Twitter at @opfweb

Our purpose is getting to an impressive photograph. So we encourage browsing and then feedback. Consider a link to your galleries annotated, C&C welcomed. Images posted within OPF are assumed to be for Comment & Critique, unless otherwise designated.
Reply With Quote
  #14  
Old June 16th, 2010, 07:09 PM
Doug Kerr Doug Kerr is offline
Senior Member
 
Join Date: May 2006
Location: Alamogordo, New Mexico, USA
Posts: 8,558
Default

Hi, Asher,

Quote:
Originally Posted by Asher Kelman View Post
The "calibration" does not I think apply to his laptop apart from setting WP.
I don't know much about laptop display systems. Are you saying that this one does not have settable LUTs in the display controller?

Quote:
So how important/useful it is to not let the profiling software do everything automatically, I don't know.
I have never suggested that we should not let the calibration/profiling software do everything automatically. My only point is that is normally does two things. First, it calibrates the display chain. Then, it makes a profile for the calibrated chain. Normally, fully automatically.

Quote:
Frankly, I'd keep it simple and just use the Eye One spectrophotometer to profile it at regular intervals.
To profile, but not calibrate? Are you sure you mean that?

I think you are still not reading what I am saying.

By the way, there are utilities that will pull the LUT settings out of the display hardware and show them (or put them into a vcgt tag). I haven't tried them on any of my laptops. (It will take me a while to remember how they work.)

Best regards,

Doug
Reply With Quote
  #15  
Old June 16th, 2010, 07:17 PM
Asher Kelman Asher Kelman is offline
OPF Owner/Editor-in-Chief
 
Join Date: Apr 2006
Posts: 34,778
Default

Quote:
Originally Posted by Doug Kerr View Post
I think you are still not reading what I am saying.
Doug,

I am paying attention. I'm just not sure about the LUT tables on laptops and regular graphics cards and monitors. But does it really matter? Test with and without attempting to calibrate and see if there's a difference. I'd be interested to know.

Asher
__________________
Follow us on Twitter at @opfweb

Our purpose is getting to an impressive photograph. So we encourage browsing and then feedback. Consider a link to your galleries annotated, C&C welcomed. Images posted within OPF are assumed to be for Comment & Critique, unless otherwise designated.
Reply With Quote
  #16  
Old June 16th, 2010, 07:39 PM
John Angulat John Angulat is offline
Senior Member
 
Join Date: Oct 2008
Location: New York
Posts: 1,272
Default

Good thread, but I think my questions have caused us to go off tangent.
For that, Cody, I apologize. I never meant to hijack your thread.

So, I'll ask - Cody, has any of this very valuable information has helped?
If not, or if not completely, please ask again.
We're all here to help.
__________________
Warm regards,
John Angulat
www.lightsimage.com
Reply With Quote
  #17  
Old June 16th, 2010, 07:46 PM
Doug Kerr Doug Kerr is offline
Senior Member
 
Join Date: May 2006
Location: Alamogordo, New Mexico, USA
Posts: 8,558
Default

Hi, Asher,

Quote:
Originally Posted by Asher Kelman View Post
Doug,

I am paying attention. I'm just not sure about the LUT tables on laptops and regular graphics cards and monitors. But does it really matter?
Well, as to the question, "Is it worthwhile to calibrate/profile the display chain on (typical) laptop LCD displays", the answer may well be "no".

I just didn't want this answer to that question to be on the table:

• There is no such thing as calibrating a laptop LCD system

By the way, on my Dell Latitude D610 laptop, which has a fairly dreadful display system, there seem to be LUT's that work in the usual way.

Best regards,

Doug
Reply With Quote
  #18  
Old June 18th, 2010, 11:29 PM
Cody White Cody White is offline
Member
 
Join Date: Oct 2008
Posts: 307
Default

Is it to my understand, if one holds the printed picture besides the monitor and it matches lightness and color your monitor is pretty much dead on?
Reply With Quote
  #19  
Old June 19th, 2010, 03:17 AM
Asher Kelman Asher Kelman is offline
OPF Owner/Editor-in-Chief
 
Join Date: Apr 2006
Posts: 34,778
Default

Quote:
Originally Posted by Cody White View Post
Is it to my understand, if one holds the printed picture besides the monitor and it matches lightness and color your monitor is pretty much dead on?
That depends. If it satisfies you, then it's O.K. If you want it to mean the same thing to someone else, then the colors just have to be measured. There's no getting around that.

