nice! good to have more videos with production quality fire/smoke , too many users are focused on the liquid part.
Announcement
Collapse
No announcement yet.
TSR stuff
Collapse
X
-
There where problems with underwater foams, they flickered pretty much and thats why I left this RnD.
and even better - render the foam separately and add it in post.______________________________________________
VRScans developer
Comment
-
Did that (disable shadow).
And adding foams in post, I have done so much of research on how to have an underwater foams as they are in beauty render (a proper fog color of the water) AND saving time, but never got the same results comparing the beauty to comp version.
I would love it if it works like this. I select above water from the foam node and render it against black with custom light rig. Then selecting under water and rendering the foams with the water to have a proper fog color.
Every time I miss some foam particles when I compose those two when comparing to beauty.
Anyway, what I dont get is how to render under water foam with fog color in less time than rendering it with the water (which takes always a lot) and have it in separate pass.
I try also try rendering the particles in fog mode but it took even more time.
Im sure I have always miss something here.
Comment
-
Thats it.
If theres no need for the fog color then its like a walking in the park
But like this canal thing, I was inspired in our summer holiday when I saw a canal to fill with water, and the foggy brownish foam looks just so cool I have to immediately make a sim out of it!
Comment
-
First test with FLIP solver, and it rocks like I expected!
Im so happy for the detail with relatively low grid.
Few questions.
As a FLIP, is it naturally adaptive (particles)? I mean, is the overall grid just a "domain" where the particles can travel and therefore be rendered with voxels?
And when my grid box is five thousand million billion, its only the FLIP particles that determines the actual voxels involved in the simulation and theres no need anymore of the voxel adaption in the grid (to save simulation time)?
My units where in cm. Foam threshold have to be under 1.0 to get any foams. Then when I try to get more places for foams to birth, the simulation crashes every time. So could I get the foam threshold somehow around 50-80 more than 1.0-0.01?
Cant wait to have all the other nice features supporting FLIP solver!
Comment
-
i'm glad that you like it
the foam is still adjusted, very probably if you simulate with newer build the foam will be different
"As a FLIP, is it naturally adaptive (particles)? I mean, is the overall grid just a "domain" where the particles can travel and therefore be rendered with voxels?"
actually the FLIP solver is combined, it uses grid based conservation, and particle based advection, so the box can't be removed, but it's much easier to make it adaptive than the pure grid simulator.______________________________________________
VRScans developer
Comment
-
the adaptive grid is still not implemented, but it will work automatically, no channel and threshold value will be used. the big advantage here is that the particles are the liquid itself. there is no convention "this channel value is considered liquid", each particle is a piece of liquid, and if some particle leaves the grid, we just enlarge it, that's all. so, the adaptive grid will have only one option - on/off. well, and camera frustum eventually______________________________________________
VRScans developer
Comment
Comment