Re: pixel's shape, and Slot Mask CRT displays [Re: Question with regards to componentAlpha]
Keith Packard <[email protected]> Wed, 25 Sep 2002 07:51:01 -0700
| Newsgroups | gmane.comp.xfree86.render |
|---|---|
| Message-ID | <E17uDVG-0000ko-00@localhost> |
Around 11 o'clock on Sep 25, Vadim Plessky wrote: > In fact, pixels in traditional Invar Mask CRTs are *round*, or close to being > round. RGB dots are placed in form of triangle. Because there's no alignment between the pixels in the graphics cards and the arragement of phosphors on the face of the monitor, there's no way to accurately model the area illuminated by a single component. A single component from a single pixel will always generate a nearly gaussian spot on the monitor which will hit one or more elements, many of which will be the right color. CRTs have many advantages, but a simple illumination model is not one of them, and attempting to model the actual pattern of light is impractical with our current mechanisms. Keith Packard XFree86 Core Team HP Cambridge Research Lab