179 comments

[ 698 ms ] story [ 4271 ms ] thread
I love this. Must be a lot of fun to work on it!
One of those brilliant ideas that seems obvious in retrospect.

Given that the results are so compelling even when pulled together by hand out of relatively primitive discrete components, I’m wondering why we shouldn’t expect to see full color, high resolution versions from every random 7-letter drop shipper on Amazon next Christmas?

It's not a new idea -- they're called swept-volume volumetric displays, and they've been around for a long time [1].

But they're mostly kind of just toys. When you're limited to transparent glowing surfaces and you can't even touch them, there's really not a whole ton you can do. You can see a much larger non-spinning version made of LEDs suspended on cables [2] and it's very cool, but the novelty kind of wears off after a while, and you realize it's not the kind of display you're going to use for anything productive.

If you actually want real 3D visualization that can render anything at high quality and that you can interact with, it seems like VR/AR headsets are the way. (Though there are also the new 3D monitors that don't require glasses, but not a lot of people have gotten to see those in person yet.)

Maybe there's some kind of toy you could make with them to sell on Amazon though? Not really sure if there's a "killer app" for these things.

[1] https://en.wikipedia.org/wiki/Volumetric_display#Swept-volum...

[2] https://www.ledpulse.com/

For [2] it does not look swept. Appears static.
Yea, specifically it looks like the LEDs suspended on threads and statically fill the volume.
The killer app is clearly sending messages via R2-D2.
This... is true.

It would actually make a fantastic Princess Leia message toy!

"You have an incoming message request from: SALES TEAM" "Help me, dev team, you're my only hope."
The innovation is spinning the entire computer, instead of just the display, reducing complexity
It seems more likely that a mechanical engineer would end up using a slip ring, so the computer doesn't need to spin.
for a 32 inch display, you gonna get a motor that may need to spin like a mofo in an area of pi*32^2.
The display needs to be pretty strong physically.
I’ve seen 1kW Chinese closed loop servomotors, with driver for < $500. And you don’t care about acceleration, just continuous torque.
I wonder if you could put small displays in an array and somehow synchronize them to act as one big display.
The centering issue could be solved by putting two led boards on top, back to back.Perhaps one board could shift the leds over by half to create an interlacing effect while doubling resolution.
There's gotta be some way to replace the rack-of-leds-spinner with something lighter. Mirrors? Acrylic light pipe matrix?
I wonder if it would be simpler to just spin two phone screens glued back-to-back. You get higher resolution in about the same weight. Bandwidth would be a problem though.
The fundamental problem is refresh rate. If you have two phones that refresh at 240Hz (the fastest I know of), and you want your frame rate to be 12 Hz, you’ll still only be getting 20 refreshes per 180 degree rotation. So your angular resolution will only be 9 degrees. Assuming you use portrait mode, that will make your outermost voxels ~5 mm arcs.

And even there, 12 Hz is probably pushing it in terms of flicker.

Yeah, well, that's not going to work. Interesting problem all the same.

For a 1 meter diameter and height cylinder at 60Hz voxel refresh and say 1mm resolution at the edge one would need the plane rotating at obviously 3600 rpm, and edge pixels switching at about 380KHz. Since rgb is nice to have that's 10 gigabit for a somewhat coarse display. Nothing impossible, but not a DIY territory yet.

I think one'd want the LEDs either on the receding or the advancing halves of the plate, depending on which is best for the cooling -- not all on the same side or on both -- that would be just a waste.

Or even just on one half of the plate, the other being just a countermass. This way cuts the bandwidth in two.

> For a 1 meter diameter and height cylinder at 60Hz voxel refresh and say 1mm resolution at the edge one would need the plane rotating at obviously 3600 rpm

The edge of such a display would also be moving at several hundred miles per hour, which creates it own whole set of problems.

Note that in your hypothetical 1m 60Hz display, the edge of the cylinder is moving at 422 miles per hour. I'm going to recommend an evacuated transparent bell jar for safety, at a minimum...
Of course it has to be evacuated and maybe filled with helium or something less viscous than air at low pressure; this has to be calculated wrt the heat rejection from the leds. Also centrifugal force shearing and tearing off components, imagine the fun.

All in all, this is what makes it interesting, no? 1x1 meter holo tank, not those tiny Voxon shakers.

