Why are the manufacture stated contrast ratio specs worthless?

I wouldn't quite say "worthless" but misleading for sure. There is no industry standard for measuring contrast ratios...so all companies use their own set of tricks to arrive at some absurd numbers for the marketing department to use. The high numbers you see are often a theoretical peak that the normal user will never reach at any time....and often some form of dynamic contrast circuitry is used...and most videophiles shun such processing and prefer to disable it if possible.


When I say the numbers aren't totally worthless I mean that if manufacturers are claiming a higher number than the competition...their set may in fact have a higher ratio during normal use. (but nowhere near the inflated number listed) When shopping for a tv it is best to use your own eyes and listen to owners of the sets you're considering. Shopping by numbers on a spec sheet can lead you down the wrong path..

The crazy numbers like 30,000:1 are useful when comparing sets by the same manufacturer. The procedure I would follow to compare any two sets on the cheap - without any meters:


1. Make sure the ambient light resembles the level you will have at home at night.

2. Select STANDARD mode.

3. Adjust the backlight to your liking.

4. Select a 4:3 channel. This will produce two black side bars.

5. Compare the side bars with the frame. If you can see the difference, don't buy that set.


If you like the set with less than amazing blacks, consider buying a backlight - a source of white light (6500) you place behind the set. Your eyes will adjust to the overall brightness and will perceive the less than perfect black as blacker. Additional bonus: less eye strain.

With all due respect to anyone trying to defend "dynamic" contrast ratios, sadly they are useless. First, let's look at what a contrast ratio is in the first place. The basic ratio is White divided by Black. In real "optic's world" ratio, the formula is White+reflection / Black + reflection. Reflection is a percentage of the ambient light (room lighting) that is getting relfected off the front of a screen. Relfected light gets added to both the White reading as well as the Black reading.


From the LCD manufacturing standpoint (and not necessarily LCD TV set manufacturers) since they cannot define ambient light, they test their displays in a completely dark room. This means that there is ZERO reflection, so the contrast ratio is "pure". Assuming a 500 nit bright LCD, and Black is, for instance .5 nits; this would calculate to a 1000:1 "intrinsic" ratio.


Not wanting to lose to the incredibly high contrast ratios of plasma, LCD TV set manufacturers started to play a numbers game. Plasma emit "essentially" ZERO light as black so they get their 500 nits of White divide by ZERO. How many ZEROs goes into 500; as many as you like; say 500,000:1. The LCD guys started playing around with dynamic numbers. They take the highest setting for White and divide it by the lowest possible output on Black. Now the 500 nits is divided by .005. It never matters to anyone that you cannot possibly get the highest white on the same screen as the lowest Black. This is the main reason why dynamic contrast ratios are worthless. (Note, with LED backlights, there are ways to drive the LED matrix with a similar pattern to the video, so that maximum white can be seen on the same image as the minimum black.... still its kinda new and expensive).


The other main reason why these dynamic contrast numbers are meaningless is that there are VERY few people watching their flat panels in a 100% dark room. As soon as there is light; there is reflection. It is quite possible to have as high as 100 nits of "black" given the right conditions. 500 nits of White / 100 nits of black is 5:1 and in some places, this is still considered acceptable.


Do the setup that is decribed above. When you are all done, turn off the set and look at it in whatever lighting is "typical". Is it at night with a lamp near you in front of the set? Is it daytime with the blinds to the window open? Whatever you see on the screen, with the backlight completely off is the darkest the screen will ever be in "Black". This is the "evil" of reflected light, and why no one ever tries to address it.


Lastly, if the screen is 500 nits, just how dark does black have to get to be "good looking. You can't make the screen any brighter, so you are stuck in making black, blacker. How black can you go or need to go (assuming no relfection).

DBLASS, Yours is the best post ever on this subject. Hopefully, it will be reduce the number of my-set-has-better-blacks-than-yours type of debates. As someone wrote, just one candle takes that zillion to one ratio down to about fifty to one or so. Further, the human eye and brain has a natural contrast limit of about 1000:1 or 60dB. The bottom line: paying a premium for that super contrast alone is as wasteful as spending big bucks on a car that can go 200 miles per hour or a 1000W per channel AV receiver.


The challange plasma manufacturers have is how to minimize the "black" cell drive as just turning it completely off is, or was, impossible. But I admit that my knowledge of this subject is not up-to-date. The ANSI checker flag test makes little sense if the plasma blacks, excluding the ambient light, are truly perfect as things become unmeasurable. However, the ANSI test would be useful with the non-LED LCDs because of the light leakage that is still measurable in the LCD sets. To Sony's credit, they do, or they used to, specify the "same screen" contrast.



Quote:
Originally Posted by xrox /forum/post/15409855


I always wonder how much ANSI contrast is affected by ambient light whose source is the screen itself. Having multiple white squares on the screen at peak white will create considerable ambient light in itself.

Click to expand...

I am not sure how the test is done but it seems logical that during the test all the ambient lights are off and the meter should be in contact with the screen. This would eliminate any errors that might otherwise be caused by the light coming from the white squares after bouncing from the walls and other objects including the person who is performing the test.


The explanation from Wikipedia does not offer much detail either by stating what we already know:

Another measure is the ANSI contrast, in which the measurement is done with a checker board patterned test image where the luminosity values are measured simultaneously. This is a more realistic measure of system capability, but includes the potential of including the effects of the room into the measurement, if the test is not performed in a room that is close to ideal.


