Wow. I like how the brevity of your statements completely hides the massive amount of work and inevitable frustration that must have gone on... transducer polarity, code, FPGA... each of these parts represents a lot of work. Well done.
I felt quite the opposite. I could see every detail of his hard work to the point where I got tired just listening about it :p but the result is worth it!
As a computer engineering student, the second I saw those Modelsim waveforms was the second I realized what level this project is truly on. This is absolutely insane and next level.
my lab is like this: i get 1/4 of that 6:50 waveform, but with no suggestive names and i should figure out in 2 hours what it all does(i need to learn how a MIPS32 processor works). I kinda managed to figure out what memory and registers do and edit them a little to see if my understanding was correct, but then i had a lab about a fraction of a pipeline, i couldn't understand a thing. i am looking forward for the times when I need to do the code myself and simulate the waveform to my willing to fix my projects.
@@joshuapeligrinoWell, sound is just pressure waves, so you technically can change the speed and direction of light using sound to rapidly increase or decrease the pressure in an area. Not sure you would be able to meaningfully effect light this way, but it's an interesting thought. Would probably also explode when doing this, cause the difference between a sound wave and a shock wave is just the amount of energy involved...
I've seen demonstrations that look like an actual physical object. The way it works is that a mirror is placed a a 45 degree angle, and it's rotated very quickly. Above (or below) it is a fast screen that changes the images rapidly, so you'll get a different image depending on the angle you are looking at it from. That was over 10 years ago. There's really no commercial application for these displays. They are novel, but that's it.
There's a WAY easier method if you only need to create a hologram in the sense of a volumetric display (which is how it's shown in Sci-Fi). You can just shoot two infrared laser beams into luminescent smoke, and where they intersect, they create a dot of visible light (since only their combined energy suffices to cause luminescence). This way, you can easily build a whole hologram. It's a well known technique for at least 20 years now. The only reason why we don't use it more often is because there are very few applications where it actually makes sense. What's so impressive about the project in this video is that it's using an actual physical object.
@@joshuapeligrino Even if this is possible, if we trapped the light in the sound waves, it wouldn't be able to enter our eyes and we wouldn't be able to see it!
I understood 0% of this, but what ever your doing don't stop, the world needs more people like you. if you could throw this together, I doubt there's not much you can't do. let the mad scientist take over.
The way you just easily say: yeah, hook two FPGAs together, just program their memory, attatch an EEPROM hook it together over SPI with a Raspberry Pi to control the 100 I/O lines for my 100 sonic transducers shows how much you actually know that you fly over those topics. Incredible! Loved it, thank you very much. :D
@@abitembedded I think you just need to adjust the frequency - Test the effects on a larger dot. Also, look for a means to hold the actual device, consider sound canceling. Funnel a single point, try salt, molten, I wanted to see your process to make a line used to make a cube.
This is pretty incredible. An actual 3D image made with sound and one little ball. In colour. Quite amazing. I wonder how complex you can go once you figure out multiple balls.
I'm 2 years too late, but this is absolutely incredible work man. Huge props to you for making all the design files available online!!! :) Are you aware of anyone else making one of these? Did you ever consider selling kits at cost to science museums around the world. This is totally amazing work you've done putting in all this leg work :)
Thanks for the kind comments! I know of one other group that is currently building one of these based on my design. I've considered making a kit but the amount of work to do so is daunting and I have so many other projects to work on (most of which don't hit TH-cam :) )
@@abitembedded I understand, I would love to see more videos from you about the projects you do now, you have an amazing amount of skill, I can't imagine how much one would learn from just looking over your shoulder as you built this stuff, the FPGA design alone let alone all the hardware, and the GPU simulation. Your a one man show :)
This honestly has to be one of the coolest projects I've ever seen documented on TH-cam. I imagine it's only a matter of time before some company comes in to sell something like this, just because of how awesome it looks.
This is incredible. The digital design alone is an impressive feat, but interfacing to the array and your pipelining of the math is really stunning work. The end result is incredible!
A really cool thing to "draw" with this would be a Lorenz Butterfly. Instead of animation frames, simply make the ball follow the trajectory as quickly as possible. I know this is a year old, but it would be cool to see if you can make it work.
This is one of the very top most impressive maker projects I've seen on TH-cam from an apparent complexity standpoint. This is from someone who has spent a huge amount of their free time finding the best ones instead of my own projects. :) I bet the video of the butterfly animation doesn't do justice to seeing it in person with real eyes.
This is just stunning and an insane amount of work across a variety of subdisciplines: fpga, firmware, electronics, layout, fabrication, the list goes on. How long did this take you?
@@abitembedded I really need to know your background and education. You're doing things I dream of doing but are so interdisciplinary I don't know how to approach. Please answer.
volumetric displays really do seem like the future. and the use of sound for it is amazing, imagine this on a huge scale, that comment at the end where you talked about how it can stimulate your touch nerve endings, imagine the VR capabilities, you could actually put a car right in front of somebody half the world away, and they could touch it. Could also be used for disabilities, imagine a blind person that doesn't need a walking stick because their entire surroundings is projected onto their hand, or a deaf person that has a persons speech turned into a touch map. taking a research paper and doing it yourself is incredible
I imagine the biggest constraint to this method would be collision. The position of the number of balls would have to be shifted such that they never overlap the same X, Y, Z. This wouldn't work for more elaborate 3D models because there is only 1 dimension of 2 forces (gravity vs pressure on the Y axis).
