At 9:48 PM -0700 6/10/04, John Willkie wrote: >Using those figures, I get similar results: 38.55 inch line width, or a 44 >inch diagonal measure 16:9 screen, and a 48 inch or so 4:3 screen. Of >course, I haven't accounted for having more than one dot of each color per >pixel, nor the actual pixel configuration. > >The issue, then, is how many sets have such closely spaced phosphors, and >how many sets have monospaced pixels. Neither appears to be universal. Phosphor spacing is a non issue, unless the phosphors for the primaries share the same real estate, as is the case for direct view CRTs with shadow masks. If you place an RGB phosphor pattern on the screen, you are in essence defining the limits of resolution - i.e. quantizing the display; it does not matter if there is a shadow mask or not. Using separate CRTs for each primary eliminates this issue. The entire face of each tube can be covered with the same phosphors. When this happens the limits of resolution are defined by the size of the spot beam, and the bandwidth of the amplifiers that carry the source signals.It is important to note that you still have the issue of spot size elongation near the edges of the CRT. Unfortunately, with projection CRTs you ALSO run into another problem...brightness. In order to get reasonable brightness levels - especially in rear projections applications - you need to maximize the beam current, and this in turn usually means a large spot size. If you follow the instructions that Joe Kane recommends for maximizing the performance of CRT-based projection systems, the first thing you do is to back off on the brightness and contrast to control the spot size so that you can maximize resolution and eliminate the problems with blooming. This of course means that you end up with a picture that is less bright, with less contrast; but you can deal with this by watching the display in a darkened room...like a movie theater. These issues all go away when you move into the world of micro displays where there are "lithographed" pixels. These display have near perfect geometry (except for any distortions introduced by lenses and mirrors), and the brightness/contrast is -in most cases - significantly improved. One other note. The desire to have more than one screen sample per source pixel is an oversampling issue. As we have seen with image capture there are significant benefits to over sampling during acquisition. The same is also true for displays. More samples help to eliminate the perception of individual pixels, and with analog sources to provide finer quantization steps as the levels change. But there are limits here too; the limits of human perception. When you reach the point that the observer cannot resolve individual pixels, additional resolution is mostly wasted. Just step a few feet away from your computer display and you will see a real world example. Regards Craig ---------------------------------------------------------------------- You can UNSUBSCRIBE from the OpenDTV list in two ways: - Using the UNSUBSCRIBE command in your user configuration settings at FreeLists.org - By sending a message to: opendtv-request@xxxxxxxxxxxxx with the word unsubscribe in the subject line.