zaghaghi wrote:
Can they make clouds with existing blender features?
I think that it's possible that make a static cloud in blender, but not its dynamics.
zaghaghi wrote:
If there is no such a feature or tool in blender that helps peach team, what kind of 3d tool can help doing this easily?
As "Campbell Barton" said, it's better that it be some kind of material. Also I think that some kind of material like Halo, but more Volumetric.
zaghaghi wrote:
Which technologies or methods do you use to create this tool?
I prefer to use a full of point mesh for cloud itself, and use point based rendering(PBR) to render the cloud and use cellular automata on points for its dynamics.
zaghaghi wrote:
Do preferred technologies or methods support cloud dynamics?
Yes, it support dynamic but not as complex as CFD.
zaghaghi wrote:
How much time does it take to render a simple and complex cloud in used technologies or methods?
I don't know
MSc. Student of Algorithms and Computation,
Engineering Science Department, University of Tehran
That would be wicked! Somehow I am imagining defining CA rules in Blender and having fun watching them evolve. However, you probably mean a CA based method for dynamics and Blender already has one such system - fluids. The current fluid solver is CA based so I'm wondering if it could be adapted to clouds.
As much as I would like to, I can't imagine Blender gaining point based rendering anytime soon. However, any proper volumetrics would be a tremendous gain for Blender artists.
As much as I would like to, I can't imagine Blender gaining point based rendering anytime soon. However, any proper volumetrics would be a tremendous gain for Blender artists.
Koba
I dunno, I'd love for someone to explain this if it's so improbable that blender would get point based rendering!
Koba wrote:Now I remember! Personally, I had been hoping that project would become part of SOC.
Anyone know what happened to it? I'm also wondering exactly how it was done - those examples don't look very wispy..
Koba
It looks like those examples lack lighting. That is pretty much the same result I got when arraying X number of planes close to each other and applying a transparancy texture to them.
I already posted this in the comments of the Blog but here it is again the thing I was using. I will keep working on it until it is as good as I can get it. Here is the latest file.
anthoni_c, its nice, but as a simple frame because all frames don't simulate clouds dynamics, but keep going on it, i think of a different particle motion that can simulate clouds dynamics.
MSc. Student of Algorithms and Computation,
Engineering Science Department, University of Tehran
For all the devs here; I wouldn't spend too much time on simulation and scientific accuracy... instead invest time in designing an instrument, a clouds/sky synthesizer with nice and interactive knobs and sliders for an artists to play with.
As for integration: It could be called 'volumetric render object', allowing a user to specify an object (mesh, blob, whatever) with a material, which - when rendered - calls an API function with 3d entry point and exit point. This then can be discretisized (line -> many point samples), or whatever math function you can think of.
AA and sampling then is part of Blender's render system. This approach also lends itself for raytracing.
A few ideas for possible volumetric creation methods (although already a few of the ideas already stated in this fourm). The idea of using a material (like the halo one) to create volumetrics might be a great idea. If you binded this material to particles, then you could use the power of particles to create such effects like smoke and fire and previsualize the movement and speed. And when using it with models, you could easily create cloulds. These are just ideas that someone could work with.
Stop talkin' and start coding.If I wasn't involved in finishing the named layers code I would do it. Time to fire up the compilers and tweaking the codebase.
Q. Where do you want to see this feature in blender GUI?
A. Preferably a material - so you could select between halo and volumetric.
I agree with this. I think making the volume something similar to voxel based and driven by a material would offer the most flexibility. Then anything could be volumetric and the material/texture could drive the appearance.
Q. There are many implementations of Volumetric Clouds, Do you need a special one, or it's not important.
I'd like to see the implementation improve volumetric lighting. Along with providing the ability of volumetric clouds.
Q. Which one is more important? Render Speed or Realistic Cloud.
A. As for rendering speed, as long as its not extremely slow, i.e. doesn't make a complex scene 2x slower or more.
seeing as how much control you get using the method shown in this page: http://www.ecn.purdue.edu/purpl/project ... /index.htm
i cant seem to help but think it would be pretty damn close to what Ton is talking about.
a few months ago, I read about this: http://blog.secondlife.com/2007/05/21/w ... cond-life/
I don't know if there's anything useful at the moment, coder-wise, but perhaps it would worth a glance for ideas, technologies, papers, etc. I didn't delved into myself because I'm not skilled in these fields. In any case, these skies are inspiring! The code for this should be made GPL, but no hint when.