@@Mildain2000 Just the refresh rate alone would be an issue even if it was possible. You're constrained by the physical speed. This is why volumetric displays will always just be things like LED's layered together. But volumetric displays are honestly not that cool for viewing anything but basic shapes, because they obviously have a limited depth, and images of things will have near and far depth of fields, and focused or blurry details. So you'd really only get a narrow range of depth and then it'd just be a cut off to become a flat wallpaper of the rest of the things in the background. This would pretty much break immersion. About all a volumetric display would be good for is things like the medical field where you want to view sections with physical topology, but you can already get that in VR or AR. For home use, it would only ever be good for giving UI elements a pop out effect. Another issue with volumetric displays is working in the 3D space itself. A simple mouse and cursor won't do. Even a touch screen wouldn't do, because eventually you'd have elements overlapping, so you'd still be limited to a 2D field of selection. Needless to say, we're not getting volumetric displays any time soon, and we'd be pretty disappointed in the practicality of it anyways. The only ones that make sense are holographic displays.
Insanely well done. I wonder if you could animate something like 3D pong, as if the ball is a mime between invisible paddles. The ability to get a couple extra spheres in there to simulate paddles would really complete the illusion. But so impressive as is!
Would be amazing but I worry about what would happen as the ball approached the paddle. Would they start to interfere? I guess just give the objects a huge hotbox.
I just had to come back to this video because it’s been living rent free in my head for months. Still probably the coolest and most impressive project I’ve seen on the site. I really think this technology can and will be used. It really is inspiring. Well done
The simulation is like the most important part of this whole project. I can't do that. Without it he could have spent months or years trying to figure it out maybe, just because of the time involved making and changing it physically vs just simulating it and knowing it will work.
This is what I was thinking. I was like, alright physics, I can keep up with this, simulations, alright yeah I kinda understand, fpga, I know what that is, 6:46, jesus christ what does this guy do knowing this much about so many different disciplines, what the hell is this
"I can't do it realiably and don't have the patience to further tune the system" Everyone reaches that point someday. I love the honesty and laugh at thr thought of reading that in a paper.
It’s amazing how this video has taken things I have learned from three different subjects this year, combined them, and made me think with them in a way I never had before. Now I’ll be able to hold physics, digital electronics, calculus, and computer science in a whole different way now that I’ve seen the connections between all of them drawn together in one video. Thank you.
What do I think? You have that rare combination of genius and an amazing work ethic! Modern technology takes a place as well of course. Mail-order PCB houses are another part of that phenomenon. Amazing! It would be great if you could juggle a couple more balls, and therefore allow more lines to be drawn, right? Would the maths be significantly more complicated? Could the hardware do it? Could you perhaps work an entire sheet of particles as a 2D plane that moves up and down? A little bit like those 3D displays that use a projector, and a screen that moves back and forth very quickly? Perhaps a video projector could replace your LEDs to colour a lot of particles at once? You could perhaps sell these to science museums, at a significant markup, if you wanted to be paid for your time. It's an example of something spreading from a lab paper into the hands of the public in one step, over the Internet. Back when we were idealistic about the Internet's use for democracy it would have been something geeks would have been proud of.
Thank you! Interestingly, the math to move more balls doesn't seem to be more complicated than adding together the "waveforms" for moving individual balls - certainly on my list of things to do...
@@AndroidFerret It's all about time :) Given all the recent interest, I think I have to prioritize coming back to this project. But I do have another project on the go right now...
This is just purely Magical. Yes, I understand 80% of everything that's happening, but it's amazing to see a single person finishing everything from acoustics to physics sim to coding to hardware programming etc.. It's just awesome!
no bad apple. 0/10. in all seriousness, this is absolutely amazing. just the fact that acoustically trapping particles is possible is crazy enough on its own let alone being able to move them around fast enough to get some - admittedly kinda janky - animations!? huge kudos to you, i cant imagine the amount of work hidden behind the veil of this simply-explained 10 minute youtube video.
any sufficiently advanced technology is indistinguishable from magic. I'm just about the most scientifically minded person I know and this is basically wizardry to me
Smoke/ aerosol particles might be interesting to observe in this. might need extra some phase cancellation in negative space to clear out the image tho
I've just stumbled upon this video and find it extremely fascinating. This is the first step into modern-day holograms for anyone unaware. The potential behind this technology is amazing, and I am very excited as an Industrial Designer to see the evolution of this tech. Great video and I would love to see an update on the multiple balls theory.
This is the first tech i have ever seen that cold evolve into a star trek style holodeck. Sound waves we can't hear manipulating the positions of possibly trillions of tiny particles to create the illusion of larger objects. AMAZING!
Alright I’m calling it: within 10 years, this will be the next technology used in user-interaction interfaces. Touch-screen will be either a dinosaur or obsolete.
This is incredible, the amount of time and skill it takes to program an FPGA let alone just do it for a project is amazing! Great Job! I know I could NEVER do that lol
The technical stuff went in one ear and out the other but ooo ball go floaty real fast! If you could physically feel those focal points that would be so cool and weird. I wonder if someday that would be used in VR to make a digital world you could physically interact with
@@abitembedded I didn't just mean it, I did it :) I'm an engineer too and you are deserving all the appreciation :) I also do appreciate your time to describe all your work in a video and inform people about such a thing. Thumbs up :)
i love how this is actual physicalised imagery using sound to "draw" one omnidirectional particle, effectively emulating the photon. nice touch with the RGB colouring as well, very impressive!
Are you telling me you can make it so that you feel that butterfly with your fingertips when it ISN'T EVEN THERE?!?! This is some cool stuff man!!! Please, keep doing this stuff!!
This is absolutely amazing. At the end you said you could feel it at targeted locations within the array. I can imagine hundreds of applications right now. Hopefully this is something that can be commercialized and improved upon to provide resolutions imperceptible to human touch much like we have done with screens today.
Very impressive, Dan. Some similarity with what I did in the 90s in two dimensions with a laser dot on a screen for animated images. The mechanical scanners were critically damped to move from point-to-point in 4ms. Images (vectors) were stored with x and y coords and the mechanical scanners and mirrors "drew" the connecting lines. When doing the real-time rotations of an image, I calculated the Z coord and used intensity modulation to dim more distant parts for depth cueing as well as making distant parts smaller, simulating perspective (that was a simple, but very effective cheat saving division to change size). . I even wrote my own multiply routines because the MUL instruction in the 6809 processor was slower than my coded routine. In an 8 bit machine, I used a 256 value angle look up table which is about 1.4 degrees I call the "Bi-gree" for Binary Degree. This allowed continuous rotations with an 8 bit pointer that auto wrapped around saving code. So many neat tricks made it work really well. Cheers.
