Marantz VP-15S1 DLP Video Projector Measurements
All of the geometry and convergence parameters here were as superlative as one expects from digital and not worth enumerating specifically. Looking at a bright crosshatch pattern there was very little color fringing or chromatic aberration visible on screen. What was there was less than a pixel in size and at the sides of the image and caused no discernible artifacts from the viewing position. Marantz' 1080p projectors have not only been excellent in this regard but remarkably consistent from unit to unit. I reviewed one VP-11S1 and bought another unit months later. Both were as close to perfection as I've seen, and this PJ isn't far off.
Also speaking to the quality to the lenses Marantz uses, as mentioned, I could not only see the pixel structure clearly I could see the dimples on each DMD clearly if I walked up to the screen. HD luma bursts at 1080p over HDMI at the highest frequencies, one pixel on/one pixel off, were as perfect as the VP-11S1, which is to say as perfect as I've seen from any display device. No softness in the line structure or rolloff whatsoever, and the sharpness patterns showed no ringing or enhancement artifacts, just pure resolution. Again, with static patterns, there was little to choose between the two, but with moving images and program material there was a refinement to the 11S1 that the junior projector couldn't quite eek out. Of course, the same could be said of every other projector I've ever seen so no shame there. Chroma bursts were also resolved to the limits. Again, simply superb.
As time goes on we gain access to more new and interesting tools for display evaluation. In particular, we now have high-def test discs from Silicon Optix, Stacey Spears and Don Munsil that were not available at the time we tested the VP-11S1. Our primary tool at that time was an Accupel test signal generator with a DVI output. While DVI is backward compatible with HDMI, I later learned of an interesting way in which this fooled me with the 11S1. Since the Accupel is DVI its output to the projector is RGB, not the Y-Cb-Cr signals now coming from most HDMI source components, including Blu-ray and HD DVD players. The VP-11S1 did not clip below black or above white with RGB, but did clip with certain Y-Pb-Pr signals. So, setting up and testing with my generator didn't reveal this and I didn't report on this, but found out later.
The 15S1 does not clip with RGB or Y-Pb-Pr at 4:2:2 or 4:4:4, and it in fact can be set up to resolve many steps between digital 235-and 255 so there's not only headroom, but resolution above white. Typically we've looked at patterns that show steps between 90IRE and 100IRE (the latter of which has a value of digital 235) to set the contrast on a display to avoid clipping, and if we have instruments we also look for top-end color shift. That's the little dance in setting up a display- you generally crank it as high as it will go without clipping whites or color shifting.
Now, program material is not technically supposed to be authored with signal information above white, or below black. But according to some this does occur and if your display clips this information it's lost on playback. With the Spears and Munsil test discs it's possible to avoid clipping far above white, but this can result in a lower contrast setting (and therefore lower light output) than if you only set to avoid clipping up to 100IRE (digital 235). I've been experimenting with the various displays I have on hand, and I'm comfortable sacrificing some light output to get this. Aside from eliminating clipping, it just seems like the projectors are operating in their optimal range, rather than being pushed to the max. But as with so many things, individual taste plays a part here, so season this to your tastes.
TJN will plot this PJ's grayscale, but my own measurements showed outstanding performance out of the box, what went to superlative after calibration (probably more accurately referred to as a touch-up). Striking was that it wasn't just conformity to 6500K that was close, but the more important adherence to the D65 point in color space was on the money before and as close to perfection from 30IRE to 100IRE as I've yet seen in any display device.