andrel wrote:

>>>   Any suggestions? ;)

>> Matlab?

> no, all floating point arithmetic (unless with special libraries).

OK. I really know nothing about it other than

1. I hate it.

2. It's meant to be fast.

> BTW for a real scene you would never need more than a precision of 
> 1:1e56 (size of the universe divided by the Planck length ;) )

Not in the final scene, no. But the intermediate calculations? Maybe! ;-)

-- 
http://blog.orphi.me.uk/

Post a reply to this message