Tweet
Hi,
Not sure if this is an "issue" or "general" since it may be expected behavior.
I would like to make a rendering with the effect seen in Anthony McCall's "Light Describing a Cone" / "Solid Light Works".
In real life, this is achieved with a projector+fog. The projected image is just a white line on black background; the conic projection of the white line creates the visual effect of a "wall" in space.
In Rhino/VRay, I've set this up as a spotlight with no penumbra, high intensity (e.g. 200), atmospheric affect scalar turned way up (e.g. 100), environment fog on (distance set to e.g. 1000). I've placed a plane with the "light path" cut out directly below the spotlight to emulate the effect of the film image (the reason for doing it this way is so that I can actually animate that path cutout in grasshopper, but that's irrelevant to this question; an alternative to this which I will also explore will be to use a generic material with an opacity texture map to let the spotlight through, but also irrelevant to this question).
Anyways, the effect is working well enough, as seen in the attached image - CPU rendering.
There are some artifacts in the upper half of the image as well as some random boxes which seemed to not render, but I think I can probably get those kinks ironed out myself.
When I tried to switch to CUDA or RTX rendering, the effect is completely ruined - see attached (though render times are much faster - ha!). I understand that it is expected that there will be differences between RTX and CPU rendering, but this seems outside of what I expected; then again, this is a pretty non-standard use case.
I tried tinkering around with some of the lighting settings, but nothing had a significant effect. I figured I would come on here to check to see if this major difference would be expected.
The issue may also be due to the size of the cutout I am using for the light to pass through (see below). It is currently set incredibly small now that I look at it - probably on the order of 0-1000nm. I can actually move the masking plane/cutout further away from the light source so that the cutout can be bigger while still getting the same final line size (needs to be only a few mm when it hits the ground). Haven't tried this yet, will try it next. I suppose using a texture with an opacity map for the masking would similarly solve the issue, assuming that is the problem...
Rhino file here
Not sure if this is an "issue" or "general" since it may be expected behavior.
I would like to make a rendering with the effect seen in Anthony McCall's "Light Describing a Cone" / "Solid Light Works".
In real life, this is achieved with a projector+fog. The projected image is just a white line on black background; the conic projection of the white line creates the visual effect of a "wall" in space.
In Rhino/VRay, I've set this up as a spotlight with no penumbra, high intensity (e.g. 200), atmospheric affect scalar turned way up (e.g. 100), environment fog on (distance set to e.g. 1000). I've placed a plane with the "light path" cut out directly below the spotlight to emulate the effect of the film image (the reason for doing it this way is so that I can actually animate that path cutout in grasshopper, but that's irrelevant to this question; an alternative to this which I will also explore will be to use a generic material with an opacity texture map to let the spotlight through, but also irrelevant to this question).
Anyways, the effect is working well enough, as seen in the attached image - CPU rendering.
There are some artifacts in the upper half of the image as well as some random boxes which seemed to not render, but I think I can probably get those kinks ironed out myself.
When I tried to switch to CUDA or RTX rendering, the effect is completely ruined - see attached (though render times are much faster - ha!). I understand that it is expected that there will be differences between RTX and CPU rendering, but this seems outside of what I expected; then again, this is a pretty non-standard use case.
I tried tinkering around with some of the lighting settings, but nothing had a significant effect. I figured I would come on here to check to see if this major difference would be expected.
The issue may also be due to the size of the cutout I am using for the light to pass through (see below). It is currently set incredibly small now that I look at it - probably on the order of 0-1000nm. I can actually move the masking plane/cutout further away from the light source so that the cutout can be bigger while still getting the same final line size (needs to be only a few mm when it hits the ground). Haven't tried this yet, will try it next. I suppose using a texture with an opacity map for the masking would similarly solve the issue, assuming that is the problem...
Rhino file here
Comment