or Connect
AVS › AVS Forum › Display Devices › Display Calibration › Adjusting preset luminane for plasma
New Posts  All Forums:Forum Nav:

Adjusting preset luminane for plasma

post #1 of 15
Thread Starter 
I'm confused as to why I would first adjust white point via a test pattern then turn around and adjust luminance via a meter using the same contrast control. Doesn't this have the potential to lower my dynamic range/ tonal range?
post #2 of 15
unlike LED/LCD displays, plasmas have no backlight and therefore there is no backlight control... therefore, the only way to control light output is with the contrast control (there might still be settings like panel brightness on Panasonics (low/med/high) and cell light on Samsungs (20-pt but should be left at max since it's really just a placeholder so that Samsung can use the same menu system for LED/LCD and plasma)

lowering contrast from the max setting that doesn't clip or discolor is desirable when your viewing environment and preferences don't require max light output and it would just cause eye fatigue... your black levels should remain largely unchanged when you move the contrast control up and down and so you shouldn't have to worry about the image losing depth/contrast just because the white level is a bit lower than max... the current crop of plasmas (especially the Panasonics) have plenty of contrast ratio to begin with (10,000:1 and higher ON/OFF CR) and so you don't need to worry about losing a little
post #3 of 15
In spite of what directions say on test/setup discs (Blu-ray or DVD), you CANNOT EVER find the right setting for Contrast using a pattern and eyes with no meter. All you can do is find the highest setting you can use that doesn't remove white levels above 235/100%. The highest setting is not the RIGHT setting. For a dark room, the RIGHT setting for 100% white is around 35 fL +/- 5 fL. You can't determine that level by eye... just not possible. All you can do is be aware of eyestrain and if you notice any eyestrain after 1-3 hours of viewing your display in a dark room, it's probably too bright and you should reduce the Contrast setting until you get to a point where there's no eyestrain. It takes a good long while to detect eyestrain though... it's not something you can estimate in 5 seconds, though if you are squinting at a bright screen, chances are the screen is much too bright.

A common scenario is for the Contrast control to have settings from 0-100. The default might be 100 and you might have 60 fL or more (for an LCD TV) when the Contrast control is set that high...some TVs may even produce 150 fL when the Backlight control is turned up all the way. Plasma TVs rarely will get brighter than 60 fL which is still double the luminance you want in a dark room.

So you display a Contrast pattern and you learn that at a setting of 85, you can see all the white steps from 220-253, but if you go higher, you start losing steps at the high end of the range. That's not a seriously bad problem, few, if any people will be able to detect whether white steps stop at 235 or whether they remain visible all the way to 254. But if you can see steps about 235, you may as well setup so they are visible. But if you had a meter, you might find that the 85 setting for Contrast produces 52 fL which is still too bright. Your meter (if you had one) might tell you that a setting of 65 produces 35 fL for 100% white in a dark room which is about where you want to be. So even though the Contrast test pattern told you 85 is the highest setting you can use without losing steps above 235, that doesn't mean 85 is the right setting. This is a hypothetical example, not based on any particular TV.

Some TVs never lose steps about 235 no matter how high you set Contrast. With those TVs, eyestrain is your only reference for finding the best setting if you do not have a meter.

And there are some TVs that ALWAYS clip steps above 235 no matter what Contrast setting you use, so once again, eyestrain is all you have to use to find the right setting if you don't have a meter.
post #4 of 15
Eye strain is considered a likelihood for most viewers with screen brightness set as low as 30 fL in a dark room. That is why multiple standards bodies still recommend implementing correct bias lighting for extended viewing sessions in dark conditions, even at such relatively low luminance levels. What Joe Kane said decades ago is still a problem among those who design, install, and calibrate video systems. Viewing environment conditions and human factors are the most frequently ignored aspect of proper display setup.