Very neat - thanks for describing the methods you used. I truly miss the times when you truly had to understand how a processor worked and had low level access to the registers - give me a 6502 or 6805 anyday :)
This is absolutely insane, I love you calm casual tone while walking threw hardware and software design that would make most people's Brains melt. Terrific work and thankyou for presenting such an awesome project, I wish you luck with any further refinements and iterations!
Seeing these amazing projects by really smart people like you is what makes me realize I'm still far away from what I could be doing, even now after graduating as a comp sci major and working as a software engineer. You also make me push myself even further, to try and close the gap if ever so slightly. Thank you for sharing, I loved this project. After you mentioned polarity markers were inaccurate, I could only imagine how debugging went.
This type of display weirdly reminds me of that water/hologram type display in the Myst opening area. It’s so cool to see “sci-fi” technology like that coming closer and closer to reality
Could you make a particle that orbits another one? like as a model of the moon going around the earth, or perhaps like a real time clock of the solar system
It may be possible to do this, however there are only two planes of transducers which is a hard constraint on the maxima and minima of the resultant pressure waves and their ability to isolate particles without interfering with the other and with a fixed radius for each transducer, further clarity is impossible because you cant decrease the diameter easily. If you had a second set of transducer planes, with a normal vector orthogonal to the original set, it would help fix this problem as you could use the second set of planes to help flatten unwanted interaction between the two pressure wells, sort of how noise cancelling works in avionics for the pilot's headset.
You said you could make it faster by double but you dont want to figure out the code? I think you should do it! a challege for you and the world will be greatful, needless to say animations will look smoother. Also, are you able to have more than one ball on at once but having a different animation? You could make rotating cubes with things inside bouncing around. it would look amazing! It would be cool to see 2 butterfly's dancing around each other! Good job man!
This is seriously impressive! I've been thinking about using something similar to make a 3d scanner using echolocation. It would work by driving the array as a directional speaker to scan an area with an array of microphones to receive the sound. I'm not sure I'll be able to write a robust algorithm to extract 3d information, so maybe I'll use a neural network
This is one of the coolest things i've seen in years. Patent this or parts of it if you can! Hasbro is gonna steal this and make a toy, but this is so much more. This could be huge. Amazing idea and execution. I'd love to see this on a larger scale, like how drones are used to make things in the air, this would be another level!
I think what usually happens is the university owns the patent on this. Maybe 25 years down the line we will get them released, like 3D printing was. Think patents last 20 years. 5 is for improvements that enable it to be commercially viable.
The distortions seem to me to be a function of the "springiness" of the air enclosure, which allows the particle to lag on acceleration and overshoot on deceleration. The only mitigation I can imagine is to use a lower mass particle and a smaller low pressure zone, i.e. higher frequency. Also, the interference patterns surrounding the sweet spot are pretty large, so controlling more than one particle would require some separation, unless you mean they would occupy other low pressure zones in the same pattern.
You might be able to tune the software for this. From what I understood from the video the inertia of the ball isnt being accounted for. You might be able to slow it down more agressively with software.
It’s the inertia. You can clearly see it when drawing the square - each edge describes a loose “S” due to over shooting. Aerogel would make a lighter particle.
@@Speaking_on_mute good idea, aerogel. @boggless2771 had a good idea about correcting for the overshoot in software too, I assume by an early stop and a brief counterforce. I saw that done in a sort of custom robot painter which was having a similar problem.
I did try moving the boards apart to reduce the possibility of reflections and still see the distortions. It turns out the original researchers see them as well and "they're working on it" - hopefully they'll publish a paper on their solution!
This is what genius looks like. I wonder what increasing the spectral resolution will do. I’d imagine it would make the motion exhibited radically more articulate. Really well done, man. I used to work with folks like you. 🇺🇸
Brilliant work.. It's going to be amazing the progress that will be made once the configuration & frequencies can be tweak on higher levels & heavier objects. Its like we are reinventing something that was lost to us a long time ago & we are finally getting to grips to be able to use it again. thanks again for all your work.
Ok actual question : What is the actual sound output of the device? I know it's ultrasonic but I'm just wondering if it is like barely noticeable or like ear-shattering deathmetal rave stuff (Also bro this is just impressive overall, thanks for documenting the entire process and sharing it with us!)
@Daniel Foisy would sealing the unit in a soundproof chamber or something similar to block the noise be possible theoretically? it would certainly improve commercial appeal lol
@@JustAGuy85 LOL no, I toured in a band as a analog keyboard player and programmer. The first time was a joke by the drummer when I had my head in his bass drum. The second time was when the bass player dropped his bass and he had his envelope filter way up, I was close to his cabs which were 8 10in'ers
Really nice project. My guess is, that if you keep multiple styrofoam balls far away enough from each other horizontally, as not to influence each other, you could just use the superposition principle. It would also be interesting, if one could make an acoustically transparent ball, like a very thin balloon. In principle one could segment the display space into cuboids and use one ball per cuboid for complex animations...
Seeing this stuff feels like humanity is not that far from some major next generation content viewing method breakthrough. Like completely next dimension. Great work wish you get some huge credit and return from it in future!
This project is amazing! Glad to see a more in-depth video about this as not many go into detail about how it works and the key components. What frame rate and shutter speed do you use to capture the circle and butterfly shape on camera so the whole animation can be seen as one continuous line?
Thanks - the frame rate on the camera was 29.97 FPS. The frame rate of both the circle and butterfly was 10FPS - it really does look like this (maybe a bit better) in real life :) No camera tricks involved - I could do a longer exposure and get a more complex shape on film but to the naked eye, it certainly wouldn't look as good...
