Gizmodo
Pioneer lined up its newest plasma display next to top-of-line TVs from Panasonic, Samsung, Sony and Sharp today. An interesting experiment, for sure. And it got everyone in the room talking about the same thing: contrast ratios. You can see why just by looking at the image above, which shows three plasma screens with power on but no image (the Pioneer is on the lower right, the Samsung HP-T5064 is on top and the Panasonic PZ700U is on the left). None of the screens were calibrated, which would make a difference. But the reason Pioneer's screen looks so much darker has to do with a lot more than calibration, or contrast ratio.
Consumer electronics companies love spec wars. Whether it's processor speeds, throughput or megapixels, gadget makers like throw around big numbers that separate money from wallet. And contrast ratios are the spec war du jour. But despite claims ranging from 5,000:1 (Panasonic) all the way to 1,000,000:1 (Sony's upcoming OLED) there is no agreed upon industry standard used for measuring contrast ratios. As a result, there are a number of tricky ways to influence the outcome of a contrast ratio test—and none of them have anything to do with the real world contrast ratio that you will experience while sitting at home...
What is contrast ratio? Simply put, it's the difference between the darkest and brightest spot on a display. This is expressed as a ratio measuring luminosity. A good way to think of contrast is like the volume on a stereo. You might have a stereo that goes all the way to 11, but that doesn't mean it sounds good. The environment has an effect too—the advertised contrast ratio has no bearing on how well the screen will perform in your house while reflecting light from a bank of windows on the other side of the room. Props to Sony for admitting as much, even if it is only in the fine print:
VESA test and measurement methods are applied yielding a contrast ratio of used 7000:1. This number represents the widest possible ratio between black and white contrast levels. Sony also measures their BRAVIA televisions with a more stringent method that measures the amount of black and white levels that can appear on the screen at the same time. This method yields a more real world measurement of 1300:1
LCD shoot-out: the Sony XBR is on top, Sharp Aquos lower left and the Pioneer plasma is on the right.
What contrast ratio is not: Contrast has little to do with color range or accuracy. Color is all in the gray-scale capabilities of a screen. But with better and deeper blacks some TVs are also better able to show the gradations of color as they fall away.
How is contrast ratio measured? In a very dark room, first of all. The big number on the front end (i.e. 20,000) is the light side, and the 1 is the dark side. The dark side, therefore, has a much greater effect on the ratio than the light. Cutting the darkest dark on a screen by .5 effectively doubles the contrast ratio. Which is why you hear a lot about "true black" and never about screens as bright as arc welding torches that you need shaded lenses to view.
How is contrast calculated? Contrast ratio is calculated in many different ways. The basic idea is to make one part of the screen dark and one part really bright and measure the difference using a light meter. Among the best known is the "full on/full off" method. There is also ANSI, which was developed as a way of measuring the contrast in projectors. Another commonly used method is the VESA (Video Electronics Standards Association) Flat Panel Display Measurement (FPDM) standard. And in Japan, the JEITIA (Japan Electronics and Information Technology Industries Association) rules the roost.
I was going to detail each one of these methods, but what's the point: they are all flawed and equally meaningless because no one runs the tests in exactly the same way. There is no ideal setting, for one thing, no room that is perfectly black. And just what black is being measured? You could turn the TV off entirely, which results in a screen that is darker than when the display is powered up. Some companies measure contrast without the filter on the front of the screen, which yields much brighter whites. Another option is to measure the black in a dark room, but then move the display to a well-lit room and turn it on to measure the light side. All of these examples are taken from interviews with people who work in the display industry and asked to remain anonymous.
Isn't someone developing a standard method for measurement? The Consumer Electronics Association is discussing the possibility, but nothing is moving forward. And it probably won't until there is enough hue and cry for a change to occur.
If there were a standard, what would it be? Instead of using one big number, the better way to talk about contrast ratio would be based on real world settings. TVs should specify contrast in terms of watching a broadcast football game during the day versus a high-def movie at night with all the lights off. More confusing, but also more accurate.
