For those who did not read my last reply till the end: 

Here is a comparison: 
http://www.cip.physik.uni-muenchen.de/~wwieser/tmp/poviso.html

Sorry, I just notized that I mistyped the time (now corrected). 
It is: 
  Original code: 47.77 sec
  Patched code: 47.89 sec
  Difference: 0.25%
Hence, it can be neglected. 

The reason is that this computation is only done one for each intersection 
and thus not in the innermost loop. 

Wolfgang

Post a reply to this message