On Fri, 12 Sep 1997 13:10:00 -0500 Joe Little writes: [...] > > video in ->----------------------------------------->mixer----> >video out > | ^ > | | > v | > tv sync detector ic | > H Sync V sync | > | | ----------- > | | | | > | \--------------------------->| | > | | | > | | PIC | > v | | > PLL --> vco --------------------->clk| | > ^ | | | > | v ----------- > | divide by N * > | | > | | > \------/ > > > * Where N is (the number of PIC cycles desired per horizontal >line) * 4 > > Am I close?? Or completely in the weeds This is what my project does. There is no need to seperate the horizontal and vertical syncs since there is always a falling edge on sync (transition from "blanking" to "sync") at the end of each line, even during the vertical sync. The divide by 4*N is implicit by calling the horizontal routine every N instructions. The vertical sync is detected by testing the composite sync line after the horizontal routine completes (i.e. in the first part of the picture area). I don't interlace the ID text so it isn't important which field it is but I'm sure that could be determined by polling composite sync at the proper times. The VCO uses the PIC's oscillator.