I have an observation from working on the score section of Death Derby...
I've noticed that when you dynamically change a sprite color when the
playfield is set to inherit its color from each side's matching player that
the hardware doesn't register the change in alignment with the playfield
pixel. You'll actually get 2/3rds or so of playfield and then it
switches. It actually looks pretty messy. It might not even be in
alignment with sprite pixel widths which would be really unusual for the
hardware. I'm thinking this has to be the slowest section of the hardware
to respond to dynamic writes, maybe because it has to go through the extra
lookup of the player colors during scanout whereas if you just changed
COLUPF it wouldn't react as slowly. Most all of the other things you can
do with the 2600 act in a snappier, predictable fashion.
The last demo version of DD that was showed last year had a little edge of
the wrong color in the top left in the gear-indicator area that is a
symptom of this.
Has anyone documented this more thoroughly?
----------------------------------------------------------------------------------------------
Archives (includes files) at http://www.biglist.com/lists/stella/archives/
Unsub & more at http://www.biglist.com/lists/stella/