Fantastic work, both on the project and the video! Seeing projects like this is one of the best things about the internet. Can the FPGA handle 64 transducers? I have no idea how much this complicated the rest of the electronics, but it would be neat to have 8x8 modular arrays to create larger arrays from array blocks.
It would be interesting to see how this would behave with dust instead of a ball. With the lower mass particles there may be less distortion due to inertia. What an amazing piece of work, sharing this incredible amount of effort to the world is to be applauded. Thank you. When art is created, it should be recognized. That butterfly animation was breathtaking.
My guess would also be that inertia causes the distortions. From you description, I get that you steer the ball by moving the position of the stable point around. If you take into account the amount of force on the ball, the positioning would be more accurate and more fluid (are very much more complicated, I have to admit).
My initial reaction is "holographic displays are getting real" but this would take 50+ years of R&D to actually become usable. Much like the first camera. Keep it up though. The progress you make in this now can have a butterfly effect (ha) in the future. For the shakiness in the image, I believe that's lag from the transducers, or the object basically hitting an invisible wall. Slowing it down just a tiny bit might improve the quality
I think I figured out the distortion issue, and it just comes down to the fact that the ball has momentum. The position the ball is supposed to be in changes instantly, but the ball takes a moment to get there, and then overshoots a little bit before bouncing back. This also applies to instant direction changes. Doing a little bit of PID action would probably fix this.
Wow. I like how the brevity of your statements completely hides the massive amount of work and inevitable frustration that must have gone on... transducer polarity, code, FPGA... each of these parts represents a lot of work. Well done.
Thanks - there is definitely more to say on this board and I'm feeling I need to write it down. If I find a few hours, I'll try to do so...
@@abitembedded Yes please. There are viewers of this channel that actually are interested in the fine details of FPGA programming.
I felt quite the opposite. I could see every detail of his hard work to the point where I got tired just listening about it :p but the result is worth it!
@@JayPixx i started questioning my entire existence. How is a mortal person able to achieve stuff like this !?
Not just programming the FPGA or the pi stuff, but creating/designing the whole board, too! Wicked engineering
As a computer engineering student, the second I saw those Modelsim waveforms was the second I realized what level this project is truly on. This is absolutely insane and next level.
exactly. _100%_ true man
Same
Another computer engineering student here and I can only agree. The amount of knowledge to make this video is insane. Congrats!
my lab is like this: i get 1/4 of that 6:50 waveform, but with no suggestive names and i should figure out in 2 hours what it all does(i need to learn how a MIPS32 processor works). I kinda managed to figure out what memory and registers do and edit them a little to see if my understanding was correct, but then i had a lab about a fraction of a pipeline, i couldn't understand a thing.
i am looking forward for the times when I need to do the code myself and simulate the waveform to my willing to fix my projects.
I’m not familiar with that, what is it?
"So that's pretty much it"
So that's pretty damn incredible.
Jeff who? :D
@@pratwurschtgulasch6662 Jeff • 10 months ago
@@pavlepetrovic989 Jeff Besos?
@@pratwurschtgulasch6662 Jeff bingo?
@@gremlin1331 look on youtube for jeff who, it's pretty funny
I always wondered how sci-fi holograms could ever be realistic, but something like this with multiple small particles... close enough. Nice!
We can try this i think since light is both a particle and a wave soo it might possible to move light particles using sound
@@joshuapeligrinoWell, sound is just pressure waves, so you technically can change the speed and direction of light using sound to rapidly increase or decrease the pressure in an area.
Not sure you would be able to meaningfully effect light this way, but it's an interesting thought.
Would probably also explode when doing this, cause the difference between a sound wave and a shock wave is just the amount of energy involved...
I've seen demonstrations that look like an actual physical object. The way it works is that a mirror is placed a a 45 degree angle, and it's rotated very quickly. Above (or below) it is a fast screen that changes the images rapidly, so you'll get a different image depending on the angle you are looking at it from. That was over 10 years ago. There's really no commercial application for these displays. They are novel, but that's it.
There's a WAY easier method if you only need to create a hologram in the sense of a volumetric display (which is how it's shown in Sci-Fi). You can just shoot two infrared laser beams into luminescent smoke, and where they intersect, they create a dot of visible light (since only their combined energy suffices to cause luminescence). This way, you can easily build a whole hologram. It's a well known technique for at least 20 years now. The only reason why we don't use it more often is because there are very few applications where it actually makes sense. What's so impressive about the project in this video is that it's using an actual physical object.
@@joshuapeligrino Even if this is possible, if we trapped the light in the sound waves, it wouldn't be able to enter our eyes and we wouldn't be able to see it!
I understood 0% of this, but what ever your doing don't stop, the world needs more people like you. if you could throw this together, I doubt there's not much you can't do. let the mad scientist take over.
Same
The way you just easily say: yeah, hook two FPGAs together, just program their memory, attatch an EEPROM hook it together over SPI with a Raspberry Pi to control the 100 I/O lines for my 100 sonic transducers shows how much you actually know that you fly over those topics. Incredible! Loved it, thank you very much. :D
It's said like the TH-cam video isn't the final product goal, and the project is what he's actually interested in.
@@asailijhijr that, as well as like he's trying to fit the description of what he did into 10 minutes so that people will actually want to watch it
totally dude. entirely right!
@@rarebeeph1783yeah otherwise it's too niche and people won't watch it
Really well done project! Thanks for showing it to us in this video.
Thanks Ben - big fan of your channel!
Hay Ben, waiting for your next video.
@@abitembedded I think you just need to adjust the frequency - Test the effects on a larger dot.
Also, look for a means to hold the actual device, consider sound canceling.
Funnel a single point, try salt, molten, I wanted to see your process to make a line used to make a cube.
@@abitembedded Have you tried with water droplets ?
@@abitembedded Truly excellent Dan. Much appreciate you sharing.
This is pretty incredible.
