Stanford Invented The Ultimate Bouncy Simulator! 🏀
Embed
- Published on Jan 7, 2022
- ❤️ Check out Lambda here and sign up for their GPU Cloud: lambdalabs.com/papers
📝 The paper "Bounce Maps: An Improved Restitution Model for Real-Time Rigid-Body Impact" is available here:
graphics.stanford.edu/project...
📝 The amazing previous works:
- Input video, output sound - www.youtube.com/watch?v=kwqme...
- Input sound, output video - www.youtube.com/watch?v=aMo7p...
❤️ Watch these videos in early access on our Patreon page or join us here on TheXvid:
- www.patreon.com/TwoMinutePapers
- thexvid.com/channel/UCbfY...
🙏 We would like to thank our generous Patreon supporters who make Two Minute Papers possible:
Aleksandr Mashrabov, Alex Balfanz, Alex Haro, Andrew Melnychuk, Angelos Evripiotis, Benji Rabhan, Bryan Learn, Christian Ahlin, Eric Martel, Gordon Child, Ivo Galic, Jace O'Brien, Javier Bustamante, John Le, Jonas, Kenneth Davis, Klaus Busse, Lorin Atzberger, Lukas Biewald, Matthew Allen Fisher, Mark Oates, Michael Albrecht, Michael Tedder, Nikhil Velpanur, Owen Campbell-Moore, Owen Skarpness, Peter Edwards, Rajarshi Nigam, Ramsey Elbasheer, Steef, Taras Bobrovytsky, Thomas Krcmar, Timothy Sum Hon Mun, Torsten Reil, Tybie Fitzhugh, Ueli Gallizzi.
If you wish to appear here or pick up other perks, click here: www.patreon.com/TwoMinutePapers
Thumbnail background design: Felícia Zsolnai-Fehér - felicia.hu
Wish to watch these videos in early access? Join us here: thexvid.com/channel/UCbfY...
Meet and discuss your ideas with other Fellow Scholars on the Two Minute Papers Discord: discordapp.com/invite/hbcTJu2
Károly Zsolnai-Fehér's links:
Instagram: twominutepa...
Twitter: twominutepapers
Web: cg.tuwien.ac.at/~zsolnai/ 
Science & Technology
This channel is the most interesting and encouraging channel I've ever watched. I'm a cg artist and I love watching this channel for inspiration and foreseeing the future of our industry :)
@PatchCornAdams723 Shit, you got myself and a couple billion others that can't even do that so you're at least you're somewhat ahead of the game. I modeled a bong in Blender about 20 years ago though if that counts. 😆
Friend I have been messing around on C4D for 3 years now and I still can't model anything organic, only hard surface models.
It's so sad :(
W
It's like watching the future unfold before it gets here. It's gonna be so cool in the next coming years to look back on videos of him saying, "I can't wait to see how much these papers will develop in the next few years." a few years from now and seeing what the future holds. Because let's be honest, you know you can't help but think how cool this is going to make video games in the future. Thanks again Doc.
My perspective on this channel changed when I started applying to college lol
I can see this work further down the line for bone simulations as described towards the end mixed with simulated muscle and tendons.
Hey there Steven, check this out, and thank you so much for your generous support! - thexvid.com/video/higGxGmwDbs/video.html
This is actually extremely interesting, I have thought about generating sounds directly inside an engine, but I've always thought it was simply impossible to do
@Mika Korhonen still tho I really like whatever is going on with this study
@Mika Korhonen Sound waves can most certainly be simulated with ray tracing
@Loljptrollergami I don't know what raytracing has to do with sound waves.
I imagine that main simulated thing is vibration frequencies of the object. You need to know material and shape of the object. Object transfers its vibration in to air movement and you hear it as a sound.
@Mika Korhonen cause I thought it would be having too many parameters to take into account, but at the same time, now that raytracing is a viable method, it doesn't sound too out there to think of raytracing applied to sounds
I once tried something like that by following all object coordinates from a physics simulation and whenever three consecutive positions could neither be explained by 9.81m/s² gravity acceleration nor by resting (e.g., on the ground), I'd add a short sound, even at the correct stereo position. In this simple manner, it didn't work too well (how loud should it be? Should it be a "bing" or a "bong"? Was that really tiny bounce or just air resistance?) Ultimately, me manually picking points in time when single more prominent bounce happen and using noise-lie sounds of frequent bounces (as in pouring pebbles) in the background sounded (somewhat) better.
