It’s Bifröst/Naiad - solver wise at least -, it took them 4 years to do it, but it is finally here.
This isnt getting as much replies as I thought it would…
Anyway, this is great news!
Wow, this is big news for 3Ds Max. I’m an avid follower of fluid software, and I was wondering how this is going to affect fluid plugins for max. Are people still interested in third party tools for fluids now that max has a native solution? I know of a few people who were thinking of porting or developing a fluid plugin for max.
Great Job! Thanks!
Maybe people have moved on - they got tired of waiting and already bought (and invested time in) something like Phoenix, or shifted to Houdini for fluids.
That it’s taken that long to implement doesn’t instill confidence that further progress on it will be quick.
Add the spectre of the CEO and his ‘everything will be in the cloud’ and the clobbering of maintenance subscribers.
The last sp for 2018 had 800 bug fixes, maybe people have just stuck with 2016 or 2017?
Take your pick.
Anyone knows what happens with Glu3D plugin?
After the GPU support, sims runs pretty nice, but it seems that they stopped developing it after 3DS Max 2013.
To bad, I’ve worked with it in 3DS Max 2011, and I liked it. It was, of course, far from realflow, but after the gpu support, I hoped it will develop really fast.
As for this new stuff, I can’t wait to try it. I am still using 3DS Max 2014, but my friend have 2018 in his studio.
I haven’t seen in the tutorials videos, but I hope that Particle flow could be used as an emitter.
I’m surprised! Bifrost is a standalone app and it took them ages to write a simple plugin for 3ds max. You’d think that if they were going to do it at all, they’d have done it years ago.
However, now that it exists for 3ds max, how does this help existing users? No modular-updates and subscription-only-model makes it the most expensive fluid sim there is and as Steve said, people either have a plugin such as PhoenixFD or a new app entirely.
Max has built in hair, didn’t stop 2 hair plugins being available.
Also It doesn’t do smoke/fire which PhoenixFD does, although Phoenix could do with Rigid Body sims within it as well.
It’s all good competition.
Spat my tea out when i got the email about the update!
Playing around with it this afternoon, i’m getting a lot of pass through by the particles (doing a simple pour into glass sim) and trying to figure which spinner increases the collision detection.
Works with vray, which is good…and pleasantly suprised at the prt exporter option, so i can try it out with frost meshing. Ironicaly which is how i used to mesh my niaiad sims back before it was bifrost.
Normally increase subdivision or samples per frame
I had the same problem, and as Duncan suggested, increasing “Adaptivity” under “Transport Steps” solved it.
thanks internet people, it was indeed the adaptivity spinner in transport steps. I have had zero crashes so far, and i usually find a way to make that happen. In terms of speed, its not great, i put that down to working within max even though it does seem to be multi threaded! Which brings me to pflow…will it ever use all my cores?
Now why can’t i get motion blur to work with vray?
Decreasing voxel size often helps, Under Transport Steps setting the Min value to 4-6 can often correct that as well for faster moving fluids. Say you have a particle traveling in space, with a min sample of 1 in a given distance the voxel could pass through your collider. If you sampled it’s location in space 6 times for that given distance it is less likely to pass through the collider.
You can try setting Min Time Steps higher or you can also set Max Time Steps higher and increase adaptivity so it can adaptively change between the min->max
Also, scale is really important to have correct…
Thanks Kelly, i’m comparing sims now on scale. Looks like for small scale stuff (wine pour into glass) mm seems to work better than cm.
You can also try turning off “delete exceeding particles” (by default it’s checked). In my test (bellow) I had particles disappear, and I didn’t had water volume at all, until I turned that option off.
Anyone will share some tests?
Her is mine first test.
I’ve tried faking foam by rendering small planes in scanline with velocity channel that drives their size (bigger velocity, bigger particle).
That didn’t turned out very well, but anyway, here’s the test.
here is the direct link: https://youtu.be/nBwuLvcH9ys
- I can’t figure out how to clone solver and cache. I want to be able to simulate once, and then to clone that solver, and render one solver as a water mesh, and second solver as a particles (spheres, boxes, planes…). However, when I clone solver, cache refuses to load. I’ve tried manually find cache, but still, it won’t load.
Cheers to all.
Oh… and btw,. this guy have some cool tests: