-

Jump to content



Photo
- - - - -

How Room Decor Can Corrupt Or Enhance A Perfect Display Calibration


This topic has been archived. This means that you cannot reply to this topic.
34 replies to this topic

#1 of 35 OFFLINE   GeorgeAB

GeorgeAB

    Second Unit

  • 468 posts
  • Join Date: Jan 28 2001
  • Real Name:G. Alan Brown
  • LocationDenver, CO

Posted May 17 2007 - 03:07 AM

The world's most perfect calibration instrument cannot measure how our brain interprets what our eyes see. Some attempts have been made to emulate how humans perceive light but science has yet to produce an instrument which tells the whole story.

SMPTE's human factors work resulted in their Recommended Practices document #166: "Critical Viewing Conditions For Evaluation Of Color Television Pictures." This is the work from which 6500K bias lighting is taken. The document actually devotes much more attention to color perception than eyestrain. Here are some links that dramatically demonstrate how ambient lighting and surrounding surface colors in the room can cause us to think we see better black levels and/or distorted colors that aren't really in the image.

http://www.echalk.co..../illusions.htm (Note particularly the "Colour perception" and "Colour perception 2" demonstrations.)

http://www.lottolab.org/ (enter the "lab," click the "Illusions" button and note particularly the "Brightness" demonstrations.)
http://www.lottolab....ionsoflight.asp (new and quicker link)

Professional monitor environments, where critical image analysis is conducted for mastering video programs, use tightly controlled lighting and neutral surfaces surrounding the display. The demonstrations above make very clear the importance of incorporating similar room conditions if image fidelity is desired. This material also makes abundantly clear how destructive to image fidelity the Philips 'Ambilight' colored light features really are.

Human visual perception is seldom sufficiently understood when consumer, and even many professional, display systems are designed and implemented. Since our human vision is so adaptive, we can think we perceive a "natural looking" image but actually do not if viewing environment conditions are incorrect. The demonstration material at the links above should provide considerable practical reinforcement for folks who have a hard time being persuaded by the theory alone or decades of imaging industry professional practice.

If you think there is some trick being used in the online images, try printing out the demonstration patterns and making your own paper masks. You will see that the only "trick" involved is being provided by your own brain. This is why even a perfectly aligned display device can indeed look different than the calibration report says. Conflicting viewing environment conditions, such as the wrong lighting or colored room surfaces within the observer's field of view, will ALWAYS distort how a video image appears. No calibration instrument can measure this function of the brain.

This technical article goes into more detail about practical applications for this topic: http://www.cinemaquestinc.com/ive.htm .

Best regards and beautiful pictures,
G. Alan Brown, President
CinemaQuest, Inc.

"Advancing the art and science of electronic imaging"

#2 of 35 OFFLINE   GeorgeAB

GeorgeAB

    Second Unit

  • 468 posts
  • Join Date: Jan 28 2001
  • Real Name:G. Alan Brown
  • LocationDenver, CO

Posted May 18 2007 - 08:43 AM

Oops! Sorry about the defective illusions link in the post above. I've repaired it.

#3 of 35 OFFLINE   GeorgeAB

GeorgeAB

    Second Unit

  • 468 posts
  • Join Date: Jan 28 2001
  • Real Name:G. Alan Brown
  • LocationDenver, CO

Posted May 20 2007 - 03:04 AM

People have been inquiring about where to get the SMPTE RP166-1995 document mentioned above. This technical reference is available from the Society Motion Picture and Television Engineers directly at: http://www.smpte.org...rIndex_3_07.pdf .

Half the cost can be saved by getting the Widescreen Review back issue of their special edition titled, 'Imaging Science Theatre 2000' from 1998, which includes a reprint of the document: http://cinemaquestinc.com/isf-mag.htm . This special edition also includes much more detail about display viewing environment conditions and imaging science principles. It should be in the library of anyone serious about ultimate video quality and video displays.