Guys... real life bouncy castle SIMULATION.
What a time to be alive.
It's been a few years since I read the paper, but weren't the bounce maps primarily about taking into account the shape of the object in the restitution model via FEM simulation done as a pre-process instead of artist control?
This. Those "crazy" versions looked sooo much more realistic! I want this in my physics engine, and it seems like it wouldn't be all that difficult to sample the bounce-map performantly. Will definitely be giving this a look.
My thoughts exactly, it seemed much less about "designing crazy hockey sticks" rather than faithfully modelling the bounciness of an *actual* hockey stick (which obviously varies depending on the contact point).
There is so much creativity in the world! Feels great to see such amazing papers especially on Two Minute Papers 😄
that audio to video simulator is insane. just imagine the possibilities of that technology with forensics. with just audio data, information can be virtually deduced about falling objects
@Bradly Cellini it's nonetheless an interesting avenue of thought that a sound could be analysed by a system to detect and infer any particular physical action using just the sound..
If It can attempt to calculate what a particular size Lego brick sounds like, it must be comparing the sound to it's bank of other 'known' sounds and scoring them based on similarity.. So if it had a huge bank of known sounds it could potentially be viable to predict wether something was a firecracker compared to a car misfire or a gunshot - although the crossover is high.
The quality of microphones and other audio artefacts are probably a problem so large, that a system like this could never be used as evidence, but it could surely be viably developed as a forensic/investigative tool/cue for deducting clues from audio alone.
I don't know a huge deal about current forensic/detective technology - but it's super interesting nonetheless..reminds me of some black mirror type plot
Also awesome video as always - Keep up the awesome papers
I would be careful about that assumption, because I imagine there are a lot of ways an object could fall to give the appearance of matching the sound. The results are very impressive, still. Especially being hand-crafted.
Ah! This could be very useful to me, I'm developing a bouncing object with some very specific counterintuitive behaviours (specifically rebound angle), and this would make conceptual testing of different configurations very easy!
Oh my God, I forgot to hold on to my papers, and they all flew away when you said the sound-to-animation was all done without machine learning!
Greate vid as always. Anyways I need this for my virtual bouncy castle experience
I for one am happy to see videos about older papers like this. Thanks for doing your part in trying to introduce a wider audience to the work!
Köszönjük az egész éves munkádat 2021.ben! Csak így tovább 2022-ben is, és még tovább! ❤👌👍 Nagyon jó lett a videó.
🙏
i really love this channel, although i would also appreciate a little more non-physics-simulation based papers
you could also take the transients of the object impact and then re sequence the sound to match any new animation, because kinetic energy decreases over time, and it's already sorted in order of impact force.
Would be nice to see how it behaves with high frequency impacts, like a spinning disk or a playing dice
I actually always want to ask a question, which simulator engine normally the engineers use to create such amazing simulation?
I love that if you see something cool that you missed you still share it. Love it. Can't wait to see this kinda stuff implemented, makes me want to do some of it myself
I love these physics experiments!
could be nice if you could run it backwards, take a video of an object being hit at various points to create a bounce map, ideally predicting bounces for areas not seen
I want to know more about the audio generator. I didn’t think that technology existed yet.
This is amazing... i wonder how this would do with that other project that tear materials, i imagine it with breaking glass and it falling on the floor or breaking in a bouncy surface
Now just add this to the current fracturing, softbody, and liquid physics engines and you’ve got yourself the ideal simulation engine!
Is there an interesting scientific reason for simulations always looking like they’re in slow motion? Just curious :)
Things fall really fast in real life, so by slowing stuff down, you can see the detail of the physics more, rather than them just flying by our eyes without being noticed
Played back in slow motion, ie. 30fps playback of 60fps footage . Them being hard to calculate is not inherently related to the frame rate. It could also be that things are larger or smaller than you think they are. Or that gravity is weaker in the simulation.
Very high cost to compute them.
Awesome video 🎉🎉How did they calculate that each object should be bounced and flipped for these many number of times or iterations ?
I love bouncing things, so i really love this paper
Imagine there was a game using all of the technology two minute papers have shown in this channel
I hope this video inspire researchers so they work on a new more accurate method regarding this topic
I seriously want Bounce Maps in game development now.
Do these get released as add-ons to Blender? Then artists can experience.
Wow that is truly Amazing!
the possibilities for creating sound to match video (thexvid.com/video/55PJtqpXAm4/video.html) and video to match sound (?t=78) are really quite amazing!