Part of this problem is evidenced by the very title of this sub-section of the forum. 'Display Calibration' should never be divorced from the rest of the system if image fidelity is the goal. Having a precise display device, with impressive calibration reports, is absolutely no guarantee of image accuracy in a system with potentially interfering elements. A correctly aligned display is only one component in the system. This area of the forum would be better titled, 'Display System Calibration.' That would serve as a good reminder for calibrators not to get tunnel vision and focus only on the characteristics of the display device. After all, human vision is the most basic foundation of the entire video industry and medium.

Best regards and beautiful pictures,
G. Alan Brown, President
CinemaQuest, Inc.
A Lion AV Consultants affiliate

"Advancing the art and science of electronic imaging"
post #5 of 15
Quote:
Originally Posted by Doug Blackburn View Post

In spite of what directions say on test/setup discs (Blu-ray or DVD), you CANNOT EVER find the right setting for Contrast using a pattern and eyes with no meter. All you can do is find the highest setting you can use that doesn't remove white levels above 235/100%. The highest setting is not the RIGHT setting. For a dark room, the RIGHT setting for 100% white is around 35 fL +/- 5 fL. You can't determine that level by eye... just not possible. All you can do is be aware of eyestrain and if you notice any eyestrain after 1-3 hours of viewing your display in a dark room, it's probably too bright and you should reduce the Contrast setting until you get to a point where there's no eyestrain. It takes a good long while to detect eyestrain though... it's not something you can estimate in 5 seconds, though if you are squinting at a bright screen, chances are the screen is much too bright.

A common scenario is for the Contrast control to have settings from 0-100. The default might be 100 and you might have 60 fL or more (for an LCD TV) when the Contrast control is set that high...some TVs may even produce 150 fL when the Backlight control is turned up all the way. Plasma TVs rarely will get brighter than 60 fL which is still double the luminance you want in a dark room.

So you display a Contrast pattern and you learn that at a setting of 85, you can see all the white steps from 220-253, but if you go higher, you start losing steps at the high end of the range. That's not a seriously bad problem, few, if any people will be able to detect whether white steps stop at 235 or whether they remain visible all the way to 254. But if you can see steps about 235, you may as well setup so they are visible. But if you had a meter, you might find that the 85 setting for Contrast produces 52 fL which is still too bright. Your meter (if you had one) might tell you that a setting of 65 produces 35 fL for 100% white in a dark room which is about where you want to be. So even though the Contrast test pattern told you 85 is the highest setting you can use without losing steps above 235, that doesn't mean 85 is the right setting. This is a hypothetical example, not based on any particular TV.

Some TVs never lose steps about 235 no matter how high you set Contrast. With those TVs, eyestrain is your only reference for finding the best setting if you do not have a meter.

And there are some TVs that ALWAYS clip steps above 235 no matter what Contrast setting you use, so once again, eyestrain is all you have to use to find the right setting if you don't have a meter.

As I've mentioned many times before, you can use a contrast pattern to determine the optimal setting on a LCD, since the highest setting is the one you want on a LCD and then you can use a meter to set light output with the backlight control.

(The very rare exception being the LCD puts too much light out even at a backlight setting of zero, which is something I've never encountered myself across various LCD TVs and monitors.)

Furthermore, targeting a value like 35 fL might still not be ideal for given viewer/viewing environment (as in there is no magic number that works for everyone... only reasonable recommendations/ranges based on viewing conditions)
Edited by PlasmaPZ80U - 8/3/13 at 6:17pm
post #6 of 15
Also, when sitting close to large screen you'll want less light output and when sitting far from a small screen you'll want more light output.

more details about light output and setting contrast covered here:

http://www.tlvexp.ca/2012/01/setting-contrast-not-what-you-might-think/
post #7 of 15
Thread Starter 
Quote:
Originally Posted by Doug Blackburn View Post

In spite of what directions say on test/setup discs (Blu-ray or DVD), you CANNOT EVER find the right setting for Contrast using a pattern and eyes with no meter. All you can do is find the highest setting you can use that doesn't remove white levels above 235/100%. The highest setting is not the RIGHT setting. For a dark room, the RIGHT setting for 100% white is around 35 fL +/- 5 fL. You can't determine that level by eye... just not possible. All you can do is be aware of eyestrain and if you notice any eyestrain after 1-3 hours of viewing your display in a dark room, it's probably too bright and you should reduce the Contrast setting until you get to a point where there's no eyestrain. It takes a good long while to detect eyestrain though... it's not something you can estimate in 5 seconds, though if you are squinting at a bright screen, chances are the screen is much too bright.