#4 of 35 OFFLINE   Kevin C Brown

Kevin C Brown

    Producer

  • 5,713 posts
  • Join Date: Aug 03 2000

Posted May 21 2007 - 12:39 PM

George- I haven't looked at the links yet, but I read what you wrote, and theoretically, I understand what you're saying.

But. Posted Image

If the room is illuminated properly, then does the surrounding color of "stuff" matter that much?

(OK, I just answered my own question, because that illumination, even as low as it should be, still has to bounce off what's in the room!)
If it's not worth waiting until the last minute to do, then it's not worth doing.

KevinVision 7.1 ...

#5 of 35 OFFLINE   GeorgeAB

GeorgeAB

    Second Unit

  • 468 posts
  • Join Date: Jan 28 2001
  • Real Name:G. Alan Brown
  • LocationDenver, CO

Posted May 23 2007 - 04:22 AM

Kevin,

Go to the links and examine the demonstrations. You will be amazed and convinced beyond any doubt how important room conditions are to reference imaging.

You will also see how fallacious the statement can be that many make in these forums, "My best judge of color accuracy is my eyes." Our eyes can be fooled very easily and adapt quite readily to changing conditions. Without an objective technical reference, such as a truly neutral color sample, D65 illumination, and/or NIST traceable color analysis instrumentation, our visual perception is unreliable as a calibration tool.

#6 of 35 OFFLINE   Brian D H

Brian D H

    Second Unit

  • 453 posts
  • Join Date: Sep 02 2004

Posted May 23 2007 - 05:05 AM

Quote:
Originally Posted by GeorgeAB
Without an objective technical reference, such as a truly neutral color sample, D65 illumination, and/or NIST traceable color analysis instrumentation, our visual perception is unreliable as a calibration tool.

Absolutely true, but not 100% useful. Unless my viewing conditions are perfect (and they aren't).

I want my whites to LOOK 100% white and my blacks to LOOK 100% black - likewise red should be red, blue should be blue, etc. It doesn't matter to me if they aren't truely the colors they are supposed to be as long as they LOOK that way to me given my viewing conditions. That having been said I will always want correct calibration as a good starting point - but knowing that my eye is going to effected by other colors in the area I can't be afraid to tweak it later to make it "look" correct. Until I get a perfectly neutral colored home theater (black), I'm going to have to tweak it off of "perfect" calibration. The illusions you posted at the top of this thread prove that.
Lurking at HTF Since 2001

#7 of 35 OFFLINE   GeorgeAB

GeorgeAB

    Second Unit

  • 468 posts
  • Join Date: Jan 28 2001
  • Real Name:G. Alan Brown
  • LocationDenver, CO

Posted May 23 2007 - 06:08 AM

"I want my whites to LOOK 100% white"

Which color of white are you referring to?

"It doesn't matter to me if they aren't truely the colors they are supposed to be as long as they LOOK that way to me given my viewing conditions."

You have no clue what the colors are supposed to look like in a program, unless you are the program's producer, or have reference viewing conditions. Image fidelity is simply faithfulness to the appearance of the original program.

The only way to insure correct perception of the original program is to follow the same rules adhered to by the program producer. Critical program content decisions in post production are made with a calibrated monitor, in a neutral viewing environment (Gray to white- not black. Black would be useful for a front projection environment, which is not typically used in post production), illuminated with D65 lighting.

You're not fully comprehending a major principle. Our human visual system is adaptive. Without an objective reference, you won't recognize what's accurate. You can only guess.

Guessing is good enough for most video consumers. Most consumers don't have their TV calibrated. They also don't care that much about artistic integrity or value image fidelity. As long as the image vaguely resembles what they consider "reality," it's good enough for them. If it "looks good" to them, then it's good enough for them.

Image fidelity is only preserved when we follow the rules established by imaging industry standards bodies. Display design and calibration must follow the rules. Display environments also must follow the rules or the perception of the image will be skewed. That's what the demonstrations above illustrate. If you think you can accurately compensate for distorted color perception by some intuitive re-adjustment of the display settings, you are mistaken.

