harrye123
10-09-2010, 08:37 AM
Hi,
I'm working on a shot for a film project of an enormous whirlpool which consumes a large area of ocean. Does anyone know the best way to go about doing this?
I've done quite a lot of different tests, mainly using the grid fluid domain method of simulating particles, however each time I have run into various problems. For example, when using the grid fluid domain, once it's meshed there is basically no detail whatsoever on the mesh and when trying to simulate splash and foam in order to add detail, hardly anything simulates because the water particles aren't really colliding with anything. Also when applying displacement to the meshed whirlpool, there is a lot of distortion in the area of the vortex due to the way the uv's for each mesh come out of RealFlow. I also tried using the RFRK to directly render the particles, but I still can't really produce anything very good.
Does anyone have any experience doing this kind of thing? I only really know about small scale water sims and have only used the Hybrido system for the first time on this project. I've avidly worked through all the great Digital Tutors tutorials relating to this topic, but I'm still not totally sure the best way to do this.
...I get the impression that the director on the project is basically expecting something along the lines of this: http://www.awn.com/files/imagepicker/1/p301_Pirates3-Maelstrom.gif
(as he's worked at Weta Digital and Dneg I get the feeling he's expecting something much beyond what I can deliver, so any help whatsoever will be GREATLY appreciated)
Obviously on Pirates 3 ILM had a huge amount of resources to produce their Maelstrom which we simply don't have, although we do have access to the render farm at the University of Portsmouth- so very large scale simulation would be possible if thats the best way to go...
Thanks a lot,
Harry
I'm working on a shot for a film project of an enormous whirlpool which consumes a large area of ocean. Does anyone know the best way to go about doing this?
I've done quite a lot of different tests, mainly using the grid fluid domain method of simulating particles, however each time I have run into various problems. For example, when using the grid fluid domain, once it's meshed there is basically no detail whatsoever on the mesh and when trying to simulate splash and foam in order to add detail, hardly anything simulates because the water particles aren't really colliding with anything. Also when applying displacement to the meshed whirlpool, there is a lot of distortion in the area of the vortex due to the way the uv's for each mesh come out of RealFlow. I also tried using the RFRK to directly render the particles, but I still can't really produce anything very good.
Does anyone have any experience doing this kind of thing? I only really know about small scale water sims and have only used the Hybrido system for the first time on this project. I've avidly worked through all the great Digital Tutors tutorials relating to this topic, but I'm still not totally sure the best way to do this.
...I get the impression that the director on the project is basically expecting something along the lines of this: http://www.awn.com/files/imagepicker/1/p301_Pirates3-Maelstrom.gif
(as he's worked at Weta Digital and Dneg I get the feeling he's expecting something much beyond what I can deliver, so any help whatsoever will be GREATLY appreciated)
Obviously on Pirates 3 ILM had a huge amount of resources to produce their Maelstrom which we simply don't have, although we do have access to the render farm at the University of Portsmouth- so very large scale simulation would be possible if thats the best way to go...
Thanks a lot,
Harry