I wonder if anyone has built multi-speed “layered” swept volumetric displays. For the center part, you would still just have one screen, but as you go further out, you’d have “spokes”, so that each screen only has to sweep a small angle (and roughly constant distance) during a frame, so that it rotates at lower RPM, and has to update fewer times.

You would still need a bare minimum of 1-2 kHz, and a lot of screens, and balancing every layer would be super difficult, but I don’t think anything about it is fundamentally impossible with current technology.

Not exactly sure what the details, but there high quality, commercial products that are probably doing something like this: https://voxon.co/products/
This is a sweeping display, it moves up and down rapidly with a projector beneath it. Same idea. I saw a demo once in NYC and it blew my mind.
(comment deleted)
Doesn't it have to be transparent?

Are phone screens (minus the phone etc) transparent?

Why do you think the weight is the problem?
Because EVIL = RPM X WEIGHT.

And we can't reduce RPM.

Get it light enough and you don't even have to center it that good.

Ah ok, I figured that by casting the whole thing in resin and adding a top bearing it would get rid of the air flow issue completely at the expense of some extra work and a bit less light output.
Or just ... don't center, and change the math to acknowledge that the LEDs aren't centered?
There's another similar thing needing correction, which is that the LEDs near the center sweep a much smaller volume than the ones at the edge, and should be dimmed in order to yield equivalent luminance. LEDs describing tiny circles very close to the center would need to be dimmed a lot since they'd essentially be stationary. Wouldn't it be better then to sweep slightly larger circles at the middle anyways?
If we're going there, note also that all the LEDs not on the edge are blocked by other LEDs (or the board) part of the time; LEDs on the edge are visible even past 90°.
I would pay $200+ for a bigger (50x50?), rugged, RGB version of this that you can make to display different images without coding.
It's not volumetric, but it's still really cool and may fit your use case https://spinprojector.com/
Only place I've seen these before is at the Atlanta airport, where they use them to show a 3d-animated sign to remind you not to bring your gun through the security checkpoint.
I sometimes see them at trade fairs as an eye catcher. With the right background you can get a pretty convincing 3d effect out of them despite the "display" only being 2d
I saw some poi that worked like that. The performer did a very good show with them
This is tight. Presently working on a list of uses.

Need one the size of a 5gal bucket.

As it is, I'd stick one on my gimcrack cabinet.

Wow this one is cool. First time I've seen it. Reciprocating screen? How the hell do they get it to move that quickly? That is very high energy.

Edit found another video showing it starting up https://www.youtube.com/watch?v=SMz4bJA47Js

Now that I think about it, I suppose if you match the resonate frequency of the display component and its carrier it should be fairly efficient. Really cool stuff.

Other videos say that the bubble is just to protect the screen because (unsurprisingly, I would) people try to touch it - so it's not evacuated; doesn't that mean it's going to be a fairly loud 30hz speaker, effectively? (or is it just not well coupled because "speakers need to actually be designed" and it's not that bad in practice?)
Having electronic candles that display a real 3D image of a burning flame seems like a great use for me! People would love these for outdoor "candle" displays.
Creative. I probably have the persistence and interest to pull something like this off, but I lack the fantasy to come up with such cool projects. Luckily there's the internet to see what impressive stuff people come up with.
Tom 7 mentioned a thing in one of his videos, which is to keep a list, and write down every idea when you think of it, regardless of how stupid it is. Winnowing is for later. That way, when you have the energy to work on a project, you don’t have to waste it on coming up with an idea.

It really makes a huge difference in what you can get done with side projects. You probably have more interesting ideas than you think you do, but if having them doesn’t line up with having the time and energy to act on them (or even to just expand on them), they’ll fade away.

100%. And the "weird" (not weird -- useful!) thing that happens with time is that once you start writing down ideas (and you should definitely note pretty much all of them - regardless of how good/bad they seem), your mind gets into "having more ideas" mode... You sort of get a second wind - or more accurately, you start developing your "what if" muscle. And the stronger it gets, the more "what if"s you'll have. And some of those will 100% be really really good (whatever good means to you).

It's also ok to have really """bad""" ideas. They're always funny to read back (a month/year) later. They're fun to share with friends and strangers. And sometimes they lead to good ones as you think about what made you think the bad ones were bad in the first place.