#8 of 35 OFFLINE   Ennsio

Ennsio

    Second Unit

  • 264 posts
  • Join Date: Feb 22 2007

Posted May 23 2007 - 08:22 AM

So if there is a window above and behind my tv, should the curtain be gray (or something quite close to gray)?

#9 of 35 OFFLINE   GeorgeAB

GeorgeAB

    Second Unit

  • 468 posts
  • Join Date: Jan 28 2001
  • Real Name:G. Alan Brown
  • LocationDenver, CO

Posted May 23 2007 - 10:42 AM

It should be neutral gray to white. A patterned fabric is actually beneficial, as long as all the colors in the pattern are neutral ( black to white through the gray scale).

For a color match, you can use one of the photographer's 18% gray cards, if you can find one at a camera store. Folks have been reporting these on back order, even in Europe. You can get a 10-step Munsell Color Order System matte gray reference book here: http://cinemaquestin...eal_viewing.htm . The Munsell system is the reference for surface colors SMPTE uses.

#10 of 35 OFFLINE   Ennsio

Ennsio

    Second Unit

  • 264 posts
  • Join Date: Feb 22 2007

Posted May 23 2007 - 10:51 AM

Thanks. Why is a patterned fabric beneficial?

#11 of 35 OFFLINE   GeorgeAB

GeorgeAB

    Second Unit

  • 468 posts
  • Join Date: Jan 28 2001
  • Real Name:G. Alan Brown
  • LocationDenver, CO

Posted May 23 2007 - 02:19 PM

It will help relieve eyestrain. Our eyes tend to get fatigued when focusing on a relatively localized area. This is aggravated when also focusing on a single focal plane (the screen) for extended periods. Our eyes also impulsively tend to wander periodically.

In a typical monitor environment, it helps relieve fatigue to have a different focal plane near and behind the TV for our eyes to re-focus on from time to time, even if very briefly. If there is a pattern or texture on the wall (or in this case a drape), the eyes can re-focus more quickly. This is, of course, with low level 6500K backlighting illuminating the surface behind the TV.

#12 of 35 OFFLINE   Brian D H

Brian D H

    Second Unit

  • 453 posts
  • Join Date: Sep 02 2004

Posted May 24 2007 - 07:31 AM

George,

Please don't misunderstand. I am completely in favor of calibrating my system AND of having a neutral viewing environment. But since my viewing environment isn't going to be perfect in the forseeable future I have to compromise. Obviously I will start with correct calibration. At that point we seem to differ. You would say to then fix my viewing enviroment. I would say again that I can't. I don't have an extra room - my set is in the living room and given WAF I have limited color and light-control options. So...

Quote:
Originally Posted by GeorgeAB
"It doesn't matter to me if they aren't truely the colors they are supposed to be as long as they LOOK that way to me given my viewing conditions."

You have no clue what the colors are supposed to look like in a program, unless you are the program's producer, or have reference viewing conditions. Image fidelity is simply faithfulness to the appearance of the original program.

Yes and no. I wasn't asssuming that I'd be subjectively judging color from a movie. I was assuming that I'd do it on a test pattern from a good calibration disc. I other words, I'd use the disc to calibrate and then use the test patterns to subjectively tweak it off of perfect calibration slightly if necessary.

Quote:
Originally Posted by GeorgeAB
That's what the demonstrations above illustrate. If you think you can accurately compensate for distorted color perception by some intuitive re-adjustment of the display settings, you are mistaken.

Correct, I don't want to compensate for my less than perfect viewing environment "intuitively". If you can invent or direct me to a device that would do the same thing more scientifically I'll certainly use it, but given that it doesn't exist I have no choice but do it this way. In theory if my room is painted a bright green (yuck) I may need to turn down the red (or is it up?). Regardless I need a device that would measure the way my eyes see the entire room's colors and tweak accordingly. Since I don't have such a device I HAVE to do it "intuitively". What option do I have? I don't want a "perfect" image that looks wrong. Obviously I'll start with perfect calibration and ONLY tweak it if something is bothering me.