Asher
__________________
Follow us on Twitter at @opfweb

Our purpose is getting to an impressive photograph. So we encourage browsing and then feedback. Consider a link to your galleries annotated, C&C welcomed. Images posted within OPF are assumed to be for Comment & Critique, unless otherwise designated.
Reply With Quote
  #20  
Old June 19th, 2010, 05:56 AM
Andrew Rodney Andrew Rodney is offline
Member
 
Join Date: May 2006
Posts: 467
Default

Quote:
Originally Posted by Cody White View Post
Is it to my understand, if one holds the printed picture besides the monitor and it matches lightness and color your monitor is pretty much dead on?
Yup, pretty much. That's the goal: Print to display match.
__________________
Andrew Rodney
Author "Color Management for Photographers"
http://www.digitaldog.net/
Reply With Quote
  #21  
Old June 19th, 2010, 06:43 AM
John Angulat John Angulat is offline
Senior Member
 
Join Date: Oct 2008
Location: New York
Posts: 1,272
Default

Quote:
Originally Posted by Cody White View Post
Is it to my understand, if one holds the printed picture besides the monitor and it matches lightness and color your monitor is pretty much dead on?
Quote:
Originally Posted by Andrew Rodney View Post
Yup, pretty much. That's the goal: Print to display match.
Hi Andrew,
Doesn't that have to assume one of the two (display or printer) is already "correct"?
Otherwise, one is simply adjusting "to match".
If the display is off calibration, now we've made the print "match" the incorrect display, or vice versa.
I'm confused...
I'll give an example -
Let's say an image is "correct" leaving the camera.
I view the image on my "incorrect" monitor (ex. monitor is too bright), so I now will adjust the print, believing my print is overexposed. Although the image may now appear "correct", in actuality it is now underexposed.
I print the image. What are my results? If my printer is reasonably accurate the print should appear underexposed. If not, then adjustments are made to the printer to produce a "correct" print.
However, the image file is still incorrect (by nature of my fiddling with all the adjustments).
It may look "correct" to me, but to anyone else it will appear underexposed.
If it is printed anywhere else, it will print underexposed.

I don't know if my pretzel logic makes any sense.
__________________
Warm regards,
John Angulat
www.lightsimage.com
Reply With Quote
  #22  
Old June 19th, 2010, 08:05 AM
Doug Kerr Doug Kerr is offline
Senior Member
 
Join Date: May 2006
Location: Alamogordo, New Mexico, USA
Posts: 8,558
Default

Hi, Andrew,

Quote:
Originally Posted by Andrew Rodney View Post
Yup, pretty much. That's the goal: Print to display match.
Perhaps, tactically (assuming that your images are never used but to print on your own printer). But that display-to-print match is hardly an indication that the display chain is "dead on".

Best regards,

Doug
Reply With Quote
  #23  
Old June 19th, 2010, 12:42 PM
Asher Kelman Asher Kelman is offline
OPF Owner/Editor-in-Chief
 
Join Date: Apr 2006
Posts: 34,778
Default Cody, Let's look at goals needs and standards!

Quote:
Originally Posted by Cody White View Post
Is it to my understand, if one holds the printed picture besides the monitor and it matches lightness and color your monitor is pretty much dead on?
I replied, without wanting to repeat what we have already presented, in the most succinct form "That depends. If it satisfies you, then it's O.K. "

Andrew then added

Quote:
Originally Posted by Andrew Rodney View Post
Yup, pretty much. That's the goal: Print to display match.
The "Yup pretty much", aside, "Print to display" for your own world might be fine enough for your standards and needs. However, there's more I wrote:

If you want it to mean the same thing to someone else, then the colors just have to be measured [somewhere]. There's no getting around that.

So John points out correctly that if one is doing this for someone else and making adjustments on the monitor until one "likes" the result, the printed output might be fine for you, the printer's preferences, but it's a roulette wheel of chance as to whether the thousands of colors are represented with fidelity to real life.

