Originally Posted by cakefoo
A test pattern with horizontal or vertical lines is not going to reveal the black and white of it all, because the LR signals are alternated in a checkered pattern. Put your DLP in 3D mode and if you're close enough with the glasses off you can see the checker pattern very clearly...
Turn the glasses on.
Resolution is measured in lines, so the test pattern is the only way to test it.
...My eyes were opened and looking at a DLP TV screen. Yours are open looking at marketing speak meant to hide the fact that the resolution is cut significantly...
I quoted a peer-reviewed scientific white paper, not marketing speak (remember your
"full 1080P" comment?), not some obscure article or blog. I tested it with the proper test pattern, looking at a DLP screen in 3D mode with the glasses on
. Try that test pattern on a passive set, and you'll see it displays 540 lines per eye.
...DLP has all 1920x1080 lines per eye, BUT since every other pixel in that line is missing, it's not full 1920x1080 per eye.
You don't seem to grasp the definition of resolution, or of 1080P. The first part is correct, except there is no "but". Again, there's no such thing as "full" 1920 x 1080. It either is 1920 x 1080 lines or isn't 1920 x 1080 lines. DLP in 3D is still 1080 lines per eye, and resolution is measured in lines. No market speak required, it's an industry standard.
You would however, be correct if you said that every other pixel "in the source" was missing, but let's see how DLP translates that to the screen.
While passive, or even Top and Bottom, skips entire lines
in order to work (and by definition, half the lines is half resolution), DLP skips pixels
, maintaining the same amount of lines
via the checkerboard pattern. The lines are still there. Thus it is still 1080p, can't argue around that.
Pixels are indeed missing from the source
, but DLP goes one or two steps further. Unlike a fixed pixel display such as LCD or plasma, the mirrors in each subframe of a DLP are diamond shaped and displayed twice as large
as the original pixel, so that it overlaps the adjoining "missing" pixels with its points. Thus the entire screen is used for each eye's view.
In this diagram, the mirrors are shown as diamonds, overlayed on the active pixels. Notice how the mirrors are twice the size of the square pixels (which are shown as gray squares behind the diamonds). The squares with dots in their center are the active pixels, the squares without dots are "blank" pixels, left out of the source because of the checkerboard sampling.
click to enlarge
Let's look closer at just one pixel. Note the different colors of the diamond pixels surrounding the “hole” where "blank" pixel 2-7 (Line 2, Column 7) should be. Although blank in the source, the color of pixel 2-7 appears as a combination of the other four diamond colors, and is formed by the intersection of their adjacent corners. (The other three corners of each of the adjacent colored diamonds are left blank for pixel clarity). It is in this way that the entire line is painted (no pixel location is left blank), unlike passive and other half resolution formats. This is not opinion or conjecture, this is how checkerboard 3D works, and why it works so well.
Call it pixel interpolation, magic, or even cheating, but it works. And by definition, it was always 1080P per eye.