A common scenario is for the Contrast control to have settings from 0-100. The default might be 100 and you might have 60 fL or more (for an LCD TV) when the Contrast control is set that high...some TVs may even produce 150 fL when the Backlight control is turned up all the way. Plasma TVs rarely will get brighter than 60 fL which is still double the luminance you want in a dark room.

So you display a Contrast pattern and you learn that at a setting of 85, you can see all the white steps from 220-253, but if you go higher, you start losing steps at the high end of the range. That's not a seriously bad problem, few, if any people will be able to detect whether white steps stop at 235 or whether they remain visible all the way to 254. But if you can see steps about 235, you may as well setup so they are visible. But if you had a meter, you might find that the 85 setting for Contrast produces 52 fL which is still too bright. Your meter (if you had one) might tell you that a setting of 65 produces 35 fL for 100% white in a dark room which is about where you want to be. So even though the Contrast test pattern told you 85 is the highest setting you can use without losing steps above 235, that doesn't mean 85 is the right setting. This is a hypothetical example, not based on any particular TV.

Some TVs never lose steps about 235 no matter how high you set Contrast. With those TVs, eyestrain is your only reference for finding the best setting if you do not have a meter.

And there are some TVs that ALWAYS clip steps above 235 no matter what Contrast setting you use, so once again, eyestrain is all you have to use to find the right setting if you don't have a meter.

Great post Doug, I think this confirms my concern. So in a nutshell proper luminance settings for eyestrain issues may not equal optimum white level settings for best tonality. You don't seem to think the trade off is that significant so I'll go with your recommendation to adjust for the eyes. Later I will post images with accompanying curves graphs illustrating the effect of raising or lowering the white point to achieve a particular luminance setting for reference.
post #8 of 15
Quote:
Originally Posted by PlasmaPZ80U View Post

As I've mentioned many times before, you can use a contrast pattern to determine the optimal setting on a LCD, since the highest setting is the one you want on a LCD and then you can use a meter to set light output with the backlight control.

(The very rare exception being the LCD puts too much light out even at a backlight setting of zero, which is something I've never encountered myself across various LCD TVs and monitors.)

Furthermore, targeting a value like 35 fL might still not be ideal for given viewer/viewing environment (as in there is no magic number that works for everyone... only reasonable recommendations/ranges based on viewing conditions)

READ THE TOPIC... it says PLASMA, not LCD.

And even if the topic WAS LCD, you STILL cannot determine the optimal Contrast setting ONLY with a test/setup disc test pattern. There are simply too many variables, including non-linearity of the backlight at low settings. You will never know that is happening without measurements unless it is a very severe problem that is very obvious. You may find that a backlight with a setting range of 0-20 is only usable within a range of 5-15 because of non-linearity issues, but that would escape you entirely if you didn't have a meter. The LCD issue is much more complex than you make it sound because you made assumptions that are not good/realistic (i.e. that the backlight is usable/fine at any setting throughout its adjustment range.

I did give a reasonable recommendation. And I specified dark room viewing to further characterize the recommendation.
post #9 of 15
Quote:
Originally Posted by Doug Blackburn View Post

you STILL cannot determine the optimal Contrast setting ONLY with a test/setup disc test pattern. There are simply too many variables, including non-linearity of the backlight at low settings. You will never know that is happening without measurements unless it is a very severe problem that is very obvious. You may find that a backlight with a setting range of 0-20 is only usable within a range of 5-15 because of non-linearity issues, but that would escape you entirely if you didn't have a meter. The LCD issue is much more complex than you make it sound because you made assumptions that are not good/realistic (i.e. that the backlight is usable/fine at any setting throughout its adjustment range.

