Thorsten Froehlich wrote:

> Hmm, now I see your idea behind this.  Unfortunately this won't really
> work because in effectively anything below 0.0000 is pretty much below the
> precision threshold you can reach with double precision floating-point
> numbers for the number of calculations performed by isosurfaces.
> 
My experience show that you actually CAN render isosurfaces with a gradient 
of less than 0.000 . 
The actual output is limited to 3 digits after the decimal point, not 4. 
Using something like %g would be appreciated. 

> I will see if I can come up with something intelligent to at least make
> the message not show up in such case.
> 
No, NOT writing the message is probably a bad idea. Because in my case 
a gradient of 0.01 already takes ages to render and I a using values 
of 0.0002 and the like. 

Wolfgang

Post a reply to this message