> The thing about speed with respect to Star Trek, the reason the later
levels seem to slow
> down and lag is because there are so many objects to draw and compute.
The Z80 CPU
> always runs the same speed regardless of the code that it is running. The
more vectors
> there are to draw, the longer it will take to draw the screen image once.
The more
> complicated the image, the slower it will seem to be.
I've wondered the same over the years, but I'm not sure it's going to be a
vector generator bottleneck. I'm thinking more likely it's collision
detection (or other game overhead) soaking up lots of CPU.
Admittedly, it's been a long time since I've been into the guts of the G80
stuff, but IIRC Star Trek runs with a fixed framerate (as far as when the VG
tries to draw the display list). It'd be relatively easy to check with a
scope-- just see how long vector drawing goes on between frames. I'm pretty
sure the VG has it's own clock (not the same as the Z80's). I don't think
the little extra vector time for some more objects on the map would result
in such large slowdown. We're probably talking 100's of microseconds for
the extra inches of travel time.
Another trick to test for that kind of behavior is to build a debug version
of MAME and give the CPU a faster or slower clock rate and see if you can
provoke the slow-down earlier (or later) in the game. (Or just stick in a
slower crystal on the CPU board. A frequency generator works well for
experimenting with turning clocks up and down too. The Z80 is pretty
forgiving about it's clock and all its internal registers and stuff are
static.)
Depending on how much timing margin there is, it might just work to pop in a
faster crystal and see what it does. If you have a multigame kit installed
that'll get rid of the slow 2716's and you could always replace the 2114's
with 2148's if it's just RAM/ROM memory access that's the bottleneck. I
think most of the rest of the stuff has LS series latches which will
probably be OK at much faster than Z80 bus access speeds.
-Clay
---------------------------------------------------------------------------
** Unsubscribe, subscribe, or view the archives at http://www.vectorlist.org
** Please direct other questions, comments, or problems to chris@westnet.com
Received on Sat Dec 3 05:03:12 2011
This archive was generated by hypermail 2.1.8 : Sat Dec 03 2011 - 05:50:01 EST