Quote:
Originally Posted by John Angulat View Post
Hi Andrew,
Doesn't that have to assume one of the two (display or printer) is already "correct"?
Otherwise, one is simply adjusting "to match".
If the display is off calibration, now we've made the print "match" the incorrect display, or vice versa.
I suppose, with exceptional luck, both monitor and printer could be off to the same extent all over the world of colors each can make. However, if you and your photographer have a perfect recall for colors you both examine the pictures together in room light and in the daylight and find the pictures to match exactly the colors in real life, then you are fine!

I have watched Nicolas Claris in Bordeaux meticulously do this with an expert printer using a giant industrial CMYK press for his incredibly beautiful marine images, and yes the two of them together can do that. However, Nicolas uses a fully calibrated and profiled chain of instruments in his photography. The printer was chosen for his exceptional quality and ability to work with the clients to get the final adjustments right even on his superbly calibrated and profiled printing machines.

So let's get back to the task at hand. What is it? We can look at the quality of the monitor settings according to the following levels of needs:

1. Casual home printing: Print without altering the color. Have a camera (with auto or correctly set white balance or not), choose nice fun pictures on the screen, send to the printer without any changes.

If you like the print, the match between monitor and printer doesn't rsally matter! You have a system that works for your needs!


2. Personal use: You adjust pictures on the monitor: Once again, if your printer's output works for your needs, great! "Print to match" is fine.


3. Professional Printing: Without knowing that what you are seeing is accurate, you are unlikely to be able to print reliably and accurately from disparate sources, and especially where folk have made corrections on their own monitors, calibrated or not. If you have good judgement of color you might get away with the output, matching print to the monitor. Unfortunately you will not be able to do this well and to compete. Why, because we already know that your output to us on the internet is, to say the least, uncomfortably odd!

To overcome this, there are two options,

1. Shoot in auto mode or set the color* appropriately to the light. Then, share the pictures without adjusting the color!

2. Calibrate/profile your monitor at the very least.

Asher

*to tungsten, flash, shady etc, (according to the predominant light used), or else take a picture of a grey card and set you camera to that.
__________________
Follow us on Twitter at @opfweb

Our purpose is getting to an impressive photograph. So we encourage browsing and then feedback. Consider a link to your galleries annotated, C&C welcomed. Images posted within OPF are assumed to be for Comment & Critique, unless otherwise designated.

Last edited by Asher Kelman; June 19th, 2010 at 02:01 PM.
Reply With Quote
  #24  
Old June 19th, 2010, 01:46 PM
John Angulat John Angulat is offline
Senior Member
 
Join Date: Oct 2008
Location: New York
Posts: 1,272
Default

Thanks Asher,
I breathed a rather long sigh of relief, for I thought for a moment all the $$ and effort expended to purchase adequate printers, monitors, printer profiles and colorimeters had all been for naught.
__________________
Warm regards,
John Angulat
www.lightsimage.com
Reply With Quote
  #25  
Old June 20th, 2010, 10:55 AM
Andrew Rodney Andrew Rodney is offline
Member
 
Join Date: May 2006
Posts: 467
Default

Quote:
Originally Posted by Doug Kerr View Post
Hi, Andrew,
Perhaps, tactically (assuming that your images are never used but to print on your own printer). But that display-to-print match is hardly an indication that the display chain is "dead on".
The goal is print to screen matching (WYSIWYG). That takes two items; the display and a print next to the display properly illuminated for that task. Doesn’t matter if its your printer or someone else’s. You edit the images and expect to view the print and display resulting in a visual match. That may or may not be “dead on”, only you the user can evaluate what that means (they will never, ever be 100%, an emissive device and a reflective print are different, we can’t break the laws of physics). People get their displays and outside printers to produce this kind of match every day. Its not easy, but it certainly is doable. It requires good profiles for the display and of course the output device and it requires proper soft proofing setup in Photoshop.

There is no way to say other than this that the display is “accurate” or dead on. If we could measure the display, a print, the illuminant of a print and end up with the magic settings to produce a visual match, one might say we have a numerical, colorimetric match. If that were true, we could set our display target calibration for say the recommended D65, 2.2 TRC, 140cd/m2 target and get a match. That rarely happens. For over a decade, we were told proofing is based on D50, yet calibrating a CRT to that aim point produced a ruddy, yellow appearance, hence we were told to calibrate to D65 despite the fact that the numbers are not even close to each other. And the fact that only true D50 or D65 device is 93 million miles from your display.