But when you write it down - please include more than two or three words - because it's hillarious to come back to a list when a lot of time has past to see an idea written simply as 'The pope of physics' <- from my own list. I have no idea what I was thinking...
I have the inverse problem. Do you have any advice for me?

I have no problem coming up with, and remembering, cool projects. I just rarely have the energy or the lasting enthusiasm to actually follow through.

I would start working on one if I were you.
Find a partner who gives energy to you, and who you give energy to. Not always easy to find, but if you do, then its easier to work on stuff like this.
I assume you mean business partner and not romantic partner. My romantic partner gives me great energy but would be like "Why are you playing with that soldering iron and not with me?"

But yes, a good business partner (or two) is, for me anyway, a requirement to move something from idea to reality.

Then play with your romantic partner first, 2 minutes later you can work on the project together :^)
One hour on average - sometimes two ;)
Nah man. It's a race. I finish first, I win. I win every time.
Creativity can be trained just like any other skills! :)

I second taking note of stuff; like most things, genuinely practicing means you will get better.

Something that's also overlooked often is (1) Technical knowledge, (2) Methodical invention (3) Motivation!.

Technical knowledge helps you know which projects are possible (or just economically feasible), and map the roadblocks on the way; impossible inventions are not really useful.

Methodical invention means methodically looking at things instead of just randomly inventing. So you analyze a problem, like: (a) I want to make a volumetric display, (b) I want to bring (virtual) 3D objects to life. Notice the subtle difference between the two. Volumetric displays (a) can range from a lightfield display, which is a very elegant theoretical foundation (and can be realized in a number of ways), to volumetric persistence-of-vision displays like the own shown here. It's likely you will eventually arrive at whatever solutions are possible when you deeply research about a problem domain. Bringing a 3D object to life (b) can mean anything from digital fabrication, VR glasses, haptic interaction, robots, and more. It's a much wider problem domain.

Motivation is also really important. Understanding why you're inventing, and what you what you want to bring to life is also significant. Are you going to bring people joy, is it just very cool, is it a useful medical device that can save lives, etc.. Focusing on worthy subjects is also something you can study and that will really increase how good you are at inventing. Of course making stuff just because [we like it] is also important and valid :)

And for me the most important part is to have fun while you're at it...

I forgot to add: it's unlikely anything you come up with (while trying to develop your creativity) is going to be legitimately useful or even fun. It's like expecting to play a violin like Hillary Hahn (or <insert skilled musician>) after you've just picked it up. But the time to play an instrument so that at least you can tolerate your own sounds and have fun is much shorter -- and frankly, necessary I think not to give up before you can play really well (if that matters at all).
I have the reverse issue, I am constantly coming up with creative ideas, so that by the time I am done building idea A prototype v1 (something like what you see in this video) or v2, idea B comes along and distracts me from seeing idea A all the way through. And when the persistence and discipline is there, bills need to be paid, and the distraction is not idea B, but merely some mundane work.
I solve it like this: when at work the soldering iron is in the garage. That little barrier helps a lot when temptation to tinker strikes.

One thing that I really like about this site is that the creator spends a ton of time on documenting it all. I am usually so busy with projects I don't get around to it. Another problem is that, especially for larger projects, that you need to store and maintain them. And that gets old pretty quickly when you're lugging around 200 pound windmills and such.

Interesting. My electronics work-bench is right next to my work/personal computer, so when I block on work stuff I can switch to doing a little on my electronics projects. It's made me much more productive.
I'm just too easily distracted. I start on a thing that I think will take five minutes, next thing I realize that it's 7 am, birds are tweeting and there is a stone cold cup of tea sitting next to me and a bunch of non-priority stuff done (but it's so much more fun...).

Most dangerous times: when the administration needs to be brought up to date for the tax man. Then I really have to force myself. Fortunately I've been able to off-load most of that but even the little bit that remains is enough to get me to spend 30 hours on something else to avoid 1/2 an hour of work.

Isn't it great how you grab the soldering iron (or whatever) at 2 am for a quick thing, and five minutes later it's 7 am?
Just that... and here I am... 4:40 and not a soldering iron in sight :) Off to bed!

