On 12/7/2016 8:41 PM, Mike Horvath wrote:
> On 12/6/2016 11:55 PM, clipka wrote:
>> Am 06.12.2016 um 08:05 schrieb Mike Horvath:
>>> I am having trouble rendering the attached scene without artifacts and
>>> gaps in the isosurface. I already increased the scale of the scene by
>>> 200 and the max_gradient of the isosurface to 100. Anyone have any tips?
>>
>> You seem to be running into /some/ kind of problem when the RGB values
>> get excessively low.
>>
>> Clipping the C parameter of the fD function to some near-zero negative
>> value seems to do the trick:
>>
>>     #declare fClip = function(X,A) {select(X-A,A,X)}
>>     #declare fD = function(C) {abs(fClip(C,-0.1)-0.5)-0.5}
>>
>> This lets you get away with a max_gradient of 20.
>>
>> (You still have a coincident surfaces problem between the isosurface and
>> the box, but I guess you can figure that one out yourself.)
>>
>
> I changed the camera angle and the problem reappeared.
>
> camera
> {
> //    #local cam_distance    = 40;    // render at 8192x8192
> //    #local cam_planesize    = 4;    // render at 8192x8192
>     #local cam_distance    = 17;
>     #local cam_planesize    = 1.7;
> //    #local cam_distance    = 20;    // render at 4096x4096
> //    #local cam_planesize    = 2;    // render at 4096x4096
>     #local cam_aspectratio    = image_width/image_height;
> //    orthographic
>     location    -z * cam_distance
>     direction    +z * cam_distance
>     right        +x * cam_planesize*cam_aspectratio
>     up        +y * cam_planesize
>     rotate        +x * asind(tand(30))
> //    rotate        +x * 90
>     rotate        -y * 045
> //    rotate        +y * 180
>     rotate        +y * RotateAmount
>     translate    1/2
>     scale        ScaleAmount
> }
>
> max_gradient is set to 200, and I am using your revised fD function.
>
> Mike
>


And seen from directly above the isosurface disappears completely.

Mike

Post a reply to this message