When you say that Krakatoa is a volumetric renderer, what exactly does it know about the space occupied by the points? Does it know the relative density of points? It seems to me that all you really need to know is the number of points per pixel in camera space (or in shadow map), but I could be wrong.
This came up in discussions we were having in my studio about what Krakatoa “knows” when it’s rendering and how that can be used to achieve different effects. Like is there any information stored about spatial density, or number of neighboring points, or anything beyond just the per-point data and the transform to the light/camera?
If all the particles have an equal density value, the volumetric density is then purely a result of the number of particles per volume. Krakatoa itself doesn’t know what that density is for any given particle, but it is tuned so that it adjusts multiplicative factors correctly depending on render resolution and distance from the camera in order that the final image is consistent.
The per-particle density modulates that implicit particle density, so the actual density ends up being the particle density channel times the particles per volume.
So two identical objects, one far and one near, would have say 50 points/pixel in screen space for the far one and 29 points/pixel for near one. If things were not adjusted internally, the nearer object would have appeared to be more transparent?
Thank you. The more I get other people to use Krakatoa, the more I have to explain things that I otherwise took for granted. Sorry about the questions, but I need to be able to get everyone else here at Anatomical comfortable.