Will TI retire the spinning filter wheel that's used in most DLP-based rear projection TVs and projectors? The old filter wheel may still have one advantage: A richness of color.
Does laser light look different?
Hearing an announcement earlier this year that Mitsubishi has unveiled the first laser-based DLP rear projection TV (see Mitsubishi launches LaserTV), my first reaction was a big hurrah. If this takes off and gets cheaper, I reckoned, TI could once and for all retire that spinning filter wheel that's used in most DLP-based rear projection TVs and projectors. The spinning wheel creates a slight bit of audio noise, has been accused of causing visual disturbances and artifacts in a small number of viewers (aka "the rainbow effect"), and is a mechanical component in what's otherwise a solid-state device.
But compared with laser illumination, the old filter wheel may still have one advantage: A richness of color.
The distinction here is a bit like the difference between illumination from incandescent rather than fluorescent light bulbs -- and is, in fact, the very reason why many people prefer the former (despite efforts to ban it, due to energy inefficiency, such as in Ireland!)
If you look at a spectrum analysis showing the frequencies of light emitted from an incandescent bulb, you'd see a nice continuous curve with emphasis in the red area (lower frequencies), but essentially including, to a lesser degree, all the colors of the spectrum right up through blue and even a little ultraviolet (at least that's how the old "black light bulb" in my college dorm room worked!)
A fluorescent bulb, by contrast, has a spectrum graph that looks like a series of vertical lines, representing some very specific frequencies of red, green and blue light which mix together to create the appearance of white. But not all colors in the visible spectrum are represented here -- far from it -- there are just a handful of specific wavelengths. This is why, when compared to incandescent light, fluorescent seems "color starved."
What does all this have to do with DLP?
When the DLP filter wheel is on the red filter, for example, the spectrum produced is not a single color, but a range of colors all concentrated around a particular wavelength of red. Its spectrum graph would look like a curve, not a vertical line. Ditto for the other colors in the filter wheel, which also includes white, producing all the colors in the spectrum (assuming the light source is a halogen bulb, which is similar to tungsten bulbs.)
With laser light, however, it's a different story. Three lasers -- red, green, and blue -- produce three very specific wavelengths. These would appear as vertical lines in a spectrum graph.
Of course, by adjusting their comparative intensities, every color within the standard video color gamut (or even a wider gamut) can be produced.
However: First, the system is not analog, but digital. So the number of available colors is actually limited to the smallest increment, typically 1/256th of full brightness (8-bit per channel color), resulting in 16-million color combinations. OK, good enough, and true with any digitally-fed display.
But what's different with laser-based color is that there's no gentle roll-off around these 16-million colors. That is to say, there's no "sloppiness" to the color accuracy, as you'd get with filter-wheel illumination, or for that matter, with CRT phosphors.
Can consumers see the difference? Does anyone care? Those questions remain open. Some may even prefer the more perfectly precise color from lasers. But just like the temporal color mixing inherent in single-chip DLP technology itself (red and blue never appear simultaneously -- you create magenta by quickly alternating between red and blue in the same spot), laser color does raise some interesting questions about what we're "feeding" the eye-brain system when we watch TV.