I have a situation where I am modelling 10mm shadow gaps in a large concrete ceiling. The grid that they follow is not regular.
The most accurate way to achieve this is to actually model 10mm joints and cut them out of the ceiling slab.
However, and alternative, and perhaps more powerful approach would be to generate some kind of vector-like graphic (something like Adobe Illustrator would produce), or failing that, a very large, high resolution bitmap image file which could be used to displace the clean geometry at render time using vray displacement.
My experience with displacement, though patchy, is growing, but I believe there are limitations with using very large bitmaps. Vector-based graphics would be better as then even if I render the final image at 10k resolution, the displacement vector map will still hold up to scrutiny. Likewise, rendering smaller test renders at 1k resolution wouldn't slow down as it wouldn't be using huge bitmap textures to derive the displacement.
At the moment, I will stick to modelling the joints, but using a vector-graphic based approach would be far more powerful and flexible if the grid changes - ie I would only need to re-paint the vector map rather than remodel the ceiling.
Does this exist?
If not, is something like this possible?
The most accurate way to achieve this is to actually model 10mm joints and cut them out of the ceiling slab.
However, and alternative, and perhaps more powerful approach would be to generate some kind of vector-like graphic (something like Adobe Illustrator would produce), or failing that, a very large, high resolution bitmap image file which could be used to displace the clean geometry at render time using vray displacement.
My experience with displacement, though patchy, is growing, but I believe there are limitations with using very large bitmaps. Vector-based graphics would be better as then even if I render the final image at 10k resolution, the displacement vector map will still hold up to scrutiny. Likewise, rendering smaller test renders at 1k resolution wouldn't slow down as it wouldn't be using huge bitmap textures to derive the displacement.
At the moment, I will stick to modelling the joints, but using a vector-graphic based approach would be far more powerful and flexible if the grid changes - ie I would only need to re-paint the vector map rather than remodel the ceiling.
Does this exist?
If not, is something like this possible?
Comment