!!! This is insanely brilliant.
Man, I dropped the CS348c class with prof Doug James this quarter. Didn't know he did such amazing work, will definitely take the class with him later on.
Great job man.
Im not convinced yet with these works, they bounce more than they have to
Some shit never gets old. 👍
This is why your are the best channel for information the public on things they need to know! Good job! Sponsor
Wish i could use this stuff in softwares like blender.
Blender could use some better simulation based on this kind of research. Someone should tell the devs =)
I think that discoveries made through papers like this can eventually filter through into addons for Blender.
Reminds me of "More Bells and Whistles" demo from back in the day.
You had me at "Let's see that bouncy knob"...
Videogames need more real physics
I wonder if the dead spots on the bounce maps for the sports sticks are engineered that way IRL to reduce the amount of vibration transmitted through the stick to the hands?
This team could simulate sounds hitting the small bones in the human ear. An accurate and high-fidelity simulation of that would have very significant implications for the medical field. Could use it to study hearing loss, hearing damage, possible materials to use for a surgical replacement, etc.
What a time to be alive!
The sound animation algorithm blew my mind.
You literally sound like a mad scientist hahaha. Brilliant XD
which programming language or framework is used to make these animations?
I was wondering the same thing
Thank you for letting us know!
I am not an artist, I know only the basics of solid mechanics before I took my engineering degree to other directions. But I gotta say, all of this used to be bleeding edge structural engineering. It took an engineering PhD a lot of time and creativity to create a feasible numerical model where useful predictions could be made, and only then could designs be validated against.
It is awesome to know technology has advanced to a point where we can create simulated solids with variable parameters and run simulations over then for the visual effects.
Nowt the knowledge is bouncing around in the internet
The sound to video is unbelievable!
Now get this shit to game devs and animators in general, and let's end voice sync issues forever, especially considering voice sync should've been solved eons ago.
your channel keeps me going
Interestingly, warm water is less viscous than cold water, so sounds different.
the 2014 paper is amazing
"what a time to be alive"
Is the software available for download?
I don’t fully understand why the bounce maps are impressive. Couldn’t you just take an ordinary physics simulator from a game engine and calculate the bounciness of a contact based on the location of the contact and a texture. The simulator that they applied bounce maps in seems impressive though, given it isn’t simply a normal simulator with a really small time step.
"that is about to stop bouncing" I think.
But great
Amazing paper
Nice video ☺️👍
so cool love it
It's only a matter of time before this is incorporated into a Skyrim jiggle physics mod.
This channel is what saves mankind...
So much useless trash on TheXvid that these channels need to be mined
😏 Adult games already are benefiting from this technology.
okay, but can it animate squahs and stretch??? no??!?! that's what I thought. #2D>3D
at 1:55 -- that's just nuts!
this is awesome
If not the, one of the best channels ever
as early as 2014 this is magic
So much creativity
1:48 "it can even shoot a music video for you! (links are in the video description)"
Where are the links to the music video?
Sure. Recognize Stanford in the title. Smaller institutions like kudos too.
How can I use all of these nice tools for *creating music videos* & other artistic outputs? Can someone help me here?
3:08
"And now, let's see that bouncy knob."
- Károly 2022
(I'm so sorry)
This will not at all be used for some features on female videogame characters.
Thighs.
lmaoooo
I came here to listen the first line ❤️😀
does this mean what I think it does? 😳
This is not a very impressive one at all, we were doing this many years ago in the Bepu physics engine.
insanity for 2017
Did this guy accidentally put too many full stops and comas in his script
This is the choice of Steins Gate.
So why do we need that?
*B O U N C E*
moar bouncy sim. yess please
Thank you and thank you Wang et al. (2017) ☺️💫
PEOPLE HAVE TO KNOW !!!!!!!!!! Thanks!!!
1:08
Woah
wow
I don't understand the point of this variable bounciness.
I'm disappointed that of all the bouncy things they have tested, they did not test the bouncy thing that we care the most
@Wout Motmans if you dont know what he is talking about, you are probably too young for that
Huh, I'm curious of what bouncy thing you are thinking of?
Is it really necessary to fluctuate every syllable of every word like that? It feels like every word is a sentence on its own. It hurt my brain.
And...
6th
am I the only one who finds the voice over so annoying that I rather watch these without sound and try to guess than listening to him?
Im the opposite really, sometimes I just listen