- Written by Chris Eberle
- Published on 28 January 2010
- Samsung SP-A900B Single-Chip 1080p DLP Projector
- Page 2: Samsung SP-A900B DLP Projector Design
- Page 3: Samsung SP-A900B DLP Projector Setup
- Page 4: Samsung SP-A900B DLP Projector In Use
- Page 5: Samsung SP-A900B DLP Projector On the Bench
- Page 6: Conclusions About the Samsung SP-A900B DLP Projector
- All Pages
On The Bench
Although color calibration was a snap with the A900, grayscale and gamma adjustments took me longer because of the many options available. I quickly discovered that the different iris settings gave me varying levels of accuracy for both grayscale tracking and gamma. Furthermore, I found more gamma presets in the service menu that allowed me to experiment a little
Equipment used: EyeOne Pro spectrophotometer, CalMAN Professional 3.3 analysis software, Accupel HDG-3000 signal generator, Spears & Munsil Benchmark Blu-ray.
All measurements were taken off the screen (Carada Brilliant White, gain 1.4) from the seating position (10 feet).
The color measurements both before and after adjustment were barely different. The sample I worked with had been bench-calibrated (meter pointed at the lens) previously so I only had to make tiny changes to get a nearly-perfect CIE triangle and luminance values which were within .3fL of the targets.
Grayscale and gamma were also very good both before and after the color calibration.
The average gamma value of 2.5 is somewhat higher than I’m used to seeing. I asked Mike Wood at Samsung about it and he passed along an explanation from Joe Kane. He sets the film gamma at 2.5 in order to better match the CRT mastering monitors still widely used to create film-to-video transfers. I have heard this argument before and it makes sense. Still, I’ve never particularly liked the look of a display set to this gamma level. I feel it flattens mid-tones too much. A gamma average of 2.2 provides a much punchier image. There is some debate among calibrators about this. Since very few displays can even achieve 2.5, the point is mostly moot. The A900 can do this easily but there is a 2.2 gamma curve available in the service menu. I assigned this curve to the video user menu slot so I could have both available from the remote.
After selecting the 2.2 gamma preset, setting the iris to a fixed value of 73, and adjusting the user menu white balance controls I recorded the following results. This is excellent performance.
Contrast performance was decent though not quite at the level of the high-end LCoS designs. With the iris set to 73, the minimum black level was .001 foot-Lamberts. I set the iris to output a peak level of around 13 fL. This gave me an on/off contrast ratio of 13,000:1. With a 300-watt lamp, there is plenty of light available. With the iris fully open I measured over 19 fL, too bright for my theater. This was with the lamp at its lower brightness setting called Theater.
I tested the video processing with the various Spears & Munsil clips played from an Oppo BDP-83 set to source direct mode and connected directly to the A900. Performance was below average. The projector failed all the source adaptive tests. Edge adaptive tests weren’t too bad with minimal jaggies on the ship ropes. The A900 would not accept a 480i signal over HDMI. 1080p/24 signals were handled correctly with the projector outputting a 48Hz refresh rate. Though this rate is lower than some other projectors, there was never a hint of flicker or motion blur.
The poor video processing performance was a surprise but not a problem for me. There are so many source components that employ quality video processing, it’s just not necessary to have it in the display. Setting my Oppo BDP-83 to output 1080p24 or 60 resulted in superb image quality in every respect.
The lens is one of the best I’ve ever encountered. It is easily the best I’ve tested at this price point. Full-field uniformity was excellent with no visible color or brightness variation anywhere on the screen. There was no sign of chromatic aberration with excellent edge-to-edge and top to bottom focus.