Just this week I took an NEC SpectraView II 2690 and a small Just Normlicht booth to a class I was teaching at Anderson Ranch. Built a custom profile for their Epson 3880 on Luster paper. Setup a soft proof using that profile and the Simulate check boxes on to produce the dynamic range soft proof. I calibrated using the above, so called “recommended” settings and the display to print match were not even close. Image on screen was too dim and way too cool. Set the target to D50, other settings were the same. Color was now matching, luminance was still too dim. Set Luminance to 150 cd/m2. Still too dim. Set luminance to 170cd/m2. Very, very close (some would say dead on) match. Booth could not be dimmed so I had to alter the display luminance for that match, otherwise I would have altered the booth. Contrast Ratio was set in SpectraView II for 300:1. If I wanted to also make a target for Matt paper, I would probably have lowered this to 150:1, maybe 200:1. Point is, I’d have to make several calibration targets, looked at the print and soft proof and seasoned to taste. And because I can build as many of these targets as I wish, and switch on the fly in the SpectraView II software, I would update this based on the material I was working with. I’d get a “dead on” match between display and print. That is all that matters!

When I took the test print from the Just booth and viewed it across the room, under Solux illumination at that print viewing area, it looked great. Did it match the display? Don’t know. The display is out of the equation, its way across the room now. When I took the print outdoors, under sun light, it looked great. Did it match the display? Who knows, its downstairs. But when editing my images, viewing the print and the display next to each other, I got a match. Again, that’s all that matters. We adapt to the white and luminance as we move the print from area to area. And the display doesn’t travel with that print. At that point, all we care about is the print continues to please (which it does). That’s all that matters.
__________________
Andrew Rodney
Author "Color Management for Photographers"
http://www.digitaldog.net/
Reply With Quote
  #26  
Old June 20th, 2010, 12:15 PM
Asher Kelman Asher Kelman is offline
OPF Owner/Editor-in-Chief
 
Join Date: Apr 2006
Posts: 34,778
Default

Quote:
Originally Posted by Andrew Rodney View Post

When I took the test print from the Just booth and viewed it across the room, under Solux illumination at that print viewing area, it looked great. Did it match the display? Don’t know. The display is out of the equation, its way across the room now. When I took the print outdoors, under sun light, it looked great. Did it match the display? Who knows, its downstairs. But when editing my images, viewing the print and the display next to each other, I got a match. Again, that’s all that matters. We adapt to the white and luminance as we move the print from area to area. And the display doesn’t travel with that print. At that point, all we care about is the print continues to please (which it does). That’s all that matters.
Andrew,

This process is exactly what I'd expect from a professional. This is the iterative process to get the intent of the photographer to print. However, to put something on the web with a far lesser monitor, one needs to have that dumb monitor at least in the ballpark.

Here the laptop display is way wrong so he is not able to send pictures to us that look right.

Asher
__________________
Follow us on Twitter at @opfweb

Our purpose is getting to an impressive photograph. So we encourage browsing and then feedback. Consider a link to your galleries annotated, C&C welcomed. Images posted within OPF are assumed to be for Comment & Critique, unless otherwise designated.
Reply With Quote
  #27  
Old June 20th, 2010, 12:17 PM
Asher Kelman Asher Kelman is offline
OPF Owner/Editor-in-Chief
 
Join Date: Apr 2006
Posts: 34,778
Default

Quote:
Originally Posted by Andrew Rodney View Post

When I took the test print from the Just booth and viewed it across the room, under Solux illumination at that print viewing area, it looked great. Did it match the display? Don’t know. The display is out of the equation, its way across the room now. When I took the print outdoors, under sun light, it looked great. Did it match the display? Who knows, its downstairs. But when editing my images, viewing the print and the display next to each other, I got a match. Again, that’s all that matters. We adapt to the white and luminance as we move the print from area to area. And the display doesn’t travel with that print. At that point, all we care about is the print continues to please (which it does). That’s all that matters.
Andrew,

This process is exactly what I'd expect from a professional. This is the iterative process to get the intent of the photographer to print. However, to put something on the web with a far lesser monitor, one needs to have that dumb monitor at least in the ballpark.