We do this with subwoofers all the time, don't we? People calibrate their speaker's volumes and then go back and turn the sub up or down to taste - especially whenever they aren't listening anywhere near reference levels. Listening to the sound turned way down is somewhat analagous to watching with the lights on isn't it? Neither one is perfect, yet both scenarios may cause you to (temporarily) tweak your system off of perfect calibration. If the room with the lights on has a color on the walls then you might want to tweak the color, too (at least until nightfall or until you can paint everything black).

(BTW I noticed that many of the illusions involve fooling the eye with "context". They draw pictures of 3-D objects with light sources and shadows to fool the eye. For example: a grey object that appears to be a floor tile in a shadow might appear white. While the same color grey depicted as a floor tile in full light might appear black. It's the context of a "floor tile" that is fooling you - which isn't really a fair illusion to use when you're talking about room conditions in a home theatre. Objects on my TV are never seen as being "in" my theatre - for example, my TV might be in a shadow, but the person on the screen won't be since the room that actor is standing in has it's own light sources and shadows.)
Lurking at HTF Since 2001

#13 of 35 OFFLINE   GeorgeAB

GeorgeAB

    Second Unit

  • 468 posts
  • Join Date: Jan 28 2001
  • Real Name:G. Alan Brown
  • LocationDenver, CO

Posted May 25 2007 - 01:34 AM

"You would say to then fix my viewing enviroment. I would say again that I can't."

We all do the best we can with what limitations may be imposed upon us. If we can't meet the standards, at least by knowing the standards we have a target to use as a reference. Get as close as you can and look forward to the day when you can achieve better. There are few consumer displays that even provide correct color primaries. Color accuracy seems to be on most manufacturer's short list of priorities.

"I have no choice but do it this way."

Understood. However, without a neutral/objective reference our adaptive vision cannot be relied on for precision. All you're left with is 'looks good enough to me.' That position substitutes subjective satisfaction for image fidelity. Both can be worthy goals but should not be confused with one another.

"...or until you can paint everything black"

Not necessary, or even desirable, in most cases.

"Objects on my TV are never seen as being "in" my theatre - for example, my TV might be in a shadow, but the person on the screen won't be since the room that actor is standing in has it's own light sources and shadows.)"

Are you suggesting the viewing environment has somehow ceased to exist?

#14 of 35 OFFLINE   Brian D H

Brian D H

    Second Unit

  • 453 posts
  • Join Date: Sep 02 2004

Posted May 25 2007 - 05:32 AM

Quote:
Originally Posted by GeorgeAB
Are you suggesting the viewing environment has somehow ceased to exist?

Not at all. I'm simply saying that while these illusions do a good job of illustrating how our color perception can be fooled, they exagerate the point. These illusions use various tricks to fool the eye into thinking that grey is black/dark blue is light blue/orange is grey, etc. But, some of the tricks they use would never exist in the Home Theater environment, and the tricks they use that can exist may not exist to this extent.

In other words, your viewing environment WILL impact how colors and brightness are perceived, but not anywhere nearly as bad as these illusions.
Lurking at HTF Since 2001

#15 of 35 OFFLINE   GeorgeAB

GeorgeAB

    Second Unit

  • 468 posts
  • Join Date: Jan 28 2001
  • Real Name:G. Alan Brown
  • LocationDenver, CO

Posted May 25 2007 - 06:00 AM

"In other words, your viewing environment WILL impact how colors and brightness are perceived, but not anywhere nearly as bad as these illusions."

So you assert. So sure are you? That's a pretty broad statement. Did you design the illusions? How do you know these things?

I've seen some severely compromised viewing environments that were featured in home theater consumer and trade publications. I've also been in some that resulted in obvious and dramatic consequences to picture quality and image perception. I've also seen VERY subtle changes in surrounding colors altering the audience's perception of a picture. I'll simply have to disagree with your assertion until you can be more persuasive.