An actual 3D image made with sound and one little ball.
In colour.
Quite amazing.
I wonder how complex you can go once you figure out multiple balls.
I dont know if thats quite possible in this format, or at least on this scale🤔 I am not sure, I am not well educated or informed in this topic
@@skateraptor12 then why comment?
@@nathanielyoungman4454 haha he said balls
@@nathanielyoungman4454 because they're giving their opinion. you don't need to be an expert to share your thoughts
@@nathanielyoungman4454 Ofc, my mistake, if only the respectful and educated were able to comment the world would be a much better place
I'm 2 years too late, but this is absolutely incredible work man. Huge props to you for making all the design files available online!!! :) Are you aware of anyone else making one of these? Did you ever consider selling kits at cost to science museums around the world. This is totally amazing work you've done putting in all this leg work :)
Thanks for the kind comments! I know of one other group that is currently building one of these based on my design. I've considered making a kit but the amount of work to do so is daunting and I have so many other projects to work on (most of which don't hit TH-cam :) )
@@abitembedded I understand, I would love to see more videos from you about the projects you do now, you have an amazing amount of skill, I can't imagine how much one would learn from just looking over your shoulder as you built this stuff, the FPGA design alone let alone all the hardware, and the GPU simulation. Your a one man show :)
Seconded! Awesome work.
@@abitembeddedlook, seriously, One of The projects must be Doctor Who's Sonic Screw driver. Pleeeeeaaaasssseeeee
Make a kit
This honestly has to be one of the coolest projects I've ever seen documented on TH-cam. I imagine it's only a matter of time before some company comes in to sell something like this, just because of how awesome it looks.
As an electrical engineering graduates, this gives me goosebumps. Well played sir. This is an artwork.
This is incredible. The digital design alone is an impressive feat, but interfacing to the array and your pipelining of the math is really stunning work. The end result is incredible!
agreed! totally, totally true man
A really cool thing to "draw" with this would be a Lorenz Butterfly. Instead of animation frames, simply make the ball follow the trajectory as quickly as possible. I know this is a year old, but it would be cool to see if you can make it work.
Yes! This!
Watch him solve the three body problem with some weird hermeneutic sacred geometry frequency thing
@@fitz3540hermetic?
This is one of the very top most impressive maker projects I've seen on TH-cam from an apparent complexity standpoint. This is from someone who has spent a huge amount of their free time finding the best ones instead of my own projects. :) I bet the video of the butterfly animation doesn't do justice to seeing it in person with real eyes.
Thanks! 👍
my man this is equivalent to like a whole research paper worthy of a PhD
thanks!
This is just stunning and an insane amount of work across a variety of subdisciplines: fpga, firmware, electronics, layout, fabrication, the list goes on. How long did this take you?
Thank you! It was about four months, working about an hour every night, somewhat more on the weekends
@@abitembedded Incredible! What is your background?
@@abitembedded I really need to know your background and education. You're doing things I dream of doing but are so interdisciplinary I don't know how to approach. Please answer.
@@thek3743 I'm a computer engineer
@@abitembedded hey keep doing love your stuff
In my experience, when someone says "the FPGA code is simple" it is anything BUT simple. Fantastic!
"just some integer math"
volumetric displays really do seem like the future. and the use of sound for it is amazing, imagine this on a huge scale, that comment at the end where you talked about how it can stimulate your touch nerve endings, imagine the VR capabilities, you could actually put a car right in front of somebody half the world away, and they could touch it. Could also be used for disabilities, imagine a blind person that doesn't need a walking stick because their entire surroundings is projected onto their hand, or a deaf person that has a persons speech turned into a touch map. taking a research paper and doing it yourself is incredible
I imagine the biggest constraint to this method would be collision. The position of the number of balls would have to be shifted such that they never overlap the same X, Y, Z. This wouldn't work for more elaborate 3D models because there is only 1 dimension of 2 forces (gravity vs pressure on the Y axis).
For a touch map you'd want to add a skin attenuating beat frequency to the signal.
Edit: Nevermind, he mentions that the paper covers that.
look up ultrahaptics, looks like they make what you are describing
@@Mildain2000 Just the refresh rate alone would be an issue even if it was possible. You're constrained by the physical speed. This is why volumetric displays will always just be things like LED's layered together. But volumetric displays are honestly not that cool for viewing anything but basic shapes, because they obviously have a limited depth, and images of things will have near and far depth of fields, and focused or blurry details. So you'd really only get a narrow range of depth and then it'd just be a cut off to become a flat wallpaper of the rest of the things in the background. This would pretty much break immersion.
About all a volumetric display would be good for is things like the medical field where you want to view sections with physical topology, but you can already get that in VR or AR. For home use, it would only ever be good for giving UI elements a pop out effect. Another issue with volumetric displays is working in the 3D space itself. A simple mouse and cursor won't do. Even a touch screen wouldn't do, because eventually you'd have elements overlapping, so you'd still be limited to a 2D field of selection. Needless to say, we're not getting volumetric displays any time soon, and we'd be pretty disappointed in the practicality of it anyways. The only ones that make sense are holographic displays.
Insanely well done. I wonder if you could animate something like 3D pong, as if the ball is a mime between invisible paddles. The ability to get a couple extra spheres in there to simulate paddles would really complete the illusion. But so impressive as is!
That would be a cool living room peice
and then in the future we could play ygo like in the anime
that might be the dumbest thing i have ever heard.
It could use 2 of those spinning display things or something else on 2 sides and show the paddles then the particle is the ball lol
Would be amazing but I worry about what would happen as the ball approached the paddle. Would they start to interfere? I guess just give the objects a huge hotbox.
I just had to come back to this video because it’s been living rent free in my head for months. Still probably the coolest and most impressive project I’ve seen on the site. I really think this technology can and will be used. It really is inspiring. Well done
thank you so much for your kind comments!