So, look at this

Quote:
Originally Posted by Cody
Well I had another go at this shot and I like it, but it looked a tad overexposed, so tonight I had another go at it.
First image, I did this one just an auto color, auto level, and auto contrast, then I think I did a color cast removal.


Quote:
Originally Posted by cody
Second image,
This one I took a little more time. Since I have the original .cr2 file.

I messed up caused I didn't keep notes on what I did to it.



So Andrew,

Working with others on the web, one has to be in the ballpark. Cody's current match system obviously doesn't work too well over the web, unless you find the second image satisfactory. Here the laptop display is way wrong so he is not able to send pictures to us that look right.

Asher
__________________
Follow us on Twitter at @opfweb

Our purpose is getting to an impressive photograph. So we encourage browsing and then feedback. Consider a link to your galleries annotated, C&C welcomed. Images posted within OPF are assumed to be for Comment & Critique, unless otherwise designated.
Reply With Quote
  #28  
Old June 20th, 2010, 02:28 PM
Doug Kerr Doug Kerr is offline
Senior Member
 
Join Date: May 2006
Location: Alamogordo, New Mexico, USA
Posts: 8,558
Default

Hi, Andrew,

It has been my understanding as follows:

The direct objective of calibration and profiling of a display chain is to make sure that the display, when rendering an image pixel recorded as a certain color in the digital representation (in whatever color space that is) will in fact emit that same color (allowance being made for the issue of chromatic adaptation considering the viewing environment).

My larger objective is likely to have the displayed image (as I perceive it) match the image printed by, for example, my printer (as I perceive it). I seek that by:

• calibrating and profiling my display, and
• profiling my printer.

But a match between the image on my display and output from my printer does not of itself tell us that the profiling of the display chain is "perfect". Both display and printer may both depart from the ideal situation in a way that compensates. That's fine if I only use the resulting files to print on my printer.

But this leaves us exposed to the prospect that an image edited (on my display) until it shows the artistic effect I seek will, when sent as a file to somebody else, and printed there on a "perfectly profiled" printer,. will produce a result that (if brought to my studio) would not match what I see on my display - would not in fact convey the artistic effect I had worked for.

What am I missing here?

Best regards,

Doug
Reply With Quote
  #29  
Old June 20th, 2010, 04:26 PM
Andrew Rodney Andrew Rodney is offline
Member
 
Join Date: May 2006
Posts: 467
Default

Quote:
Originally Posted by Asher Kelman View Post
Working with others on the web, one has to be in the ballpark.
I don’t know what “ballpark” means, especially when dealing with non color managed apps.

IF the image is tagged in sRGB and someone views said image in a color managed web browser, its going to be in the “ballpark” and preview the same for all users.
__________________
Andrew Rodney
Author "Color Management for Photographers"
http://www.digitaldog.net/
Reply With Quote
  #30  
Old June 20th, 2010, 04:28 PM
Andrew Rodney Andrew Rodney is offline
Member
 
Join Date: May 2006
Posts: 467
Default

Quote:
Originally Posted by Doug Kerr View Post
But a match between the image on my display and output from my printer does not of itself tell us that the profiling of the display chain is "perfect". Both display and printer may both depart from the ideal situation in a way that compensates. That's fine if I only use the resulting files to print on my printer.
Not if all sets of profiles are good profiles. If you don’t like the image appearance, change it (the sole reason we use products like Photoshop). You don’t have to soft proof and produce a matching print. You can view the image in its working space, represented as it appears on the display (sole output). But if you do want to print it, you can, and it will match what you saw when you soft proofed the image, using the printer profile for the simulation.
__________________
Andrew Rodney
Author "Color Management for Photographers"
http://www.digitaldog.net/
Reply With Quote
Reply

Bookmarks

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are Off
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -7. The time now is 12:09 PM.


Posting images or text grants license to OPF, yet © of such remain with its creator. Still, all assembled discussion © 2006-2017 Asher Kelman (all rights reserved) Posts with new theme or unusual image might be moved/copied to a new thread!