What other variables are there, aside from the one you mentioned? I'd like to know if I'm missing something here.

Also, let's assume the backlight is non-linear at settings of 0-4 out of a max of 20. Would the average LCD even be able to produce 35 fL at a backlight setting of 0% to 20%? I know I have never owned an LCD (mainly entry-level and mid-range TVs and monitors) that can produce even 30 fL at such extremely low backlight settings (20% or less of max). Perhaps high-end LCDs have brighter backlights (more LEDs)?
post #10 of 15
Also, do you have any examples of LCDs with non-linear backlights? brand/model?

And by non-linear backlight are you referring to a grayscale/gamma tracking issue or something else?
Edited by PlasmaPZ80U - 8/5/13 at 9:29am
post #11 of 15
Quote:
Originally Posted by Steve Ruddy View Post

Great post Doug, I think this confirms my concern. So in a nutshell proper luminance settings for eyestrain issues may not equal optimum white level settings for best tonality. You don't seem to think the trade off is that significant so I'll go with your recommendation to adjust for the eyes. Later I will post images with accompanying curves graphs illustrating the effect of raising or lowering the white point to achieve a particular luminance setting for reference.

Getting the Contrast setting correct, within that 35 fL window, does not compromise ANYTHING. It's just where you want the luminance to be for viewing content in a dark room without eyestrain or perhaps with minimal eyestrain is a better definition. To truly eliminate eyestrain, you want a backlighting system with correct d65 light surrounding the area behind the TV to make the illuminated area larger. And you want that light to be about 1/3 as bright as the TV screen (plus or minus a bit).

If you have instrumentation, you simply calibrate at the correct 35 fL level. During calibration, that measured level may change a bit. You can either reset it and re-cal a bit so you maintain 35 fL or if it measures 32 or 37 fL after calibration... that's not likely to be critical for eyestrain.

Eyestrain comes from having a black area surrounding the screen with the screen considerably brighter. Your eye/brain has a hard time deciding whether to close the iris for the bright screen or open the iris for the dark background. There is a constant "fight" going on and that leads to eyestrain... which is why backlighting can be so effective. The other thing that can be effective is sitting so close that the screen fills almost all of what you see when you look at the screen... there's then no black area around the edges to throw your brain off. But that's really too close for HDTV resolution.
post #12 of 15
Non-linearities in the backlight are pretty difficult to avoid. White CCFL, white LEDs, and RGB LEDs can all be non-linear, not just in brightness but in color. White light sources can be quite different in spectral distribution of frequencies of light present as you dim and brighten them. And they may not be linear re. luminance... meaning if the setting range is 0-20, that a setting of 10 probably doesn't produce half the light output of the 20 setting, at least not precisely half. And a color LED can shift spectral distribution just as easily. A Red LED might drift towards orange or crimson as it is dimmed or made brighter so you can't assume that dimming a backlight to minimum is the best backlight setting even if the LCD display will produce 30-40 fL at the dimmest setting.

You also need to determine how the LCD pixels interact with the backlight. If the LCD pixels don't produce good black levels, you have to figure out if you can dim the backlight more than you would prefer, but still be able to achieve a good calibration. Not all LCD pixels are created equal. The ones that block more light when they turn black will produce better-looking images, but those tend to exist in more expensive TVs.
post #13 of 15
Quote:
Originally Posted by Doug Blackburn View Post

Non-linearities in the backlight are pretty difficult to avoid. White CCFL, white LEDs, and RGB LEDs can all be non-linear, not just in brightness but in color. White light sources can be quite different in spectral distribution of frequencies of light present as you dim and brighten them. And they may not be linear re. luminance... meaning if the setting range is 0-20, that a setting of 10 probably doesn't produce half the light output of the 20 setting, at least not precisely half. And a color LED can shift spectral distribution just as easily. A Red LED might drift towards orange or crimson as it is dimmed or made brighter so you can't assume that dimming a backlight to minimum is the best backlight setting even if the LCD display will produce 30-40 fL at the dimmest setting.