(Ear training exercises are fun. I just built an ear training module into the beta of pianojacq.com and it's a hoot, what's also interesting is that it seems to have an immediate effect on playing ability, I really didn't expect that)

That seems interesting, too bad I don't have a MIDI device (or any musical ability, but that's less important) to try it out...
That is cool.

The code is lean too, seems like one could learn a lot by trying to do this themselves.

Nice example of persistence of vision display. You can use large stick versions of this to “paint” images by waving them around.
It seems like this might Work even better with a two sided OLED screen (2 screens back to back) rather than led array, for better resolution? I think that are sufficiently wide angle, at least the monochrome ones.
I think OLEDs might have a refresh rate issue. The nice thing about LEDs is that you can drive them at a very high refresh rate.
Don’t oled displays support pretty high refresh rates? And also have way lower GtG latency?
with this kind of spinning display you'd need to multiply the frame rate you want by the angular resolution you want. So if you're happy with only 20 different angular views and only a 20fps refresh rate of the 3d display then that's 400fps required of the flat display
Oof. I think that might tax the typical driver ASICS that are used on those.
The rest of the site is filled with dozens of similar very cool projects
Pretty cool, but the bigger the display, the more that refresh rate becomes an issue. Say you can refresh the outer pixels on the horizonal axis at 100 fps. Then if you want to have 50 pixels on the outer half circle, then you can change them at 2fps.
conventional leds can be driven at several megahertz pretty easily. the white type uses a yellow phosphor that fades slowly but they can still do kilohertz
The RP2040 tiny is a Nintendo Switch modchip in disguise (47 ohm resistors on 3 specific pins).
Using blender to generate the frames is really clever.
This would look better with the LEDs on a transparent PCB, right?
I bet you could to it with those little fiber-optic wires. Not the fancy telecommunications ones, the cheap stuff they use to make decorations.

I wonder how they respond to being dunked in epoxy.

LEDs connected directly by rigid wires (no PCB at all), then encased in resin might be worth considering too.
holy hell is this guy productive. it is worth scrolling through the various one-off projects he has completed and documented:

https://mitxela.com/projects

(comment deleted)
I thought so too but look at the dates and the scope of the projects. They're just good at managing their time :) I probably have just as many projects accumulated over the years.
managing time, and ensuring every bit of work is showcased on a single platform, rather than scattered across many public and proprietary venues
I feel like a bit of an incompetent after looking through all that.
Just a catalogue of all the skills on display is already the stuff of a couple of lifetimes.
Looks Pretty cool, Interesting how blender was integrated into such workflows for animation in hardware
(Great YouTube channel if you’re into this sort of thing!)

I wonder if a cheap oled display could be updated fast enough to achieve a much higher resolution. It might work but could look worse since the space between radial slices would be much larger than the pitch of the pixels, most likely.

Ok I just went down the rabbit hole of this guys projects over the last decade and I am now thoroughly impressed and utterly depressed that I'll never have as much free time as he does.
Free time is an excuse for most people - I'm not calling you out because I don't know you but most people spend a lot of time on YouTube - Hn - Watching TV etc and they could be working on a cool project instead.
Watching TV? YouTube? Sounds like your peer group is biasing you. How about struggling to work from home for 8 hours a day with two kids with zero familial support, an endless array of bills, and every possible life obstacle coming your way 24/7?
I wasn't calling you out. I don't know you. (look what I wrote) - best of luck in your situation!
(comment deleted)
He even has entry on this topic: https://mitxela.com/rants (scroll down to "Spare Time and Hard Work")
His work is even more inspirational after reading this rant, as he describes the therapeutic aspect of his projects and his base emotions. Still, I can't help but excuse my inferior output by admiring his incredible fine motor skills and pre-existing expertise in electronics.
Those fine motor skills will go though, use them while you have them! Especially for fine soldering work you need a very steady hand and that which used to be trivial suddenly is a real challenge.
Indeed! His soldering work is of the highest art.
This is a neat project. With such a low resolution, am I wrong in thinking that the flame simulation could be done in real time on-chip using a cellular automata algorithm? It would still be nice to support pre-rendered video for other purposes, of course.
I’d prefer this kind of spatial computing, I like it more than AR/VR – especially that this relies more on human sensory input, as opposed to working against it. E.g. no need to simulate head tracking.
A couple ideas for improvement. If you have extra identical motors and disassemble two, you can fashion a rotating power transmission system from the brushes in two assemblies. You sacrifice two motors for each unit, but it's a perfect fit and very reliable (with a cap and rectifier), and you don't have to worry about batteries anymore.