The simulation is like the most important part of this whole project. I can't do that. Without it he could have spent months or years trying to figure it out maybe, just because of the time involved making and changing it physically vs just simulating it and knowing it will work.
I'm also cheap - I wanted to make sure I understood that I can build this without spending a dime... :)
Absolutely Incredible! This could be someone's masters thesis and from what i can tell you just did it for fun😄
indeed. precisely right dude
Except this is Ph.D level...
This is what I was thinking. I was like, alright physics, I can keep up with this, simulations, alright yeah I kinda understand, fpga, I know what that is, 6:46, jesus christ what does this guy do knowing this much about so many different disciplines, what the hell is this
This is mind-blowing, a true pursuit of scientific curiosity. The amount of effort that went behind making this video is highly understated.
"I can't do it realiably and don't have the patience to further tune the system"
Everyone reaches that point someday. I love the honesty and laugh at thr thought of reading that in a paper.
It’s amazing how this video has taken things I have learned from three different subjects this year, combined them, and made me think with them in a way I never had before. Now I’ll be able to hold physics, digital electronics, calculus, and computer science in a whole different way now that I’ve seen the connections between all of them drawn together in one video. Thank you.
This is like physical vector graphics; insane work, thank you for sharing!!
What do I think? You have that rare combination of genius and an amazing work ethic! Modern technology takes a place as well of course. Mail-order PCB houses are another part of that phenomenon. Amazing! It would be great if you could juggle a couple more balls, and therefore allow more lines to be drawn, right? Would the maths be significantly more complicated? Could the hardware do it? Could you perhaps work an entire sheet of particles as a 2D plane that moves up and down? A little bit like those 3D displays that use a projector, and a screen that moves back and forth very quickly?
Perhaps a video projector could replace your LEDs to colour a lot of particles at once?
You could perhaps sell these to science museums, at a significant markup, if you wanted to be paid for your time. It's an example of something spreading from a lab paper into the hands of the public in one step, over the Internet. Back when we were idealistic about the Internet's use for democracy it would have been something geeks would have been proud of.
Thank you! Interestingly, the math to move more balls doesn't seem to be more complicated than adding together the "waveforms" for moving individual balls - certainly on my list of things to do...
@@abitembedded and ? Can it be done ? Video when ?
@@AndroidFerret It's all about time :) Given all the recent interest, I think I have to prioritize coming back to this project. But I do have another project on the go right now...
@@abitembedded I'm glad the algorithm shoved you into my recommendations! Subscribed to see more awesome, well documented projects!
This is just purely Magical. Yes, I understand 80% of everything that's happening, but it's amazing to see a single person finishing everything from acoustics to physics sim to coding to hardware programming etc.. It's just awesome!
no bad apple. 0/10.
in all seriousness, this is absolutely amazing. just the fact that acoustically trapping particles is possible is crazy enough on its own let alone being able to move them around fast enough to get some - admittedly kinda janky - animations!? huge kudos to you, i cant imagine the amount of work hidden behind the veil of this simply-explained 10 minute youtube video.
any sufficiently advanced technology is indistinguishable from magic. I'm just about the most scientifically minded person I know and this is basically wizardry to me
Smoke/ aerosol particles might be interesting to observe in this. might need extra some phase cancellation in negative space to clear out the image tho
I've just stumbled upon this video and find it extremely fascinating. This is the first step into modern-day holograms for anyone unaware. The potential behind this technology is amazing, and I am very excited as an Industrial Designer to see the evolution of this tech.
Great video and I would love to see an update on the multiple balls theory.
This is incredibly awesome. Please continue sharing your projects with the internet!
The fact that this man gives us the time stamp of the time the video starts is a blessing
This is one of the coolest things I’ve seen on TH-cam in a long time. Super cool experiment, very inspiring
Absolutely beautifully done and explained! Hoping to see more projects from you. Thank you for sharing your research.
This is pretty darn cool. I can't find the right acoustics to describe how impressed I am.
This is the first tech i have ever seen that cold evolve into a star trek style holodeck. Sound waves we can't hear manipulating the positions of possibly trillions of tiny particles to create the illusion of larger objects. AMAZING!
I was not ready for how hard this was going to blow my mind, and then you introduced color.
Alright I’m calling it: within 10 years, this will be the next technology used in user-interaction interfaces. Touch-screen will be either a dinosaur or obsolete.
This is one of the coolest things I've seen recently, I wish I had the time and knowledge to build something this complex.
I am absolutely blown away by this. Thank you for making the files available, that is so cool.
This is incredible, the amount of time and skill it takes to program an FPGA let alone just do it for a project is amazing! Great Job! I know I could NEVER do that lol
I can’t tell you how grateful and happy I am that you released the design files online. You are a gift
This is incredible. Beam-forming a foam ball to create a volumetric display. What a brilliant idea!
The technical stuff went in one ear and out the other but ooo ball go floaty real fast! If you could physically feel those focal points that would be so cool and weird. I wonder if someday that would be used in VR to make a digital world you could physically interact with
This video is going into my "art collection" youtube list.
That is probably the nicest thing anyone has ever said to this engineer :)
@@abitembedded I didn't just mean it, I did it :) I'm an engineer too and you are deserving all the appreciation :) I also do appreciate your time to describe all your work in a video and inform people about such a thing. Thumbs up :)
wow, I would've never imagined this being possible with such accuracy as well. Awesome work!
Oh wow. This channel rivals applied science, alphaphoenix and old tech ingredients. I’m so glad I found this.
Wow - that's amazing but definitely undeserved praise - thank you!
i love how this is actual physicalised imagery using sound to "draw" one omnidirectional particle, effectively emulating the photon. nice touch with the RGB colouring as well, very impressive!
Are you telling me you can make it so that you feel that butterfly with your fingertips when it ISN'T EVEN THERE?!?! This is some cool stuff man!!! Please, keep doing this stuff!!
yup :)
Beautiful engineering, incredible dedication! Keep it up.
