HD81 Glitches, Calibration & Benchmarks

If I unplugged power to the video processor and reconnected before I started my video source, the cableTV system would incorrectly inform me that the display did not support HDCP. The error occurred with every channel and the Scientific Atlanta 8300HD DVR box was connected directly to the HD81's outboard video processor via HDMI (with low-bandwidth 480p resolution as the lowest common denominator. Subsequent success with other displays, cable integrity was easily ruled out as a potential cause of the error). The only solution involved disconnecting and reconnecting power to the projector itself - except that this didn't always work. Power cycling any of the other devices or terminating power to the video processor did nothing. After rebooting the processor and projector everything tended to again work perfectly, but sometimes the entire system would have to be brought down. After a few more days of viewing this happened again when switching channels - about an hour into watching television. This happened several more times over the course of 30 days and I could not seem to discern a pattern. Switching sources from cableTV to DVD did not seem to help - the HDCP error was still present, ruling out the source. No other projector installed into this reference system has ever behaved in this manner.

The other thing you'll need to be aware of is a rather slow switching process when the resolution changes on an input (think HD cable boxes). At about ~7 seconds, you will experience some delays when going from Fox HD (720p) to CBS HD (1080i) channels. Channels surfers will be most affected and there is no delay when switching between identical formats (i.e. the 480p digital SD channels).

Calibration - Getting Those Black Levels Right!

Though the HD81 is the best Optoma has to offer currently, it will take a few tweaks to get it behaving perfectly. I found that even when set to Warm color temperature and having Brite Mode disengaged, the system could not quite measure flat across the grayscale, nor did it have a smooth luminance response (we'll explain this later). Real-world performance was excellent, but if you use numbers to determine the best projector - you'll need to spend some time tweaking to get optimal results. CIE showed a lack of blue, green and red saturation in varying degrees.

We utilized datacolor's ColorFacts Professional 6.0 software to calibrate and measure the grayscale response of the HD81. Starting positions were a bit off what we expect to see in a high-end projector, but calibration proved successful for the most part as we quickly dialed in the HD81 to approach ISF levels. For the "before" measurements, Auto Iris was engaged and the settings were configured for best film response in a light-controlled room.

Both before and after settings looked fantastic to be honest. The calibrated image was much warmer, which is what I've grown accustomed to. We noticed a lack of smooth grayscale response. While 30 and 80 IRE were dialed in, areas in between seemed to run a bit loose. Further adjustments are recommended and you may want to set your lower level at 40 IRE for calibration. This projector had more of a "rubber band" effect than most, where adjustments in one area seemed to bring down that IRE level, but areas of higher and lower light levels wanted to stay where they were.

Our best "real world" measured contrast after calibration was 1648:1, pretty decent given our parameters and the resolution of our meter (we had several errant measurements showing over 20,000:1 but I think it's safe to throw those out and blame an unstable low light reading). Brite Mode will drop the color temperature a couple hundred K so take this into account when calibrating for various uses. At maximum output, we measured over 152 cd/m^2 reflected off the 100-inch Stewart StudioTek 130 screen. Yeah, it's bright.

Audioholics/HQV Bench Testing Summary of Test Results

Perfect Score is 130
Optoma HD81 Benchmark Score: 71 (The Optoma should be considered a suitable video processor for almost all uses. If you happen to have a high end player, simply feed progressive out of that component and the Gennum will handle the scaling of the input signal to the native resolution of the projector.)

* While the HD81 locked on to 2-2 cadence but had periodic "jumps" where it would quickly lose and reacquire the correct lock.

Comments on HQV Testing

All tests were done by feeding the projector a 480i signal via HDMI. Right off the bat we realized that the Gennum processing, at least this GF9351 chipset didn't have the smoothest jaggie reduction we'd seen. It was definitely reducing jaggies, but in a way that produces smaller jagged edges, rather than a truly smooth line. As a result we gave partial credit for those tests. This manifested itself not only in the Jaggies tests, but also with the more real world Flag test as well.

Failing the noise reduction tests really took me by surprise. I cranked up the Noise Reduction controls all the way to 10 and so absolutely no difference in the picture (even using the Demo mode for comparison). I would have to say that I saw no evidence for this function being enabled over HDMI.

The HD81 was a mixed bag on the AVIA Pro Moving Zone Plate tests, failing the standard Zone Plate test, but doing very well on the 2-3 tests in horizontal, vertical, diagonal, and circular directions (it seemed to take a while to catch on to the vertical movement, but afterwards didn't skip a beat). The 2-2 Zone Plate test went very well in horizontal, took some time to adapt in vertical and showed periodical glitches on that pattern as well. Overall, it sported impressive results that translated to excellent real-world performance.

In testing for rainbow effect (RBE) I found the unit to be pretty good unless you dart your eyes. If you have a "white on black" shape and move your eyes left to right, you're going to see rainbows (if you are susceptible to this). What this tells me is that people who sit very close to the screen (you know who you are!) and have to physically move their eyes to take in the left-to-right action may have some issues. The majority of viewers will not and I was never distracted during normal viewing.