You also need to determine how the LCD pixels interact with the backlight. If the LCD pixels don't produce good black levels, you have to figure out if you can dim the backlight more than you would prefer, but still be able to achieve a good calibration. Not all LCD pixels are created equal. The ones that block more light when they turn black will produce better-looking images, but those tend to exist in more expensive TVs.

is it necessarily a problem if the backlight control is not perfectly linear with regards to luminance (50% backlight doesn't produce exactly 50% of max light output)? as long as you are using a meter to measure light output, couldn't you just use whatever backlight setting produces the target level of light output (assuming you've already set contrast with a white clipping pattern)?

and for color non-linearities, couldn't you just compensate for any spectral distribution changes with the available calibration controls?

I guess what I'm asking is whether there are backlight settings that must be avoided completely for the most accurate picture or whether you just need to compensate for any non-linearities using the available picture controls.
post #14 of 15
You might be able to correct the different spectral distribution... depends on what happens. If the measured Red point is moved inside the Rec 709 triangle (undersaturated) you can't fix that. If it is over saturated or too bright or too dim, you MAY be able to fix that, but the calibration controls may not do what you want them to do at low backlight settings.

It's only a problem if the control itself is non-linear if you are expecting it to operate linear (without a meter). If you have a meter and measure what is happening, you surely can get the luminance within that 30-40 fL range... as long as the backlight is bright enough.

It all depends on how the backlight behaves and how you EXPECT it to behave. If you want a dark backlight setting for movies in a dark room and a bright backlight setting for daytime viewing in a family room, you may need 2 separate calibrations... if that is even possible. Sometimes you can calibrate 2 different modes and use one for dark room and the other for bright room and have different calibrations for each one. The problem comes from ASSUMING if you calibrate 1 mode, you can just make the Backlight brighter or darker without affecting the calibration... that might be possible over some limited portion of the backlight setting range, but could be problematic at the darker or brighter setting because spectral distribution changes so much.
post #15 of 15
Quote:
Originally Posted by Doug Blackburn View Post

You might be able to correct the different spectral distribution... depends on what happens. If the measured Red point is moved inside the Rec 709 triangle (undersaturated) you can't fix that. If it is over saturated or too bright or too dim, you MAY be able to fix that, but the calibration controls may not do what you want them to do at low backlight settings.

It's only a problem if the control itself is non-linear if you are expecting it to operate linear (without a meter). If you have a meter and measure what is happening, you surely can get the luminance within that 30-40 fL range... as long as the backlight is bright enough.

It all depends on how the backlight behaves and how you EXPECT it to behave. If you want a dark backlight setting for movies in a dark room and a bright backlight setting for daytime viewing in a family room, you may need 2 separate calibrations... if that is even possible. Sometimes you can calibrate 2 different modes and use one for dark room and the other for bright room and have different calibrations for each one. The problem comes from ASSUMING if you calibrate 1 mode, you can just make the Backlight brighter or darker without affecting the calibration... that might be possible over some limited portion of the backlight setting range, but could be problematic at the darker or brighter setting because spectral distribution changes so much.

so you're basically saying it's best to confirm the impact (if any) the backlight setting has on various areas of calibration like grayscale, gamut, (and obviously light output numbers if you are targeting a specific number or range of numbers) with a meter/software and then you can decide whether to increase or lower the backlight a bit on the fly depending on viewing conditions, source material, and eye fatigue (avoiding any backlight setting changes that hurt the calibration in a significant way/require a re-cal)...
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Display Calibration
AVS › AVS Forum › Display Devices › Display Calibration › Adjusting preset luminane for plasma