RE: Star Wars Tie fighter issues

> Video shows not only the tie fighters having issues but
> possibly the death start and trench too. It really looks
> like a scaling issue but it recovered a few seconds into the
> trench run. Hummmm

That's pretty neat. ;-)

Smells like Mathbox to me... It's basically 'zoomed' in for the transforms
and as a result is drawing all of "space" at a much larger scale which takes
longer (dropping the frame-rate, hence why it's flickering more) and drawing
off screen which gives the flashes of beam scatter illumination on the CRT.
It's probably not the AVG since the rest of the HUD is drawing OK, so things
are executing like they should by the time it's ready to display
(subroutines, vector list, etc.). Connecting it to an oscilloscope might be
neat-- you'd probably see all of 'space' being drawn out to the limits of
the DAC voltage range.

It's almost like there's a bad value getting into the transform math. 5F
and 5H (mathbox RAM) would be easy to swap if you haven't tried that
already. A problem with the serial multiplier and/or MAC clock could result
if a scaling issue (multiplies are off by an order of magnitude or
something), so 10C, 9C, 6C, 7C could be a cause. I wouldn't think the data
registers would be too likely since you'd think it'd mess up some other
coefficient in that case and you wouldn't have a good image (ie, it'd be
wrong on a translate or something too and the shapes would be in the wrong
position, not just at the wrong scale).

(I wouldn't think it's related to the accumulator since it seems to be
getting the basics of the math right, just with the magnitude off-- I'd
focus on the multiplier.)

This will be interesting to hear the root cause when you find it!

-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 Thu Aug 2 20:09:48 2012