This is absolutely amazing. At the end you said you could feel it at targeted locations within the array. I can imagine hundreds of applications right now. Hopefully this is something that can be commercialized and improved upon to provide resolutions imperceptible to human touch much like we have done with screens today.
Very impressive, Dan. Some similarity with what I did in the 90s in two dimensions with a laser dot on a screen for animated images. The mechanical scanners were critically damped to move from point-to-point in 4ms. Images (vectors) were stored with x and y coords and the mechanical scanners and mirrors "drew" the connecting lines.
When doing the real-time rotations of an image, I calculated the Z coord and used intensity modulation to dim more distant parts for depth cueing as well as making distant parts smaller, simulating perspective (that was a simple, but very effective cheat saving division to change size). .
I even wrote my own multiply routines because the MUL instruction in the 6809 processor was slower than my coded routine.
In an 8 bit machine, I used a 256 value angle look up table which is about 1.4 degrees I call the "Bi-gree" for Binary Degree. This allowed continuous rotations with an 8 bit pointer that auto wrapped around saving code.
So many neat tricks made it work really well.
Cheers.
Very neat - thanks for describing the methods you used. I truly miss the times when you truly had to understand how a processor worked and had low level access to the registers - give me a 6502 or 6805 anyday :)
The idea this could be a haptic system is amazing! Think of the vr and ar possibilities!
This is so FAR over my head I feel Acoustially TRAPPED ...lol awesome stuff.
That is amazing, I couldn't imagine the time and effort that you put into this project, nicely done
Well according to him only about an hour every day for 4 months and more on weekends. So like 160 hours give or take?
This is absolutely insane, I love you calm casual tone while walking threw hardware and software design that would make most people's Brains melt.
Terrific work and thankyou for presenting such an awesome project, I wish you luck with any further refinements and iterations!
Its 3am, and im watching this video as it was reccomended to me... i have no idea whats going on engineering wise, but damn thats cool
Seeing these amazing projects by really smart people like you is what makes me realize I'm still far away from what I could be doing, even now after graduating as a comp sci major and working as a software engineer. You also make me push myself even further, to try and close the gap if ever so slightly. Thank you for sharing, I loved this project. After you mentioned polarity markers were inaccurate, I could only imagine how debugging went.
I'm completely floored, that is amazing 👏 👏 👏
This type of display weirdly reminds me of that water/hologram type display in the Myst opening area. It’s so cool to see “sci-fi” technology like that coming closer and closer to reality
Phased arrays are literally magic.
This stretched my brain to follow along with. You are a brilliant engineer.
Wow, that is some serious tinkering! Most people have no idea how impressive that is
Wow, that's a really impressive bit of work, well done!
3:36 question, does sealing it in different gasses change its speed? as we know air friction is affected by gas dencity etc.
I suspect so but a lower density gas would (probably?) have less ability to lift an object
@@abitembedded a much heavier gas would be perhaps better than? Sulfur hexaflouride?
@@abitembedded fluorescent aerogel for the particle?
Could you make a particle that orbits another one? like as a model of the moon going around the earth, or perhaps like a real time clock of the solar system
Yes please do this
I second this. That would incredible.
Can it have 2 or more particles doing separate things at once ?
@@a_3718 theoretically yes, but the closer they get to each other, the more they interfere with each other, either causing a collision or an ejection
It may be possible to do this, however there are only two planes of transducers which is a hard constraint on the maxima and minima of the resultant pressure waves and their ability to isolate particles without interfering with the other and with a fixed radius for each transducer, further clarity is impossible because you cant decrease the diameter easily.
If you had a second set of transducer planes, with a normal vector orthogonal to the original set, it would help fix this problem as you could use the second set of planes to help flatten unwanted interaction between the two pressure wells, sort of how noise cancelling works in avionics for the pilot's headset.
You said you could make it faster by double but you dont want to figure out the code? I think you should do it! a challege for you and the world will be greatful, needless to say animations will look smoother. Also, are you able to have more than one ball on at once but having a different animation? You could make rotating cubes with things inside bouncing around. it would look amazing! It would be cool to see 2 butterfly's dancing around each other! Good job man!
Are you kidding! you make the insanely complex sound so easy. I followed up to the FPGA and from then on I was flat on my face :)
This is seriously impressive! I've been thinking about using something similar to make a 3d scanner using echolocation. It would work by driving the array as a directional speaker to scan an area with an array of microphones to receive the sound. I'm not sure I'll be able to write a robust algorithm to extract 3d information, so maybe I'll use a neural network
Best of luck - I'd love to hear how that goes
This is one of the coolest things i've seen in years. Patent this or parts of it if you can! Hasbro is gonna steal this and make a toy, but this is so much more. This could be huge.
Amazing idea and execution. I'd love to see this on a larger scale, like how drones are used to make things in the air, this would be another level!
There's already a paper describing it, so there's prior art. Either the team of scientists already filed the patent, or it's not patentable anymore.
I think what usually happens is the university owns the patent on this. Maybe 25 years down the line we will get them released, like 3D printing was. Think patents last 20 years. 5 is for improvements that enable it to be commercially viable.
The look of this reminds me of a 3d vector display. I'd love to see where this tech could go in the near future!
Discovered this in Dec 2023. Great project - huge respect!
You're from the future?!
That's freakin incredible... I just can't say much more - you're so cool for being able to do such stuff!
The distortions seem to me to be a function of the "springiness" of the air enclosure, which allows the particle to lag on acceleration and overshoot on deceleration. The only mitigation I can imagine is to use a lower mass particle and a smaller low pressure zone, i.e. higher frequency. Also, the interference patterns surrounding the sweet spot are pretty large, so controlling more than one particle would require some separation, unless you mean they would occupy other low pressure zones in the same pattern.
You might be able to tune the software for this. From what I understood from the video the inertia of the ball isnt being accounted for. You might be able to slow it down more agressively with software.
