Not so Black and White?
Contrast ratio testing is not only one of the most important tests for validating the performance of a display system, it is also one of the most difficult to do with some display types. Recently I've been in the trenches working on a review of a new premium flagship level projector for Sound and Vision. Due to the nature of the design, I've been mulling over different ways of measuring contrast to provide the best approximation of true performance as possible.
I've often seen it mentioned that the popular sequential or "On/Off" contrast ratio is nothing more than a marketing gimmick. While I certainly agree that it can be abused, as all specifications can, when measured properly it can provide a lot of detailed information to the end consumer. It tells you the full dynamic range of the display, and gives you a direct representation of what you can expect for black level performance in the display system when you know its peak light output capabilities.
But one way this can be manipulated is with designs that can turn off their light source. This can be a projection system with solid state illumination like a laser or LED light source. Or it can be a flat panel display with a zoned back light system. With products like these it is easy for a manufacturer to simply turn off the light source to provide for an infinite contrast ratio specification. This does work with a full blackout in content, but it is a fallacy when it comes to the display's true dynamic range. What happens when you illuminate so much as a single pixel on the screen? What typically happens is the black floor of the display jumps considerably to its true native contrast performance. Depending on that number, that jump can be quite jarring for the viewer, especially when you are watching content that jumps in and out of black for a sequence or has a fade in and out of black, which is pretty common.
To counter this I worked with Stacey Spears from Spears and Munsil. They are the producers of the popular HD Benchmark Blu-ray that is a must have for any enthusiast wishing to not only calibrate their display using the common settings, but also evaluating its performance with fantastic test patterns and tutorials. Stacey generated a test pattern for me that is pure black with only a single pixel illuminated. This is a native 2160p pattern (3840x2160), so only one of the 8,294,400 pixels is actually illuminated. But I thought this would be the perfect pattern to test for a better approximation of the dynamic range capabilities of a display that uses a light source that can simply be shut off.
The testing was simple enough, but it also uncovered an interesting issue. With the display I was currently evaluating, it not only revealed the native dynamic range of the display, it also affected the contrast ratio performance with the dynamic light source disengaged. I wasn't expecting this from a single pixel, but the measurements showed exactly that.
Since my review has not been published yet, I went back and performed the exact same test on a projector I had just reviewed recently that I still had on hand; the Sony VPL-VW885ES. Here are the results:
The Sony VW885ES has two dynamic modes for its laser light engine; FULL and LIMITED. The "Native" column represents the performance without the dynamic mode engaged. When comparing the sequential contrast (ON/OFF) to the single pixel, you see how that one pixel actually affects the black floor enough to lower overall dynamic range. This means that while the sequential contrast measurement will tell you the best black floor you can expect, it only pertains to true absolute black only (similar to the dynamic measurement). But a single pixel will change the dynamic range and the black floor you will achieve with any content on screen, regardless of how small. I will be including this measurement in all my display evaluations going forward, regardless of the technology. Keep an eye out for more on this down the line!