|
 |
Don't know if this is a bug or not, but something odd is happening.
Got an isosurface generated from a jpeg image (using Christoph Hormann's
'earth_simple_free.pov' from http://www.imagico.de/pov/earth_scene.php). Just
trying to check how it renders, so I'm using KF3, KFF180 (60x3 frames) rotating
360*clock*axis - changing the axis >2=y, >1=x, else z.
Each time I render the isosurface, I get a max_gradient bailout. Initially this
was set at 1.2, but after the first couple of frames I got a warning that the
max_gradient should be 45. I set this and restarted the render... the same
thing happened, stating that the max_gradient should be >105... /on the same
frame/. Then, it states that the max_gradient should be >199!
Any ideas on what's going on?
Post a reply to this message
|
|
|
|
Am 16.04.2012 13:20, schrieb Al Feersum:
> Don't know if this is a bug or not, but something odd is happening.
>
> Got an isosurface generated from a jpeg image (using Christoph Hormann's
> 'earth_simple_free.pov' from http://www.imagico.de/pov/earth_scene.php). Just
> trying to check how it renders, so I'm using KF3, KFF180 (60x3 frames) rotating
> 360*clock*axis - changing the axis>2=y,>1=x, else z.
>
> Each time I render the isosurface, I get a max_gradient bailout. Initially this
> was set at 1.2, but after the first couple of frames I got a warning that the
> max_gradient should be 45. I set this and restarted the render... the same
> thing happened, stating that the max_gradient should be>105... /on the same
> frame/. Then, it states that the max_gradient should be>199!
>
> Any ideas on what's going on?
It's hitting some part of the object where the max_gradient is
unexpectedly high.
In the first iteration, it was only able to determine that in order to
reliably render that detail of the object correct you would have to
specify a max_gradient of 45 /or higher/.
In the second iteration, the "max_gradient 45" increased the precision
of the computations, enabling POV-Ray to find out that actually 45
wasn't enough either, and that 105 would be closer to the mark.
My guess is that it's hitting some singularity there (at the poles
maybe, or the seam where the left and right edges of the image meet?),
where the gradient of the function (i.e. the derivative) approaches
infinity, so the max_gradient will always be too low, no matter what
value you choose.
You can probably ignore the warning, unless it occurs in significantly
more than half of all the frames.
Post a reply to this message
|
|
|
|
clipka <ano### [at] anonymous org> wrote:
> Am 16.04.2012 13:20, schrieb Al Feersum:
> > Don't know if this is a bug or not, but something odd is happening.
>
> It's hitting some part of the object where the max_gradient is
> unexpectedly high.
....
> You can probably ignore the warning, unless it occurs in significantly
> more than half of all the frames.
Brilliant. Thanks - didn't know if it was something that had been introduced.
Currently using RC5, but my last attempts at doing stuff with isosurface mapping
was around 10 years ago... with 3.1... so effectively I'm a noob. Ho hum.
Post a reply to this message
|
|
