@sam8404 Here’s the main roadblock as of now: real-time rendering. We currently don’t have the capacity to render something as complex as a human real-time. The other thing being that we don’t really have a "convenient" way of scanning someone’s entire body
"radios evolved so the user could change stations with the wave of the hand. which ment that if users wanted to stay om the same station they would have to hold a steady hand awkwardly and uncomfortably in the same position"
Reminds of the early UA-cam days, when it’s was just regular or cleaver people doing things, now billion dollar entertainment corporations, treating UA-cam like the next cable channel.
Theoretically you can double your pixel density by putting another LED panel behind the first facing the opposite direction with the pixels of the second panel being placed in between the pixels of the first LED panel. So when it spins the second panel can fill in the gaps of the first panel. This has quite the potential I must say.
@@grabble7605 wow, if we'd all think like you. We'd never have invented/discovered anything. " A bow and arrow you say? How is this better than a spear?" " A car you say? How is this better than a horse?" Maybe think a little more on the key word of Potential not on what it will be, but what it could be. Dosnt mean it's better now or will even replace the thing you think it will. Just that it's interesting, innovative and moving tech forward not stagnating...
I'm kind of surprised how balanced it seems to be. It would be pretty brilliant to see this in a glass tube replicating the look of a vacuum tube or Nixie tube. Very impressive as always!
Why would your orientation matter? It's spinning so one of the lights is always on. You'll just be seeing it in a different position in the rotation no matter where you're standing.
@@NarattoRadians two words: frame timing. the display needs to spin at a good RPM relative to the framerate of the display's refreshing. ninja edit: well, "needs." it doesn't NEED it, it just looks almost immeasurably more... correct is the word i'll chose.
What I love is the speed control is the "rotation" control to rotate the "image". The slower it goes or the faster it goes, it rotates the "image". Such a simple design but absolutely brilliant!
you could possibly balance out the offset of the leds in the rotation center by sandwiching the leds between their PCB and a bit of acrylic. That way, no weight is needed, but the leds are still easily visible. Amazing project :)
Exactly, yes, perfect Comment Ideally with a Material with the same density should give the best balance and could also be used to protect the led s in between In a bigger version you could also drill out holes in between the led s to reduce atleast somewhat air resistance for better battery performance
@@genshinF2Play if its getting your finger caught in it you could just put it in a case, if its possible to make parts lighter that'll also help with making it shake less? or maybe improve the motor itself
@@MrVijayMadhavan Yeah, probably used as a sync reference, because the DC motor doesn't seem to have any connections that would allow the microcontroller to know when it completes a full revolution, so instead it probably measures the intervals between something being close to the IR sensor to determine how long is one revolution and to have a reference.
@@pvic6959 There's an infrared sensor on the side. The device uses the infrared sensor to check where his finger is many times a second, and then uses that information + some math to make the image face towards his finger
Well, since everything is spinning and that sensor has a definite location, it would automatically rotate the image while using the right speed as a function of the sensor location.
Adding a few degree tilt to the display should let you double the number of apparent rows that are swept through on each rotation, which might help improve the resolution 🤩
For the dumb people in the back, are you saying tilt it like a box onto its corner, so that e.g. the leftmost pixel in a row spins higher than the rightmost pixel and it's like it's an extra row?
I quite like the "bonfire" effect over if it would be just a flame of a candle. But I suppose it might be worth trying to get that one done as well, for the practice at just getting better at programming and tweaking. This is super impressive. Imagine a set of 20 of these, lined up in a grid fashion, and flying on a drone. It's almost like a flying hologram(ish) moving picture sign!
Yeah but the drones are not static, and they don't allow viewers to rotate the image, and the image density is severely restricted by the drone's bodies. Something like this we could probably scale up to 4k/8k image quality over time. We already have pretty high quality 2D spinning displays like this available on Amazon. By 90's standards this tech is already futuristic as hell.
@@bobriquardo5317 At first I thought even with one display like this over the body of the drone it still wouldn't be high resolution because the wings ocupy some space... but then I came to a conclusion, the wings are literally rotating motors! If the wings themselves were volumetric displays, then we would see the future my friends
Recommend the each row contain an odd count of pixel elements with the center row right on the primary spin axis. Or maybe even better, put an odd number of elements on one side of the board with an even number of elements on the backside. This could be two boards glued back to back of course. Bottom line is that as it spins, it would appear as if there were no gaps in the pixel density. (note that the even board's rows should also be slightly off from the odd number to make sure the horizontal gaps disappear along with the vertical gaps. BRILLIANT display. I love it!
Overall adding offsets to this would allow it to fill space much more fully. It's interesring how such a great thing still has the potential to get even better
Thats really clever! From making pixel art i see tge advantage of odd and even number of pixels/voxels but putting them back to back so they fill the gaps and eachothers weaknesses is so smart
Now im wondering if its possible to add color by having 3 sides with different color LEDs. Of course thatd require faster rotation and break your idea, but itd still be cool and interesting Also ruins the beautiful minimalism i think. But itis a direction this display could be taken in
No need to glue two boards, just have a double sided PCB made. Also being double sided guarantees alignment. He should also do RGB and as small as possible for maximum density.
