Tests
Unlike most of the testing sections in our reviews, I present a summary first in this case. For more detail, keep reading.

Summary:
The ClearPix2 produces a subtle, and essentially insignificant, loss in the extreme high frequencies.

None of the deviations I measured in the ClearPix2's color performance are of more than academic interest. The screen is superb at retaining a projector's color quality and balance, both on and off axis.

The screen's brightness is extremely uniform from corner to corner, but its overall gain is very low. This must be carefully factored into your choice of projector and screen size.

Details:
I took audio measurements on axis through the screen alone and also with its transparent black backing. Neither had any significant effect up to 6kHz. At 8kHz, the response with the screen by itself was -0.8dB from the response with the screen retracted. With the screen and backing, it was -1.2dB. At 12.5kHz, the highest frequency I measured, the response was not reduced at all with the screen alone; with both the screen and backing, it was -1dB. I was surprised to measure less rolloff at 12.5kHz than 8kHz, but the result might be explained by comb filtering (an acoustic phenomenon that can cause closely spaced ripples in the response). Screen Research specifies no measurable comb filtering, but in any event, the deviations I measured, even if comb filtering was a factor, were acoustically insignificant. Yes, you can hear a change of 1dB if it occurs over a wide enough range. But it's subtle and more than compensated for by the behind-the-screen advantages already noted.

The video performance of an acoustically transparent screen is just as important as its acoustic signature. To check the ClearPix2 for color consistency, I first measured the post-calibration gray-scale tracking of a Yamaha DPX-1200 DLP projector on a Stewart Studiotek 130 screen (arguably something of an industry standard for projection quality). Our Photo Research PR-650 colorimeter was located behind the projector and aimed directly at the center of the screen—the usual technique I use when measuring color temperature. The results were very close to the D6500 standard. I then repeated the measurements on the ClearPix2 using the same projector settings. The maximum deviation was -113 kelvins at 30 IRE. At no other point in the brightness range was the deviation from the Studiotek's numbers greater than –60K; the average difference for all points was –51K.

Next, I checked how the color temperature changed at the far left and right of the screen, measured both from the position directly behind the projector and then by moving the meter about 15° off axis to the left. The off-axis position approximates a seat on one side of a large sofa positioned about 12 feet back from the screen. I took measurements at four points in the brightness range, from 40% to 100% peak white. Over a total of 20 readings, the maximum change I observed in the color temperature was +168K. The average of all these measurements was +105K.

Finally, I took another set of measurements from a more extreme 45° off-axis position—arguably farther to the side than most viewers would likely sit. Nevertheless, the maximum deviation from the straight-ahead, on-axis reading at this extreme position was +135K; the average was +50K.

I also measured the red, green, and blue color points from the ClearPix2, again comparing them with the results from a Studiotek 130. Neither the x nor y coordinates changed by more than 0.001—for all practical purposes, they were identical on both screens.

Keep in mind that all these measurements include not only the effect of the screen, but any changes due to the program material and projector as well. The projector used for these measurements was the Yamaha DPX-1200. Had the deviations been large, it would have been necessary to try to isolate these non-screen factors with additional measurements. But that wasn't necessary.

And what about the consistency of the light output from various parts of the screen? At the far corners of the image, the ClearPix2 retained an average of 90% of its on-axis output. In comparison, the Studiotek 130 retained 73%, and the Stewart FireHawk retained 53%.

It's also useful to know how the light output of a screen holds up as you move to an off-axis seat. From about 15° left of center—again where you'd be on the side of that comfortably large sofa—the brightness from the center of the Screen Research was still 83% of its on-axis level. Compare this with the FireHawk, which is down to 63% from the same location. Of course, keep in mind that the FireHawk's off-axis performance is intentional. It is designed specifically to focus its output toward the center viewing area and minimize off-axis output, in order to limit possible reflections from the sides of the room and back onto the screen.

Up to this point, the Screen Research excelled in meeting its design goals. But it does have one Achilles heel: gain. To evaluate this parameter, I projected a 100 IRE white full-field pattern on the ClearPix2 and Studiotek 130 (both with 80-inch-wide images). I then calculated the ratio of the brightness readings and used that number to calculate the gain of the ClearPix2 relative to the Studiotek's published gain of 1.3. In an average of five measurements taken with different projector settings and the light meter in the same position behind the projector, I determined that the ClearPix2 provided about 62% of the brightness of the Studiotek, which yields a gain of 0.83. Of course, this assumes that the Studiotek actually has a gain of 1.3, which I am presently unable to determine experimentally.

While we have not tested all the single-chip DLP projectors on the market, I would be very reluctant to use most of them on anything but a relatively small ClearPix2. Even 80 inches wide will be marginal on lower-output projectors fresh out of the box, ignoring the inevitable dimming you can expect as the projection lamp ages.

I did experience good results with the ClearPix2 together with the Yamaha DPX-1200, achieving just over 9 foot-lamberts peak-white output with an 80-inch-wide image (mid iris setting, maximum lamp output, maximum white peaking, and 80 hours on the lamp). So it can be done. But I strongly recommend choosing one of the higher-output DLP projectors if you plan to use this screen.

We hope to provide you with screen-compatibility information in our future projector tests, but obviously we can't review all the projectors on the market. The smart buyer will test the projector he or she plans to use on the ClearPix2 of the intended size to insure a satisfactory result. This is good advice for all projectors, not just single-chip DLPs.