|
|
楼主 |
发表于 2012-10-2 09:00:41
|
显示全部楼层
Weaknesses and limitations of DLP (1 of 2)
本帖最后由 MySense 于 2012-10-2 09:02 编辑
Color wheels can produce rainbow artifacts. The problem people point to most frequently as a weakness in DLP is its tendency to produce "rainbow artifacts." Rainbow artifacts (sometimes referred to as color separation artifacts) are momentary flashes of banded color that look like rainbows. They occur at random, and they only last for an instant. But for people who are sensitive to them, they can be quite distracting. If you are engrossed in a film or video, they can take you entirely out of the video experience.
Rainbow artifacts are a problem only on single-chip DLP products, and for the most part, only those using slower speed color wheels. They can also occur on LED-based models due to the sequential strobing of red, green, and blue LEDs. Typically the problem manifests itself when the viewer is watching movies or video. When viewing static images such as presentation charts or photographs, people generally do not experience the problem.
The rainbows occur because of the sequential color updating from the wheel or LED. As the color wheel spins or the LEDs change, the image on the screen is either red, green, or blue at any given instant in time. The technology relies upon your eyes not being able to detect the changes from one to the other. However, when your eye moves rapidly in response to some movement in the picture, you can get a red, green, and blue update on three different points on your retina, thus producing the impression of a rainbow. Not everyone perceives rainbows the same way. Many people have less sensitive eyes and cannot detect rainbow artifacts at all. Others see them quite readily. There is no way to know whether you are among those who can or cannot see them except by watching a DLP projector yourself.
Since LCD projectors and 3-chip DLP projectors always deliver a constant red, green, and blue image simultaneously, they do not create rainbow artifacts.
On DLP projectors with color wheels, rainbow artifacts are reduced by increasing the speed of the wheel. The first generation DLP projectors incorporated a color wheel that rotated sixty times per second, or 3600 RPM. With one red, green, and blue filter in the color wheel, updates on each color happened 60 times per second. This rotation speed in the first generation products was known as a "1x" rotation speed. In second-generation DLP products, the color wheel rotation speed was doubled to 2x, or 7200 RPM. The doubling of the color refresh rate reduced the time between color updates, and so reduced the visibility of rainbow artifacts for most people. But a 2x rotation speed was still not fast enough for products to be used in home theater and video applications.
Today, some DLP projectors being built for the home theater market use a color wheel containing two sets of red, green, and blue filters. This wheel still spins at 7200 RPM, but because red, green, and blue are refreshed twice in every rotation rather than once, the industry refers to this as a 4x rotation speed. And by increasing the physical rotation speed beyond 7200 RPM, some projectors now have 5x or 6x speed wheels. For the large majority of users, the 5x and 6x speed wheels in most current home theater models have reduced rainbow artifacts in video display to the point where they are of little or no concern.
However, most DLP projectors built for commercial/presentation use still use 2x speed wheels because they are less expensive. This is perfectly fine if the presentation matter is static charts, graphics, photography, or anything that does not stimulate rapid eye movement. We do not recommend DLP projectors with 2x speed wheels to buyers for whom video display or part time home theater are important intended uses.
Color saturation/color brightness. Some DLP projectors have excellent color saturation, and some are exceptionally poor. This is related more to the vendor's implementation than anything inherent in the technology itself. Advocates of 3LCD technology have been quite vocal about the lack of color brightness on single-chip DLP products, particularly those that have white segments in the color wheel. This phenomenon is worth commenting on.
When a color wheel has a white (clear) segment, the lumen output of the projector is increased dramatically, and the ANSI lumen rating skyrockets. Most business class DLP products have white segments in the wheel to boost the all-important lumen rating. Conversely, most DLP projectors built for home theater have no white segments because they can compromise color saturation and the overall balance of the video image. Moreover, the lumen rating is not a big driving factor in the sale of home theater projectors.
When you use a light meter to measure the brightness of red, green, and blue on an LCD projector, the sum of the values usually adds up to the brightness reading you get for white. This makes sense because on an LCD projector, white is created by turning the red, green, and blue channels all fully on. But on a DLP projector, this is often not the case. Due to the presence of a white segment in the wheel, the white reading can be as much as double the sum of the brightness readings for red, green, and blue. In other words, if an LCD projector measures 2000 lumens of white light, you will also get 2000 lumens of color light out of it. If a DLP projector measures 2000 lumens of white, you might get only 1000 lumens of actual color light from it, the rest being white light.
Because of this, proponents of 3LCD technology have been lobbying for color brightness specs to be included along with ANSI lumen specs on the industry's specification sheets, and support for this has been building in the industry. In the spec wars, quite clearly this would be one metric on which LCD has a commanding advantage over DLP. Not surprisingly, Epson and Sony have already begun to publish color brightness specs on their LCD projectors to drive home the point. The color spec is always the same as the ANSI lumen rating, and the specs will read, as an example, "2600 lumens color light output, 2600 lumens white light output."
As a rule, DLP projectors with a white segment in the wheel do not look very appealing when measured for color brightness. The color reading always fall short of the white reading, and sometimes it falls short by 50% or more. This is especially true when the color wheel contains the basic red, green, blue, and white filters only. Many DLP projectors have complementary color filters such as cyan, magenta, and yellow. In this situation the color brightness measurements become more problematic. Thus we can understand why Texas Instruments and the DLP projector vendors have little interest in publishing color brightness specifications.
From a practical perspective, we have mixed feelings about all of this. Clearly, the 3LCD camp is correct that the traditional ANSI lumen spec does not tell the whole story. But neither does the color brightness spec. To be sure, the color on some DLP projectors looks dull and anemic next to some LCD models of the same lumen rating. Ironically, this can be particularly true when the "BrilliantColor" feature is enabled. Though BrilliantColor boosts the brightness of the image, it can substantially reduce color saturation in the process. It is peculiar that in order to get the richest and most saturated color from many DLP projectors, one needs to turn BrilliantColor off. (This is not universally true of all DLP projectors with BrilliantColor, since the BrilliantColor system can behave quite differently based on how it is implemented by the vendor.)
Oddly enough, on some DLP models with white segments in the wheel, even those on which color brightness falls far short of white, we see a rich, vibrant color that can easily match an LCD projector in the same price and lumen class. One reason is that the color filter configuration of the wheel has a lot to do with the end results. Another reason is that, though the DLP's color brightness may fall short of white, the effect of the DLP's inherent contrast advantage helps to compensate for it. That compensating effect cannot be quantified in a spec. Even when color brightness falls very far short, the picture sometimes does not end up looking much dimmer at all when put side by side with an LCD projector of the same white light output.
When a DLP projector's color vibrancy looks poor next to a comparably priced and spec'd LCD projector, it is due to a variety of design and product cost decisions made by the vendor, and not anything inherent to DLP technology per se. DLP can look truly spectacular or downright dismal depending on what is done with it. With so many variables in play, the specs can't tell the whole story, even if a color brightness spec were added to the mix. The publication of color brightness specs would be interesting, and would certainly draw attention to a noteworthy technical difference between LCD and DLP. But it is not conclusive information that would help an astute buyer sort out which model to buy.
|
|