#16 of 35 OFFLINE   Kevin C Brown

Kevin C Brown

    Producer

  • 5,713 posts
  • Join Date: Aug 03 2000

Posted May 25 2007 - 12:23 PM

This *is* an interesting discussion. I can understand both sides in that there is a way to perfectly calibrate a display to what the director intended, but that the room can change the perception of that calibration such such that tweaking that calibration within someone's home might actually appear as an improvement to the viewer. Might not be "accurate", but it would still "appear" to be better!
If it's not worth waiting until the last minute to do, then it's not worth doing.

KevinVision 7.1 ...

#17 of 35 OFFLINE   GeorgeAB

GeorgeAB

    Second Unit

  • 468 posts
  • Join Date: Jan 28 2001
  • Real Name:G. Alan Brown
  • LocationDenver, CO

Posted May 26 2007 - 08:11 AM

"...there is a way to perfectly calibrate a display to what the director intended..."

That statement is a bit convoluted. Display calibration has nothing to do with anyone's "intention." An electronic display is simply an instrument, a tool. The TV should be aligned to imaging industry standards in order for it to faithfully and accurately convey the signal fed to it without changing that signal.

Picture controls can help compensate for viewing environment conditions, and the occasional poor broadcast signal, but only up to a point. Some controls can be over or under adjusted to the point where operational linearity suffers. One example is: no picture control or service menu setting can remove screen reflections from the image. The reflections (or diffused veiling glare on anti-reflection type screens) will always be present, interfering with the picture. It's much better to eliminate the cause of the reflections.

"Might not be "accurate", but it would still "appear" to be better!"

That's what consumers do every day. They attempt to adjust the picture to look "better" to them. It's called a WAG ( wild-ass guess) and strictly based upon their subjective judgement about what "better" is.

The points I have been making are all in light of the pursuit of image fidelity via display accuracy. Personal preference is another matter entirely. To understand calibration correctly, potential customers have to keep those two issues separate. If you pay for a calibrator to align your TV according to imaging industry standards and you don't particularly "like" the results, the technician did his job and earned his fee. He can attempt to alter an accurate display according to the customer's preference, but that's another thing entirely. That personal preference can be affected by: time of day, mood, lighting, decor, the program being viewed, etc., etc.

#18 of 35 OFFLINE   GeorgeAB

GeorgeAB

    Second Unit

  • 468 posts
  • Join Date: Jan 28 2001
  • Real Name:G. Alan Brown
  • LocationDenver, CO

Posted May 30 2007 - 04:00 AM

The National Institute for Standards and Technology (NIST) has labs near me in Boulder, CO. They have assigned one of their physicists to evaluate methods to help professionals and consumers select the right flat panel display for a given application. One very big factor that was considered is the impact the display's environment has upon the picture. NIST provides a nifty tutorial on their web site to illustrate how different flat panel display types handle various viewing environment challenges. Here's the link to the tutorial: http://www.fpdl.nist.gov/tips.html . It features a PDF slide show that you can save in a file. Many of the images are very good examples of the kinds of interferance to picture quality that poor viewing conditions can present.

Here's a link to a recent article in the Rocky Mountain News featuring an interview with the man who runs this department at NIST's Boulder labs: http://www.rockymoun....558486,00.html . This newspaper article will only be available for a limited time.

#19 of 35 OFFLINE   Brian D H

Brian D H

    Second Unit

  • 453 posts
  • Join Date: Sep 02 2004

Posted May 30 2007 - 04:33 AM

George,

Some points:

Quote:
Originally Posted by GeorgeAB
"...there is a way to perfectly calibrate a display to what the director intended..."

That statement is a bit convoluted. Display calibration has nothing to do with anyone's "intention." An electronic display is simply an instrument, a tool. The TV should be aligned to imaging industry standards in order for it to faithfully and accurately convey the signal fed to it without changing that signal.

Agreed. Correct calibration does AND SHOULD simply correctly display the signal sent to it. But, this statement is based on current technology. In the future I would hope that better measuring instruments allow a technician to calibrate the display for the environment it exists in.

