So, fiddling with the code came up with this:
I can't really find where the error is placed. I'll
try to cut down the source-code to the bare
minimum which'll reproduce the error with
#QNAN . I've tried using the same pretty close
numbers in the calculation where this error
is returned. In the scene, it uses 0.47... with
pow(0.47.... , 1/.03).
Works fine outside of scene, but not inside of
it.

Only thing I noticed is that when commenting out
an array declaration (trying to fill it with UV-Vectors),
the error is gone.
So, if I do:
#declare Array=array[2] { <1,.03>,<.9,.02> }
everything is working, but with
#declare Array=array[2]
#declare Array[0]=<1,.03>;
#declare Array[1]=<.9,.02>;
the error pops up.

Anyways, I'll post the code once its stripped. If anyone
has got any suggestions what to look for, I'd be happy
to attempt that.

-- 
Tim Nikias v2.0
Homepage: http://www.digitaltwilight.de/no_lights
Email: Tim### [at] gmxde

> So, while fiddling with trace() and objects declared into
> arrays, I've noticed the following: when outputting the
> specular value (asked that in a recent thread), I get,
> when using str(Value,1,20) this:
>
> 1.#QNAN000000000000000
>
> Now, I expect NAN is Not A Number. I'm not too
> sure how this can happen... I'll look some more into it,
> but perhaps some information on NAN and possible
> origins (dividing by something near 0 or multiplying with
> something too large) might aid me in looking for the actual
> error (which might be in my code, can't tell for sure yet).
>
> Nontheless, shouldn't POV have some means of avoiding
> that?
>
> -- 
> Tim Nikias v2.0
> Homepage: http://www.digitaltwilight.de/no_lights
> Email: Tim### [at] gmxde
>
>

Post a reply to this message