Phoenix Drag force / Better turbulence

[ben_hamburg]

I´ve done a lot of R&D for my short film on how to make smoke look like its underwater and two things that I missed were:

1. My first thought about underwater behaviour was to add some drag force. But the standard 3ds max drag force seems to not work very well with phoenix. At least I couldn´t get it to work, maybe its just super sensitive.
So I´d love a phoenix drag force, thats easier to control.
2. I wasn´t quite happy with how the phoenix turbulence force looked. Again: Maybe I wasn´t using it right, but it creates a look with sharp edges I didn´t find natural, I was expecting something that would look smoother and more like standard max noise maps
I stayed away from the standard max wind force because I think it wasn´t recommended, but I usually find the results with standard max wind turbulence and for example particle flow to look more natural.
My dream-turbulence would also give me some option to make the turbulence field even more random, by being able to add a number of different sized turbulences at random positions into one field, kind of like adding several separate spherical turbulence objects with different sizes and strengths.

[Paul Oblomov]

phxmapper + noise

[ben_hamburg]

Oh, well, will you look at that...somehow I managed to completely miss the PHX mapper!
I have a couple of questions though, but thats not for teh wishlist...

[Svetlin.Nikolov]

Btw, we will add more images and examples to the docs for the mapper too

[ben_hamburg]

Ah, that would be great, I still have a couple of questions for that, but I´ll start a new thread for that...

[joconnell]

Oh thanks for this! Very nice that a colour value is interpreted as a vector direction, this and the distance tex might give me something very useful!

[Svetlin.Nikolov]

Mapper + distance tex is a very strong combination. I am thinking about bringing a copy of the distance tex to Phoenix as well.

Cheers!

[joconnell]

I'm going to look into it today as a way of limiting wind to a specific height of the grid. It's very very specific but I want to generate clouds with these qualities - the vertical ones that turn into horizontal streaks at the top:



It's great that vray colour acts as a vector input for direction, can I assume that R,g,b = x,y,z with a positive or negative value meaning left / right / up / down?

Again it's quite a specific thing but I haven't seen many simulators that are able to do "shelter" from forces - say for example you've got a wind blowing but your sim is behind a wall that blocks the wind up to a height. I've tried a very quick setup using particles to push velocities into specific areas of the grid but of course the velocities gradually diffuse out and rather than a sharp directional turn in the sim, you get a gradual bend and influence.

Distance tex fun for me today!

[joconnell]

Yep, mapper and distance tex does it

[joconnell]

Mapper + distance tex is a very strong combination. I am thinking about bringing a copy of the distance tex to Phoenix as well.

Cheers!

Would it be worth putting in a boundary geo object control plus falloff into the phx mapper for this? I've got a sim that was running pretty quickly at 100 voxel resolution but fairly quickly bogged down when I went to 250ish - there's two phx mappers, one that's modifying velocity at a height and another that's multiplying down smoke to fade it off gradually in the same areas!

[Svetlin.Nikolov]

Oh yes, this kind of control is in line with what I was thinking A note just in case - the distance tex is several times faster with V-Ray Next - are you testing with it or with V-Ray 3?

[joconnell]

Yep, we're on 3.6 in work for now unfortunately!

One thing I notice is similar to using a wind force in fume, the mapper seems to fill all the voxels with values right away so it slows things down a fair bit from the start - a lot of artists used particles instead of forces to try and keep their grids small for as long as possible. Thinking about it though my adaptive grid was based on smoke and I had one mapper to influence velocities and another to multiply down smoke to fade it off so it's probably the second one expanding the grid from frame 1!

Do you think it'd be possible to use phxmapper on things like diffusion and drag / buoyancy? Or are those parameters heavily tied in with the fluid equation and more difficult to localize?

[Svetlin.Nikolov]

It would be harder to have a force tinker with the internal workings of the solver such as the advection, conservation and vorticity, but specifically in the cases of diffusion, drag and buoyancy, it would be easier. This also contributes to the list of new additions, together with the bounding volumes for the forces...

[joconnell]

You guys are moving pretty quick