So what should I make of all this contrast ratio bragging going on? The best bet is to ignore it entirely. It is marketing hype. Contrast ratio is the least important number to look at when making a purchase. Judge with your own eyes—it's worth noting that your eyes recognize contrast far more clearly than resolution, which is why a 1080p screen with low contrast may not look as good as a 720p screen with high contrast. Enthusiast magazines such as Perfect Vision and Home Theater do their own contrast ratio tests on screens they review, which is a good option. But ultimately, those publications agree with me.
This topic always generates some hot commenting action. Fire away and I'll answer questions the best I can. But I've got a question for the readers as well: Is contrast ratio a useful, albeit flawed, way of comparing TVs and announcing new TV technologies, or should Gizmodo ignore that number from now on?
Comments
It would have been nice to list what the three screens claim as their contrast ratio's, just to get an idea of scale.
Personally, I think you should still list contrast ratio's even if they are a bit flawed. It sounds like it is sorta like the old vacuum brag about who had the most amps, thinking that meant they could clean better, but its still another factor to consider.
If the process was not flawed it would be more technically important but by itself it is very important for people to at least understand what it is and how it relates to displays. Those that buy high definition displays are already looking at the finer points of the display such as color saturation or sharpness and many understand that these settings can and should be optimized.
I have experienced many times where people have bought a display simply because it looked sharper or cleaner or had better color and when they got it home and setup they realized black areas always looked gray for some reason and they couldn't figure out why. Even if flawed contrast ratio's can give customers at least an idea of what to expect be it well lit room or dark theater testing in order to get the result ... at least at some point in the testing phase the results were acurate.
Good point Capt. The Pioneer is 20,000:1. The Samsung HP-T5064 is 15,000:1. The Panasonic TH-50PZ700U is 5,000:1.
The LCDs are both measured using Dynamic Contrast Ratio, which can't really be compared to static contrast ratio and usually results in a higher number (this helps explain why the Sharp has a much higher contrast ratio than the Panasonic plasma, even though the Panasonic has a deeper black level. The Sharp Aquos LC-52D92U has 15,000:1. And the Sony Bravia KDL-52XBR3 is 7,000:1.
I also ran across an excellent article full of a more technical explanation of all this stuff at Practical Home Theater Guide:
The Contrast Ratio Game
What we need is for someone, I don't know, some technology website webmaster perhaps, I can't think of what one, who is independent and who can come up with a fair set of tests and test the various TV's they review.
As for contrast ratio, you can measure how dark the screen is when all the pixels are black, and how bright it is when they are all white, but that doesn't tell you much.
A better test would be displaying a checkerboard pattern of black and white pixels and determining exactly how bright those black pixels are, and how bright the white pixels are.
You could then use this to come up with a contrast ratio, but as stated already contrast ratios are wildly dependent on small changes in black level. It would be much better to just list the black level, and the white level. So a TV with perfectly black pixels might be 0/1000 where 0 is the black level and 1000 is the white level of those adjacent black and white pixels.
You could also measure the reflectivity of the screen. You could say how diffuse it is, vs how shiny, and how much light is reflected on average back at a viewer sitting directly in front of the TV from a variety of light sources at different angles relative to the TV.
But really the best way might be to do what you're doing now, but under more controlled conditions, which is taking photos of the TV's. If you set up a room with some lights, and then set up a camera on a tripod with the exposure fixed so all the pixtures come out the same, and did a white balance correction for the room light, then we could see how each TV actually looks in a real room with a number of different pictures on the screen, like a color test, and an actual still scene from some DVD where there is some good variation in light and dark and color all in the same scene. Then we could at least see from a visual standpoint which TV's are darkest, and which are the brightest, and which have the best contrast, and which have the best color reproduction.
I made a post in a previous comment that is relevaant to this one.
http://gizmodo.com/gadgets/exclusive/adwatch-pioneers-proj...
Display manfacturers almost always quote sequental contrast which is showing a full frame of a white image then a full frame of a black image. You can cheat a lot when making up these numbers adjusting the backlight and what not to make bigger numbers.
What you want to see is simultaneous contrast which is measured with a black and white checkboard on the screen. This makes it more difficult to cheat with backlight adjustments and gives a sense of the display flare.
Display flare is any light that is present that isn't supposed to be there. The grey you see on a LCD panle when you are supposed to be seeing black is flare. You also get flare from the display it self. When you have a bright white dot on a black screen there is typically a halo around it. This extra light is a type of flare also. Finally you have the room the display is in. If it is brightly lit you are going to have ambient light bounce off the face of a display.
