>>
If it requires 16 pixels and we use a centered 32 pixels of playfield
graphics (i.e. 2 pixels per bit) (vs. 40) it MIGHT work. I just don't know
if the signal has to extend across the full width of the screen or not. I
suppose it does. I don't think the average decoders are tolerant enough to
allow for a squashed signal that starts well into the scanline and ends well
before the end. You'd really have to try to evenly scale out those 16
pixels evenly across the full width of the VCS's screen and even then the
VCS's left and right border is fairly wide as it is.
<<
If it really requires the full scanline, you could use self-modifying
code in the RAM. Load the X-register with white and the Y-register
with black, and then have a series of "STX COLUBK / STY COLUBK ...".
You can change the opcodes for STX/STY as you need the pixels. A
STX to the zero page takes 3 cycles or 9 pixels, so you should make
every third command use absolute addressing, which takes 4 cycles
or 12 pixels. That might be close enough to the 10 VCS-pixels per
CC-pixel to be decoded properly.
Couldn't anyone with a TV with a V-hold knob or a TV card that shows
all scanlines have a look at a TV show with closed captions and find
out how much of line 21 is used for the CC signal and how many pixels
there are?
Ciao, Eckhard Stolberg
----------------------------------------------------------------------------------------------
Archives (includes files) at http://www.biglist.com/lists/stella/archives/
Unsub & more at http://www.biglist.com/lists/stella/
