We’ve been posting a lot here recently, so here’s yet another update
So we ran out to the store to get us some Leap Motion devices yesterday, and put them to the test today to see if we could come up with a nice showcase that demonstrates integration of Leap Motion with our Tendr soft-body plugin. We think it’s getting better and better by the day, check out the result:
made be laugh :), I now have a strange urge to pet frogs.
Really cool use of the leap! You may want to adjust some of the FABRIK rigs in the character animation blueprint, it seems the shoulders are sometimes a bit eager to move, may be due to how the rig is set up or the default leap offset in the character. Did you capture with the leap facing up or attached to a HMD? If you used it on the table facing up, you can adjust the ‘virtual leap’ in LeapBasicRiggedCharacter to better fit your table offset, see post for details.
Cheers for the mention, always happy to see the yellow guy out in the wild!
I’m interested to know if could be used as a real time solution for foliage interaction (i.e. running through tall grass realistically without a major FPS drop.)
We just want to let you know that we’ve finished our first alpha release candidate (0.8.0) and have begun testing. If you’re interested in testing out release candidate with us, let us know through Twitter or send us an e-mail at info [at] so we can fix an invite for you!
Good suggestion, will try to put that in next time
We would recommend other middleware options for that (Speedtree?), since grass is mostly flat-ish and hardly has any volume for Tendr to be useful. Flat objects can already be done quite well with existing surface-based techniques. You’d have to think more in terms of objects that have some degree of volume, such as frogs, tires, etc.
Don’t worry, we are not dead, the project is still alive We are working with our alpha users to fix some critical bugs right now, but we have limited resources and therefore things are taking longer than we would like. There may be a lack of updates, but we’re still working hard on our plugin
Actually we are working on soft tires right now. There are a couple of issues we need to fix before we can put it up on the marketplace, but we’re aiming for a release of soft tires within the next couple of weeks.
That is exiting! If you do not have a test case for physics-based cars, download the latest version here.
Would be interesting to see how the soft tires behave at higher speed.
1: How applicable could be? (Without a game lagging) Could I have a city full of cars, all with collapsible tyres?
2: Multiplayer? Is it possible that the objects themselves are replicated, or do I need to replicate the collisions exactly? (That would be bad)
3: Could we have an update?
Thanks.
Also, I saw someone mentioned gooey fluids. Just use Nvidia FleX. The source code is open and you could probably make a goo demo right now!
Sorry for the late response, but we have good news. One of them is that the project is not dead We had a bunch of alpha testers that we would like to thank for their input, and now we have decided that a public beta would be in order!
Go to our website to download a beta version of our plugin with the frog model included You cannot generate other models with it but we would appreciate it if you would try it out!
Please do mind the system requirements:
Newest Unreal Engine (4.8.3) launcher version (not the Github version)
NVIDIA GTX400 series or newer (compute capability 2.0 and higher)
I’m trying to make a game with metal vessels that could run into each other. I became interested in the possibility of plastic deformation with the plugin.
Interesting. I’m trying to make soft cubes that the player can assemble into kinetic structures. Can the soft bodies oscillate in volume? Do you expose Young’s modulus or equivalent var?
What about having a skeletal mesh with simulated skin parts.
Softbody would have multiple layers of vertices between animated bones and the character skin.
vertices close to the bones would stick to the animated bone movements.
The rest of the vertices are simulated, with compression, bending / bulge near bones articulations, stretching.
Vertex weight painting would not be needed anymore. The skinning would be automated when computing the vertex layers / volume.
The artist would still be able to disable simulation on some vertex by painting weights or some specific vertex color.
With common softbodies, the softbody object behaves just like an inflated balloon, but skin is not only a surface, there is a volume composed of multiple layers between bones and the outer skin, in the case of internal organs, you could have vertices with various attributes that would mimic the organ’s rigidity/bounciness in that zone of the volume.
In movies like avatar they use muscles simulation with skin sliding, it gives good visuals. I think its called weta digital’s tissue system :
My gtx 970 runs Nvidia’s flex samples easily with 65536 particles of Water for example.
Im sure tendr on skeletal meshes would show some good results and performance that would look similar, maybe better than “Dyna: A Model of Dynamic Human Shape in Motion (SIGGRAPH 2015)” :