Flare degrades the contrast ratio of a display device but can not be avoided. The TV that looks excellent in a well designed home theatre may not be sufficiently bright to have a good image in a living room that gets sun light.
So when looking for a TV look to make sure that the maximum brightness is sufficient to overcome the room ambient light. Then look for the non-dynamic contrast ratio.
Is this helpful to anyone? I can write more :)
A note on dynamic contrast--
_______________________________________________
Dynamic contrast is a tricky concept. Almost all HDR display can be considered to be dynamic contrast in some way but it is done in what I consider to be a good way.
Standard dynamic contrast looks at the image average level and adjusts the back lighting or projector bulb voltage to create a blacker black or brighter white. Since it changes the whole image you can get undesired visual artifacts.
Elliuotatar- I'm right with you all the way to the point where you start talking about photographic the displays.
Digital cameras have sensitivities to near IR light much greater than that of human vision. For the most part it actually helps make pictures look better. However when you are looking for small changes in light in drak areas it will register the heat of the display as a fog over the image. The backlights for LCD's can also refelect a significant portion of non-visible light back at at the camera.
It would take an extansive setup of camera to make an accurate pictures of just the black levels. Color displays also give no indication of the display quality because you have to look at it through the limitations of photoshop resizing, jpg comression and the display you are using.
Photos of displays are only good to show off the case design.
Silly me...I always thought plasma screens could have pixels be in a "true off" state. Odd, since it's an electric discharge in there...you'd think pixels could be shut down with the TV on.
Techguy1138 and Elliuotatar have it right: the most honest way to measure contrast ratio is to use a black/white checkerboard pattern on the screen. Measuring the blacks and whites at the same time.
The most common use of "Dynamic" contrast ratio -- in which the whole backlight is dimmed from one time to another -- is not a good measure of image quality and should be CLEARLY listed by manufacturers IN ADDITION to static contrast ratio from a checkerboard type test.
Any method of measuring contrast ratio that involves altering room lighting between light/dark measurements for a single ratio, or removing parts of the TV (light filters on the front, etc) is basically just a lie.
I'd also like to see TV manufacturers list "reflectivity" of the screen -- that'd be a good measure of how room lighting will effect things. Usually not an issue with the "matte-finish" LCD's...but whenever I walk past the TV department at stores, I notice these strange clear plastic, slightly warped, very reflective coverings on the CRT HDTV's -- they reflect room lights like crazy, and the fact that they're warped/curved makes it look even worse.
Once the Brightside technology (array of individually-adjustable LED backlights. Technology now owned by Dolby) takes off for LCD's, I think things will be quite nice. This sort of "dynamic" backlight -- that can adjust the overall brightness with time, and the local brightness at different areas of the screen -- has great potential in my opinion. It should be able to "win" the Contrast Ratio war with an honest checkerboard and in terms of the "Dynamic" cheater method.
After that, something like OLED (which doesn't use a back light, and can have individual pixels be truly off) or SED (if it ever makes it to market) might be the technology that trumps an LCD with LED-array backlighting. If plasma's are still struggling with "true black," I don't really see a dramatic improvement coming soon...
Very helpful article. I spent several weeks researching HDTVs before biting the bullet and going with a 40" Bravia XBR2 before Christmas time. I found the different contrast ratio measurements used by the manufacturers to be confusing and occasionally misleading. Ultimately, I had to make the decision based on what looked best to my eyes.
Don't hold your breath waiting for an industry standard. Even those can be misleading. Remember back in the day when we thought clockspeed was the ultimate determinant of how well a CPU performed?
First of all, this is a FANTASTIC thread. I've been DYING to buy an HDTV but have constantly found flaws in each one that have prevented me from dropping the dough. So all of this is very helpful.
I find it interesting that contrast ratios are all the rage in current HDTV marketing. I guess there is always a new angle to pimp to the consumer. Anyways, I agree with EQC - I think OLED (and SED, if it ever stops being vaporware) is going to end the contrast ratio battle once and for all. And then it will be onto another marketing angle. Awesome!
...but it goes to 11.
This Brightside stuff got me interested.
