Linards Ticmanis wrote:
Michael J. Mahon wrote:Of course, an NTSC processor doesn't care about boundaries, only about streams of bits--which I why I proposed a shift register "window" that is more than 4-bits wide, like 8 or 12, even, for greater fidelity to analog color processing.Interesting approach. How would the forumlae for determining one output pixel's color work in this case? In principle, I mean.
You would do it just as you have done in the past, but with a wider "window" and a lowpass filter applied to the chroma before demodulation. That "smears" the color out. Of course, the delay of the lowpass must be matched in the luminance channel to keep things in register. The width of the window is the only limit on the "context dependency" of a pixel's appearance. This would amount to a "canonical" analog NTSC simulation for inputs that are binary strings at 14MHz. The computed R, G, and B values for, say, 12-bit substrings could then be stored in RAM and indexed by the "shift register" for output to the DACs. -michael Parallel computing for 8-bit Apple II's! Home page: http://members.aol.com/MJMahon/ "The wastebasket is our most important design tool--and it is seriously underused."