Hughes, B. wrote:

> was expecting to find instead was a bright spot where the top of the dome
> meets the part atop it, since that's where I notice the most slowdown occur.
> Or so it appears to me as I watch the rendering take place.


I think you can improve the renderspeed of the dome part considerably 
with a bit of manual bounding. Try adding something like
    bounded_by { box { <-60, 0, -593>, <60, 655, 0> } }
to the declarations of Dome_A and Dome_B in the stpaul.inc file.

(Back in '96, when I made the model, I obviously wasn't as aware of the 
limitations of auto-bounding, as I am today :)

/Ib

Post a reply to this message