Found a really good article as well. Mentioned alot of what has already been said here. Thought I would share anyway.
http://www.hardwaresecrets.com/article/208
Article says it was last updated Sept 05. Why have I not heard more about this and why are we not seeing it yet? I guess if consumers are going gangbusters for flat panels, why spend money on R+D.......
I'm still waiting for LED backlit displays before I get rid of my old CRT computer display. I guess Apple is doing it this year, of course more will follow. I'll find a nice 24" second Gen LED display for a good price.
I agree, and have known for a while, that the contrast ratios published by manufacturers are BS and not very comparable between brands. But one good thing is that it is the latest marketing trend, so they are sinking a lot of effort to improving their contrast rations, whatever they may be, and across all types of TV/monitor technologies. So the consumer wins, as every brand competes and we end up with much higher contrast than we had a year before, for the same price.
Just an FYI in case nobody mentioned, but it looks like the picture at you refer to as being the top with the plasmas is actually the 2nd pic on the page - 1st one looks to be the all LCD group, 2nd one has the Samsung and Pioneer (lack) of blackness.
Reminds me of the whole HP wars that plague car companies now - does anyone TRULY need 500HP in a daily driver? With accepted "speed limits" of 85mph on most roads....heck no! Will the average consumer know / see the difference between most of the specs that exist on TVs? Not too often.
But its good to keep everyone outdoing one another...or so "they" say....
Agreed about the photos - while they are useful, they should not be taken as a replacement for seeing these screens with your own eyes. However, while the photos may exaggerate things, these photos are a good representation of what I saw at the event. The differences between the screens was obvious.
Also, any shoot-out event put together by one vendor should be regarded with some suspicion.
Very true! Although there are so many variables at work for what each consumer sees, there's still something to be said about a setup that is truly more advanced or more accurately produces a realistic picture.
Put on by the mfr or not, still is better than a retailer showfloor :)
Consumer's need to read respected review sites that compare contrast among different models. They will almost never be able to see properly calibrated displays in a store.
For projectors, www.projectoreviews.com does contrast comparisons in it's reviews and www.cine4home.com does calibrated measuring of contrast for projectors
That photo idea is a good one but how would you keep out ambient light and what about shutter speed and aperture. I suppose you could just use the same setting but how would you be able to tell if it was the right setting in the first place
So, Gizmodo basically took all the posts from myself and Techguy1138 in all the previous threads where we decried the nonsense that is "contrast ratios" (which I, at least, have been doing for MONTHS) and made a whole thread about it? Sheesh. Hey Techguy...maybe we should apply to the Giz? ;)
Hopefully now people will stop worrying about this ridiculous spec until there is a more standardized understanding by the general public and a modicum of technical honesty on the part of the manufacturers.
If you want an example of how misleading this spec can be take a look at any of the photos above and pick out the WORST one in your opinion. Now realize that that TV may actually have the HIGHEST contrast ratio. All the manufacturer has to do is turn up the brightness to, as Noah stated, "arc welding torch" like levels while keeping the black point reasonably low and, viola, instant jump in contrast ratio. Don't worry about the fact that you'll need a welders mask to watch tv, just be happy that you've got that 40,000-1 contrast ratio!
The brightside stuff is not for home distribution. Putting a standard dynamic range image on a high dynamic range display does not improve it's appearance. Consumer based equipment does not have high dynamic range information stored. Once resolution and color space(xvYCC) standardize dynamic range is the next area of image improvement.
Noah- Even though the display illustrate different black levels that doesn't mean that the ones with high black levels don't have better contrast ratios or image quality. There is an implicit indication that high black levels equate to poor image quality. When place in a living room, tv's with higher black levels have an opportunity to produce better looking pitures at given ambient light levels.
I never said that contrast ratio were non-sense. The ratio that Sony gives for its Bravia seems spot on. It's just that people need a context to know how contrast fits in to whole of an "image quality" metric. As a rule of thumb divide any contrast measurement by 10 to get a more accurate measurment. It seems Sony took the correct measurement in the first place.
I still stand by the fact that photos of display are near useless for general use.
If you look at the photos of the 3 displays with no image than the ones with image the Samsung got deeper blacks and the Pioneer far worse ones. Wordy descriptions would serve people better.
