Thank you jascha, good to know…I will keep my eye on it.
The workaroun of converting to vdb is a no no for me, due to the work behind it, and also quality and renderspeed issues with vdb, and the fact that it will always be non changable after the conversion.
Previous versions of TFD has´t met up to what I require, partly missing the particle advection, but also important things like weight map emission, and no fire and smoke openfl at the same time, that only exists as either smoke or fire.
The multiscattering was very slow and recalculates each little tweak, and I had a lot of crashes when tweaking the shader curves.
Thus so it has come to the point that I´ve been focusing more on blender fluids, which allows for weight paint ( partly also because the weight paint is better in blender) and it can use that as emission source.
Blender also have both fire and smoke presented in openGL, unlike TFD.
Another aspect is that the blender forces that is used to push the fluids, will also push particles, and it will also push dynamics such as cloth or rigid bodies etc, this is not the case with TFD, neither can it´s forces affect lightwave particles, or bullet dynamics, neighter can Lightwaves bullet forces or old legacy forces affect the fluids, so there is a constant mismatch between the Lightwave fluid dynamics, it´s legacy dynamics, and newer bullet dynamics.
This is something you off course have no controll over, just wanted you to know where I see issues, where it in fact is a bit relevant also to workflows including fluids.
That said, somethings TFD is doing better, seems to simulate much faster at higher resolutions, better convectional and turbulence control overall, easier to use textures to controll channels, I especially like the option to use multilayer procedurals with varied modes set, so I can get varied cloud height raising when doing such stuff, this is something that I do not find how to do in blender.
Ive´been writing on the old forums, never got any answers on my later post…so I kind of lost interest and trust in the development of TFD, but glad to see it continued.
It all depends on how Lightwave itself develops, a big part of me that has driven me to using blender more and more, is also the fact I have both CPU and GPU…and the GPU render in blender is quite decently fast,
if it at some point goes out of memory because of to high resolution, short volume step rate, then I just switch to the CPU, the point is that I at least can get fast feedback at a decent volume step with while tweaking the shaders in blender.
Blender is also now in experimental mode with a mantaflow branch, but that one I am not comfortable with, difficulties clearing cache and some other issues.
So…still a little interest of TFD, my first points would be…
- Volume render engine PBR.
- Weigth emission.
- Fire and smoke at the same time in openGL
- Particle advection.
- No crashes while tweaking curves for the shading.
That´s my input.
Good Luck on the development.
/Michael