The rectifier also provides a signal that the assembly has completed a rotation, so you can maintain image stability based on actual position, rather than guessing how long a cycle is.

Transmitting power via induction might work, but I was never able to deliver it efficiently enough, so to make it work I had to turn up the source voltage so high that I worried about fires.

This advice comes from my 2001 Burning Man art project. A very sad early prototype is pictured here: https://github.com/sowbug/tqw/blob/master/photos/side.jpg. The final installation worked great.

If the drive motor was asynchronous you could use the inductive windings on the rotor to power your rotating electronics, but the phase difference may get pretty extreme
Another idea: why not put a tiny TFT display on the motor? You can get cheap ones at aliexpress.
Would be cool to see how that would look like. Or even a transparent OLED or such.
Standard displays are 60 Hz. You need a much higher framerate because not only do you want 60 frames a second, but you also want some number of frames per angular rotation. For 1° of angular resolution, you would need: 360°*60 Hz = 21,600 Hz display. Liquid crystals can be modulated at KHz speeds, but you're not going to find associated driving circuitry to do that. It's not easy, and there's no demand for it.

A TI DLP DMD can modulate at those high speeds, and there's readily available driving circuitry for it. However, it's a small reflective based display designed for projectors, and you would then need a light source to reflect off of it.

MicroLEDs would let you increase your pixel pitch with fast modulation frequency, but the display area is still small at the moment because of low yields. You also need a custom chip to drive the microleds at the required high framerate.

You seem to know a lot about this. Have you worked on similar things?
I work on VR/AR headsets where there's a drive for higher refresh rate displays to do things like multiple focal planes with the extra frames.
What do you think is on the horizon of cool or interesting within your work?
The big push is currently towards microLED displays because of the high brightness and fast modulation speed. Yield is not yet good, and there's difficulty in growing all colors on the same substrate. Picking and placing different color LEDs onto the same panel is not cost effective. Red microLEDs are also not yet as bright as blue and green.

There was a push towards micro liquid crystal panels with laser illumination to create holographic images with depth. There are several startups still pursuing that, but the image quality isn't very good at the moment.

The latest advancement has been in the optics with moving to pancake lenses to reduce the length of the optical path from the display to the eye. The Meta Quest 3 has a smaller form factor than previous generations because of this.

I saw a cool design where they stacked transparent displays
If anyone is interested in that then search for Sean Hodgins on YouTube, he made one like this.
This is interesting info. I guess you can get higher angular resolution if you are able to turn off the screen fast enough, and turn it on at the correct angle. Of course, you won't be able to light all voxels during the same revolution, but perhaps that is not a problem.
What about keeping the motors intact and using the second as a generator?
I might be misunderstanding the idea, but I don't think it addresses the original problem that wires can't connect the two parts of something that rotates in a single direction.
I think what he means is having the second motor act as a generator driven by the first motor (maybe with some geared arrangement to have the generator shaft rotate at a different rate from the body), so that the resulting power can be used for the rotating electronics.
Sorry if I'm being dense, but the second motor (the generator) must stay stationary relative to the first motor to do the generating. Which means that the generator is on the same "plane" as the power source. The problem all the solutions are trying to solve is how to transmit power from one plane to another rotating plane, because physical wires can't do it without getting twisted up. So if the generator is on the same plane as the power source, then it can't contribute to solving the problem.
It shouldn't have to stay stationary I think? As long as the shaft rotates at a different rate to the body, there should be some power generated. That's why I mentioned the need for gears.

Overall though it's certainly not going to be good enough to justify the complexity compared to either slip rings or wireless transmission.

I think I get it now. Maybe the main rotor rotates at 6,000 rpm, the geared assembly has a 10-1 reduction so it rotates at 600 rpm (giving the LEDs 10 fps), and the 5,400 net rpm difference is used to generate power. Yes, maybe that would work.
That's all true, but I liked his solution of using an IR sensor for the sync. It was so cool how he was able to rotate the display by changing the position of his finger. No guessing involved.
He has actual positioning built in with the ir transceiver.