For all of these measurements, the Stewart Studiotek 130 screen referenced in the comparisons (78 inches wide, 16:9, gain 1.3) was used only to a width of 72 inches, a close match to the 73-inch width of the Supernova. All light readings were made off the screens with a Minolta LS-100 light meter. The color temperatures and light falloff readings on both screens were made in a completely darkened room. The off-axis viewing position used for some of the measurements was equivalent to the left side of a large sofa centered on the screen and about 10' back from it.

Compared to a gray scale calibrated on a Stewart Studiotek 130 screen, the color temperature measured at the center of the Supernova did not vary significantly. From low to high brightness, it remained within 100 degrees Kelvin of the readings on the Studiotek. However, as I moved to the off-axis seating position (toward the left), the color temperature of the Supernova increased by up to 450 Kelvins, worst case, compared to the Studiotek. The Studiotek's color temperature was also more uniform from left to right at this off axis position (approximately +/-110K) than the Supernova (approximately +/- 300K). I did not notice any significant problems, in normal viewing, from any of these deviations. But they might be visible on some program material.

A screen image is always brighter at the center than at the edges due to the limitations of all projectors, but higher gain screens contribute to this effect as well. The Supernova was no exception. Measured directly on-axis, the drop in brightness from the center to the edge of the screen was 70% (the edge measurement was an average of readings taken at six points, three on the left and three on the right). That is, a detail at the center of the Supernova that measures 10 foot-Lamberts in brightness would measure approximately 3fL if it were located at the edge. The same measurement made on the Studiotek 130 resulted in a drop of only 24% (or a reduction to 7.6fL at the edge for the same hypothetical 10fL source).

When I moved the meter to the off axis (left) viewing position, the brightness on the right of the DNP was only 22% of the brightness on the left. With the Studiotek 130, the right side was 85% of the brightness on the left.

This uneven brightness as viewed from an off-axis location on the Supernova was, as you might expect, visible to the eye. The light falloff to the sides as viewed from a center seat was far less so. In neither case, however, was the effect a significant distraction on normal program material, though I was occasionally conscious of it at the off-axis location. (I have observed a similar off-axis performance from the Stewart FireHawk screen). While this uneven light output might suggest that the DNP will suffer pronounced hot spotting, I did not find the Supernova's hot spot characteristics anywhere near as obvious as it is on many rear projection sets.

Finally, I measured the respective peak contrast ratios (peak white vs full screen video black) of the Supernova and Studiotek 130, with both the mild ambient lighting used for most of the viewing tests and, for comparison, with the room lights off:

(* peak white/video black, in foot-Lamberts)

It's obvious from this result that the Supernova produces a significantly improved contrast over the Studiotek 130 with some ambient light—though remember that this is modest room lighting, sufficient for most social activities but not "let's open all the drapes and Let the Sunshine In."

Also notable here is confirmation of the Supernova's rejection of stray room light. Notice that while the peak white reading with the lights on is just over twice that of the Studiotek, the black level from the Supernova is lower.

It's also clear from the results that even a modest amount of room lighting seriously degrades the full contrast potential of a good projector—on either screen.