The Bravia just flat out looks like it wasn't setup correctly. If the image looked like that on my display I'd return it to the store. Unless of cousre the other two displays had artifical amounts of contrast added to make them more easier to sell.
Having worked many years with light measurements in refrigerated display cases, a standard for measurement could be set really quick if there was a governing body.
In refrigeration (the indutry I work on) we go by ASHRAE Standard 117 or 72 for open or closed refrigerators.
When it comes to testing the refrigerator, the light in the test chamber has to be 75 LUX when measured in the center of the refrigerator display area and 12" away from the front (the most protruding feature of the unit)
I cannot see why a standard requiring a set parameter of ambient lighting measured in Lux or Candles and with a specific Kelvin temperature, and a set distance from the screen could not be set forth.
Then you would measure the contrast ratio.
What ever the result is, that is it.
This ratio would be affected by the light scatter effect of the screen surface material (light infiltrating the measuring device by the light scattered from the edge exposed to the ambient light, through the screen material, from the outside of the measuring device perimeter).
I really do not care if the screen surface has a filter (to reduce light scattering) or the TV has dynamic contrast (variable light intensity from the back lighting system).
These are just techniques to produce blacker blacks and whiter whites. They do however make a difference if the overall contrast ratio but add cost to the unit.
How the TV achieves the contrast ratio, I don't care. The one with the highest ratio wins the game. Whether it's dynamic, static, or by filters to reduce glare, it's not the point.
In the end, having a set environmental light level and one standard device to measure light intensity at any point with is the only way of measuring and to even the playing field.
which is the best brand to get? the one with the best lcd displays?
You probably know more about this than I do...but I have followed the brightside stuff a bit and their technology does not only apply to HDR imaging. The article I like about it is here:
http://www.bit-tech.net/hardware/2005/10/04/brightside_hdr...
from the last page, I get this quote:
So, if an LCD manufacturer wants to make "true black," whether or not they are going for HDR, then I think for now the best way to go is with Brightside's IMLED backlighting.
I think, then, that if the manufacturer did not want to go HDR, they could use weaker LED's than brightside used (so the display wouldn't be capably of blinding brightness), bringing the component cost of the LED's down, and still producing a TV that can make "true black" regions and also have a super high *realistic* contrast ratio [neglecting reflected room light, of course].
Techguy is right that you can't really upconvert a standard dynamic range image to HDR...but current LCD's cannot produce the full range of colors that can be encoded on movie disks, and an IMLED-equipped LCD will bring us close to that capability. In the case of a manufacturer actually making an HDR set, we'd be well beyond the specs for endoding color on movie disks.
...and I do remember Samsung showing off a few IMLED-based monitors at a recent consumer-electronics show...they didn't say if they were HDR or not...but they did have a very high *true* contrast ratio.
They mention it here:
http://www.engadget.com/2007/02/01/samsungs-15-4-30-and-40...
from which I extract this quote:
news like this should be taken with a grain of salt, of course, but it sounds promising to me :)
hmmm...might also look into the Samsung 81 series via a google search...it appears as though if/when (did it already?) those LCD tv's come out, they'll likely be using brightide technology.
mentioned on gizmodo here:
http://gizmodo.com/gadgets/notag/samsungs-81+series-lcds-h...
on AVS here:
http://www.avsforum.com/avs-vb/showthread.php?t=788306
(with lots of pictures)
Head on over to Gizmodo to read an excellent “contrast ratio” write-up: What is contrast ratio? Simply put, it’s the difference between the darkest and brightest spot on a display. This is expressed as a ratio measuring luminosity.
The problem with "judging with your own eyes" is that going to BestBuy and comparing models is not a good way to compare. Manufacturers know that the TVs that sell the most on the floor are those that appear to the brightest, so they come with default settings all jacked up way too high. Most consumers don't know this and so the TV looks all messed up at home without calibrating.
Most editors of magazines use the http://www.hqv.com/benchmark.cfm to test/review picture quality.
What I would love as a consumer is for someone to post all test results of this benchmark on the web similar to what these guys have done for DVD benchmark results here: http://www.hometheaterhifi.com/cgi-bin/shootout.cgi
Start a discussion:
Login with your username and password below. Or comment on this post via email.
Forgot your username or password? New User?