This is a really dumb question, but I genuinely want to know. Could you create a 3D model of a butt, generate wind coming out of it and calculate the interactions of the various surfaces, thereby creating a simulated fart?
Wow this is really incredible! The only thing this is missing is sympathetic resonance from the rest of the crumpled object. I believe that would make the sound much more natural. Incredible paper!
That's amazing, I hope simulated sounds get more research. Specially on liquids, there's a ton of 3D Physics simulation videos out there that have no sound. Hopefully soon, simulations will come with their respective simulated sounds.
As a foley artist and sound designer, I found the simulated sounds too crispy and thing, lacked low -mid and mid range frequencies with sharp high-mid attacks that dominated. The warmth of the sound in the simulated sounds was gone and sounded very synthetic.
Very interesting project! It is easy to cheat human with another sound, but this can be used to help syncronise sound perfectly or restore sounds from video.
The simulated sounds lack depth and "air". Especially any materials that also form an actual "membrane" such as tin foil and a plastic bottle, lack quite a bit of timbre and tonal details when synthesized.
Wouldn't it be a lot better to simulate the object and use the simulation data instead of inferring deformation from an animation? Either way the object's deformation must be simulated so simulating it while synthesising the sound is a win-win.
*From the paper:* "We seek to automate the synthesis of crumpling sound to accompany a given crumpling animation. The input to our method is an animated triangle mesh produced by some external thin shell simulator."..... "....For this, we target a physically based approach. Physics-based sound synthesis is appealing because it essentially leads to an “encoding” of the physics behind the phenomena as an algorithm, while avoiding the need of recorded sounds" Yes it is physical simulation, not input from rendered image.
im very late to this and dont do 3d art. But i feel like one of the parts to this that make it sound less realistic is that all of these things are LOUD. they sound too quiet
Volume of the object is also important imho. Sound should get more high pitched when the object shrinked.
I love these sound synthesis papers, great tech.
a tad scarry
Fantastic, the plastic bag and bottle were really good but some part in the back of my head kept thinking "I hope they recycle all this plastic."
This is a really dumb question, but I genuinely want to know. Could you create a 3D model of a butt, generate wind coming out of it and calculate the interactions of the various surfaces, thereby creating a simulated fart?
Wow this is really incredible! The only thing this is missing is sympathetic resonance from the rest of the crumpled object. I believe that would make the sound much more natural. Incredible paper!
That's amazing, I hope simulated sounds get more research. Specially on liquids, there's a ton of 3D Physics simulation videos out there that have no sound.
Hopefully soon, simulations will come with their respective simulated sounds.
I can't believe this is a thing. Incredible
As a foley artist and sound designer, I found the simulated sounds too crispy and thing, lacked low -mid and mid range frequencies with sharp high-mid attacks that dominated. The warmth of the sound in the simulated sounds was gone and sounded very synthetic.
Please release it as a software to public, so Sound Sdesigners/Visual Artists/Musicians Could Use It.
This is genuenly amazing. Unbelievable
Just remember that sound is also affected by the medium it's traveling through. But I like it so far.
OMG, this is INSANE
This is so cool, so no more foley sound effects for animated film
amazing, i can't tell which is real
Very interesting project! It is easy to cheat human with another sound, but this can be used to help syncronise sound perfectly or restore sounds from video.
so good
The simulated sounds lack depth and "air". Especially any materials that also form an actual "membrane" such as tin foil and a plastic bottle, lack quite a bit of timbre and tonal details when synthesized.
Its a bunch of garbage with an amazing synthetic sound! :) impressive!!
Wouldn't it be a lot better to simulate the object and use the simulation data instead of inferring deformation from an animation? Either way the object's deformation must be simulated so simulating it while synthesising the sound is a win-win.
*From the paper:*
"We seek to automate the synthesis of crumpling sound to accompany
a given crumpling animation. The input to our method is
an animated triangle mesh produced by some external thin shell
simulator.".....
"....For this, we target
a physically based approach. Physics-based sound synthesis is
appealing because it essentially leads to an “encoding” of the
physics behind the phenomena as an algorithm, while avoiding
the need of recorded sounds"
Yes it is physical simulation, not input from rendered image.
im very late to this and dont do 3d art. But i feel like one of the parts to this that make it sound less realistic is that all of these things are LOUD. they sound too quiet
Teacher: So we all know we have 5 senses
Me: I can hear pictures
Teacher: You can't silly !
Me:
why does the audio sound like its coming through the cans
ok that is good an all but do tell how you animated like stop motion.
theshuman100 move something, take a picture, move it again, take a picture, and do that a few thousand times
theshuman100 thats an computer animation
the plastic bags should make SO much more noise
Can't stop thinking about how AI will use such tools to fool humans.
What is the practical application of this?