Another bit from Wikipedia offers this point of reference: A clean print at a typical movie theater may have a contrast ratio of 500:1. Yet another reason why chasing the super contrast is hard to justify.

Everyone is trying to come up with a digital answer to an analog question. Everyone wants the "A" is number one or "B" sucks, here's why. The answers are "maybe, kinda and sorta". The ANSI standards commitee has long tried to take the fluff and ambiguity out of the equations, in their attempt to make a solid digital answer.


By doing this, their tests, as is all Mil rated tests for displays, defines the EXACT environment to take the test. Again, they are all in absolute dark rooms. Also, they define the light source, the distance away it is, the angle to the screen. They even define the light meter.


Did you know that even light meters vary their scores. Inside of meter are pre-set calculations and algorithms that define the how the meter comes up with the nit rating.


The truth (if there is such a thing) is to find out the make and model of the actual display being used in the set. There are about 6-8 OEM LCD display manufacturers with Samsung, LG, Sharp and AUO leading the pack. Read the product specification for the OEM model. The model that is rated with the highest contrast, response speed, etc. in that document will be better in any given environment than a display with a lower score. Getting the product specification on the OEM cell is easy. Finding out which display is inside the TV is a different challenge.



Quote:
Originally Posted by Ktulu6569 /forum/post/15411846


You must be a mathematician. The last time I heard such terms used was when I was watching the science channel.

Click to expand...

Definitly.


but I thought malleable meant "ability to be pounded into thin sheets".

Manufacturer numbers tend to go as close to outright lies (178 degree viewing angles anyone?) as they can get away with. Pumping the backlight down to some unusable under any circumstance level to measure the black and to maximum to measure white.


That is why we could really use some good reviews that measure Ansi contrast or at least their own standard checkerboard measurement.


Where are the good reviews that do this? In fact where are the good reviews period. Most don't get beyond the level of advertising.


The ANSI test is the only one that really makes sense to me. It is a fair representation of what the screen can represent at any one time.


It also gives you a good indication of how good your blacks will look when displaying images on screen. Also in my book it is preferable to shut off the dynamic contrast behavior and not have the back light pumping from scene to scene.


Some may have a point, that if we ever get to near absolute black, contrast measurements become meaningless (infinite) but then we have the holy grail and we are a long way from that, especially with LCDs.



Quote:
but I thought malleable meant "ability to be pounded into thin sheets".

Click to expand...

mal·le·a·ble


1. Capable of being shaped or formed, as by hammering or pressure: a malleable metal.

2. Easily controlled or influenced; tractable.


3. Able to adjust to changing circumstances; adaptable: the malleable mind of the pragmatist.

-The American Heritage® Dictionary of the English Language


Synonymns: flexible, pliant, adjustable

Quote:
My question is... Where does one set that bar? I know how bright, and dark it will will if I view HD in my own LR. [ii] Where is the standard then? [iii] what does one look for?

Click to expand...

(i) Well, you figure in your viewing environment's lighting (black, dim, moderate, bright, brilliant)*, and get a monitor that will produce enough light to create convincing whites in that environment (brightness ratings are sufficient, here). Then, buy and test the monitor in a black environment to get a sense of the depth of black that it actually produces. If satisfactory for your purposes, keep it. If not, return it.


*Highly technical, I know. Sadly, there isn't a standardized way of doing this.


(ii) It doesn't exist. I believe that meaningless specs reinforce the current order in the market--if objective comparisons cannot be made, what does the rational agent do? She goes with a name brand, putting her faith in a corporate reputation and presumably low failure rates.


(iii) BLACK. Contrast is good to the extent that black is deep. I think that is roughly accurate because brightness is not much of an engineering obstacle. If your black is gray, you have a flat-looking, lifeless image. Shadow is shape--without convincing darks and good shadow detail, you experience none of the crucial three-dimensional feel. In today's flat panel market, this is the difference between good and not good. To take an obvious case--it is the difference between a Pioneer KURO and a Sharp 55U.

Quote:
You must be a mathematician.

Click to expand...

Well, actually I am a philosophy major with concentration in mathematical logic and philosophy of mathematics.



Quote:
Originally Posted by JeffMHT /forum/post/15448038


mal·le·a·ble



*Highly technical, I know. Sadly, there isn't a standardized way of doing this.


(ii) It doesn't exist. I believe that meaningless specs reinforce the current order in the market--if objective comparisons cannot be made, what does the rational agent do? She goes with a name brand, putting her faith in a corporate reputation and presumably low failure rates.


(iii) BLACK. Contrast is good to the extent that black is deep. I think that is roughly accurate because brightness is not much of an engineering obstacle. If your black is gray, you have a flat-looking, lifeless image. Shadow is shape--without convincing darks and good shadow detail, you experience none of the crucial three-dimensional feel. In today's flat panel market, this is the difference between good and not good. To take an obvious case--it is the difference between a Pioneer KURO and a Sharp 55U.

Click to expand...

I disagree. ANSI contrast is a standardized and simple measure that gives a visually significant number as opposed to the meaningless dynamic numbers.


The reason ANSI is not used has nothing to do with difficulty and everything to do with it not being as suitable for marketing (AKA lies).


I measured this on my own LCD computer monitor. It took me about 30 seconds. The measuring device (getrag eye one) goes right on the screen so is somewhat immune to ambient conditions.


LCD constrast tends to remain constant throughout the back light range as the ratio of panel blocking ability is a constant value. My calibrated NEC has about 700:1 contrast whether you are running the backlight at full or at zero. You then use the backlight to best match ambient conditions.