It’s the inertia. You can clearly see it when drawing the square - each edge describes a loose “S” due to over shooting. Aerogel would make a lighter particle.
@@Speaking_on_mute good idea, aerogel. @boggless2771 had a good idea about correcting for the overshoot in software too, I assume by an early stop and a brief counterforce. I saw that done in a sort of custom robot painter which was having a similar problem.
Your proposed software corrections would be effective, Im certain. They would not be easy to implement, I suspect, and would be intensive at runtime.
@@Speaking_on_mute yeah, I'm not sure it's really worth it, it's really more of a puzzle than a practical device. Interesting challenge though.
Excellent! Thanks for sharing!
The distortions could be due to reflections, or differences in transducer radiation patterns.
I did try moving the boards apart to reduce the possibility of reflections and still see the distortions. It turns out the original researchers see them as well and "they're working on it" - hopefully they'll publish a paper on their solution!
@@abitembedded well I'm well below this project, but would some sort of dampening help?
Could the distortions be related to signal quality, air quality, or manufacture quality of the transducers?
This is what genius looks like.
I wonder what increasing the spectral resolution will do. I’d imagine it would make the motion exhibited radically more articulate. Really well done, man.
I used to work with folks like you. 🇺🇸
I had an acoustically trapped particle lying around and just couldn't think of what to do with it. Thank you!
Brilliant work.. It's going to be amazing the progress that will be made once the configuration & frequencies can be tweak on higher levels & heavier objects. Its like we are reinventing something that was lost to us a long time ago & we are finally getting to grips to be able to use it again. thanks again for all your work.
Whoa neat! I’ve seen the ultrasonic units before, but this is the first one I’ve seen used as a display like that.
I love it and really appreciate the effort put in this project. Awesome work!
Ok actual question : What is the actual sound output of the device? I know it's ultrasonic but I'm just wondering if it is like barely noticeable or like ear-shattering deathmetal rave stuff
(Also bro this is just impressive overall, thanks for documenting the entire process and sharing it with us!)
a high pitched squeal, caused by all sorts of resonances - extremely annoying
I was thinking this as well, i'd imagine it could be quite piercing at just the right spot. I've been knocked out twice from sound concussion.
@Daniel Foisy would sealing the unit in a soundproof chamber or something similar to block the noise be possible theoretically? it would certainly improve commercial appeal lol
@@swordinhand8356 Wow, how'd that happen?
Please don't say the government is after you and that you're a TI.
@@JustAGuy85 LOL no, I toured in a band as a analog keyboard player and programmer. The first time was a joke by the drummer when I had my head in his bass drum. The second time was when the bass player dropped his bass and he had his envelope filter way up, I was close to his cabs which were 8 10in'ers
Gotta love our persistence of vision for things like this
Wow.. I haven’t been this impressed with a guy on TH-cam since I discovered Shane from ‘stuff made here’. This is incredible work! Instant sub
Such an inspiring video and super well explained! Really curious about how long it took you to accomplish this project from start to finish?
It took about 4 months
I’d add a “power off” function so the ball doesn’t go flying, but instead lowers to the center. Other than that, pretty cool!
Really nice project. My guess is, that if you keep multiple styrofoam balls far away enough from each other horizontally, as not to influence each other, you could just use the superposition principle. It would also be interesting, if one could make an acoustically transparent ball, like a very thin balloon. In principle one could segment the display space into cuboids and use one ball per cuboid for complex animations...
Amazing project, a shame that it's video doesn't get more views simply because it's to advanced for 95% of humanity.
Seeing this stuff feels like humanity is not that far from some major next generation content viewing method breakthrough. Like completely next dimension. Great work wish you get some huge credit and return from it in future!
That’s still beyond my understanding, but congratulations! That’s completely fantastic!
This project is amazing! Glad to see a more in-depth video about this as not many go into detail about how it works and the key components. What frame rate and shutter speed do you use to capture the circle and butterfly shape on camera so the whole animation can be seen as one continuous line?
Thanks - the frame rate on the camera was 29.97 FPS. The frame rate of both the circle and butterfly was 10FPS - it really does look like this (maybe a bit better) in real life :) No camera tricks involved - I could do a longer exposure and get a more complex shape on film but to the naked eye, it certainly wouldn't look as good...
Fantastic work, both on the project and the video! Seeing projects like this is one of the best things about the internet.
Can the FPGA handle 64 transducers? I have no idea how much this complicated the rest of the electronics, but it would be neat to have 8x8 modular arrays to create larger arrays from array blocks.
I don't understand a word of what you said but your inteligence amazes me, incredible!
Sounds so easy when seeing it, but there are tons of experience, knowledge and hard work in there.
You are well skilled. That´s a great job!
All the bats in this guys neighborhood: “WHAT THE FUCK IS MAKING ALL THAT RACKET!?!”
It would be interesting to see how this would behave with dust instead of a ball. With the lower mass particles there may be less distortion due to inertia. What an amazing piece of work, sharing this incredible amount of effort to the world is to be applauded. Thank you. When art is created, it should be recognized. That butterfly animation was breathtaking.
My guess would also be that inertia causes the distortions. From you description, I get that you steer the ball by moving the position of the stable point around. If you take into account the amount of force on the ball, the positioning would be more accurate and more fluid (are very much more complicated, I have to admit).
My initial reaction is "holographic displays are getting real" but this would take 50+ years of R&D to actually become usable. Much like the first camera. Keep it up though. The progress you make in this now can have a butterfly effect (ha) in the future. For the shakiness in the image, I believe that's lag from the transducers, or the object basically hitting an invisible wall. Slowing it down just a tiny bit might improve the quality
That's absolutely amazing. And the idea of invisible haptics. Incredible
I think I figured out the distortion issue, and it just comes down to the fact that the ball has momentum. The position the ball is supposed to be in changes instantly, but the ball takes a moment to get there, and then overshoots a little bit before bouncing back. This also applies to instant direction changes. Doing a little bit of PID action would probably fix this.