Quote:
Originally Posted by GeorgeAB
Picture controls can help compensate for viewing environment conditions, and the occasional poor broadcast signal, but only up to a point....One example is: no picture control or service menu setting can remove screen reflections from the image. The reflections (or diffused veiling glare on anti-reflection type screens) will always be present, interfering with the picture. It's much better to eliminate the cause of the reflections.

Again, I agree. Screen reflections cannot be compensated for and must be eliminated.

Quote:
Originally Posted by GeorgeAB
"Might not be "accurate", but it would still "appear" to be better!"

That's what consumers do every day. They attempt to adjust the picture to look "better" to them. It's called a WAG ( wild-ass guess) and strictly based upon their subjective judgement about what "better" is.

Whoa! (Here's where I disagree) "Appear" to be better IS better. The entire reason we have our sets calibrated is so that the image appears to be accurate. Heck, (not to get too philosophical, but) vision itself is only a matter of perception and while the processing of an image inside the human brain may not be able to be compensated for, my room color can be. For the record, I do NOT want to do this by WAG, but do I have a choice? What I want is a calibration tool that can "see" that I have puke-green walls and flourescent lights, or blue walls and candles, or an ugly yellow rug right in front of the display. Why can't there be a calibration tool that is set up right where I sit that looks at my entire field of view and then tells the calibration expert to adjust the colors slightly to compensate? Until this tool is invented I have to make a WAG.

Quote:
Originally Posted by GeorgeAB
The points I have been making are all in light of the pursuit of image fidelity via display accuracy. Personal preference is another matter entirely. To understand calibration correctly, potential customers have to keep those two issues separate. If you pay for a calibrator to align your TV according to imaging industry standards and you don't particularly "like" the results, the technician did his job and earned his fee. He can attempt to alter an accurate display according to the customer's preference, but that's another thing entirely. That personal preference can be affected by: time of day, mood, lighting, decor, the program being viewed, etc., etc.

Again, room decor is NOT customer's preference (Well, it was certainly someone's preference, but not as far as the display image is concerned Posted Image). Do I care if my TV is displaying an image "perfectly" if I can't see it? Time of day is too much of a variable to attempt to compensate for, but my room color is not. While the technician certainly has "earned his fee" when he calibrates correctly it is simply because the device I describe below has not been invented. When it has I would expect him to use it to earn his fee. (Unless my room is a perfectly neutral area with dim 6500K lighting)

After all, what is so difficult about this device?
1) You get a camera set up in the viewing location to look in the direction of the display with the appropriate len(s) to mimic the human field of view.
2) It has software built in that allows the technician to set the area of the display in this field of view (this area would be disregarded for analysis).
3) The area it "sees" (not counting the display) would be analyzed for color/brightness (with more "weight" given to colors closer to the display) and an "average" room color/brightness would be computed.
4) This ambient room color/brightness value could then be used to modify the calibrated settings.
This is not a WAG.

Heck, I can walk into any paint store and have a cushion analyzed for it's color and buy a gallon of paint to match. Why can't I have the same thing done in my room as a whole?
Lurking at HTF Since 2001

#20 of 35 OFFLINE   GeorgeAB

GeorgeAB

    Second Unit

  • 468 posts
  • Join Date: Jan 28 2001
  • Real Name:G. Alan Brown
  • LocationDenver, CO

Posted June 02 2007 - 07:12 PM

Brian,

Wishful thinking won't make it happen or make it affordable. In the context of current technology, the best way to preserve display accuracy and image fidelity is to provide a neutral surround and correct lighting for a calibrated display.

There's no need to feel bad if viewing environment conditions have to be compromised due to circumstances beyond your control. The important thing is to understand what imaging science principles are involved and do the best you can in the circumstances you face. Without an understanding of the fundamental principles involved, image fidelity is unobtainable in any reliable way. Fortunately, there are methods and solutions available for the individual who has the option to do it correctly.

Best regards and beautiful pictures,
G. Alan Brown, President
CinemaQuest, Inc.

"Advancing the art and science of electronic imaging"


Back to Display Devices (TVs/Projectors/Screens)



Forum Nav Content I Follow