@dougcox835 RGB ruins the minimalism but yeah Another thing thatd be interesting is doing 3 sided display with red green and blue LEDs on each side, i feel thatd be neat. Itd have a side effect of slowing down rotation being a way to animate color change
i would pay to see this in real life. I watched some modern tech art galleries and this kind of stuff is amazing to see. Would be very cool also to see a double sized or tripple sized version, maybe smaller leds? I respect your work on this a lot.
I reckon you get a really high resolution hologram by basically doing this but using one of those SPI OLED displays (maybe a transparent one?) instead of the LED matrix board. Although you'd have to cover/paint it to make it yellow if you wanted it to look like a candle
I looked into this great idea, but unfortunately for this application its not super viable :( let's say you wanted your pov display to spin at 20 rps (20fps basically). These spi displays are usually 64x128. if you wanted reasonable pixel density, i would use the radial pixel count x π to find how many times per rotation you would need to update the display. In this case, the desired circumference resolution would be 201 screen refreshes per rotation (a refresh for each "pixel"). This resolution at 20 fps would require a screen framerate of around 4020 fps, and the theoretical max speed of these spi oleds is around 200fps at 8mhz. This math is for if the display was either offset or in landscape. If you centered the display vertical, you would only need 2000fps, but thats still unfortunately out of range:( I love the idea though!
I really need to be careful what videos I click on. This is one of the coolest things I have ever seen and I'm afraid it's going to re-ignite my electronics hyperfocus until I build one of these
@@USureAbtThat Study electronics in your spare time, if you have access to the internet then you have the knowledge of millions of people at your fingertips. Leonardo Da Vinci never had that.
This is not a new idea. I made such a display in 8th grade in a science contest. Here is a video from like 8y ago from another youtuber: ua-cam.com/video/8MFhoGbUYPY/v-deo.html&pp=ygUYbWljcm8gbGVkIHZveGVsIHNwaW5uaW5n
How about a super thin PCB and leds on both sides, would fix the balancing problem and give you higher resolution of you mapped the actual points the LEDs end up at. Really cool project, feeling inspired!
Maybe offset the reverse side to fill in the blank areas of the first. The wiring is a challenge I’d imagine, because a ribbon would probably be easier, but I’d imagine the current is not inconsequential.
Cool stuff. For balancing, Consider adding weight 'strips' to the left and right edges all the way to the top, pointing outwards to the led side. Or conside adding a metal arc. If you attach it from bottom left to top right, it might average out the amount of light blocking, making it invisible. Alternatively, it could end up with significant interference. Edit: the acrylic layer cover is a much better idea :) Maybe another idea: if you change the board color gradient from the centre to the outside, to reduce the difference in color of the board due to the local speed.
I might recommend having the LEDs be flat and reflected on a half-silvered mirror that's vertical, to mitigate the thing where the center of the volume obstructs voxels in the rear (creating an "opaque central column" illusion in the current version)
if i understand it right (and that's an ask) i think the mirror can be placed vertically where the led board is currently, the led grid being placed horizontally at the bottom facing upwards and the mirror's half silvering will create the 45 degree angle necessary to remove the columnar effect?@lolithighs
Impressive work! It's amazing to see how you've managed to bypass traditional challenges like power transmission in spinning volumetric displays. The rotating cube and fluid simulation look really cool, especially considering the low resolution. I can imagine how satisfying it must be to see it come to life after all the hard work. The next steps, especially with centering the LEDs and balancing the motor, will be exciting to watch. Keep it up-looking forward to seeing how this evolves! written with GPT
Рік тому+15
Woah, impressive ! To get a more "candle" effect, the "flame" needs to be thin, and to dance around it's center on the lower part, letting the higher part going off the "center of mass". It will need your whole LED on the upper parts but not on the lower parts. And moves are generally erratic and quick, with phases of more stable moments. Thanks for sharing this, this is awesome!
yeah, in fact, half the time the screen is facing away from us, so why not put another screen on the opposite side showing a mirrored image of the first screen.
could you do that by spinning a smartphone. use the smartphone accelerometer for orientation & program the display on the screen. maybe even spin it by hand? so the only hardware you would need is some kind of pivot for the phone
Nice! a rotary transformer is another option for coupling power, like they used in VCR head drums, though can get mechanically complicated combining it with the motor. I suspect balancing at the bottom may cause some imbalance about a tilted axis. Maybe a thin PCB with some heavy inductors/ferrite beads between the LEDs could balance it fully.
This is amazing. I am no expert so discard my ideas if useless, but perhaps a clear circuit board to eliminate the black "fog", and a clear resin over the LEDs to act as the counterweight for the board after shifting it so that it can still be higher for balance? Getting that right would still be tricky, though. Also heat may become an issue if they are encapsulated. Really cool project! ❤
i wonder if a clear board and resin might transfer light much easier. you might have a full form image when spun with this method or it might just completely glow like with light throughout.
A clear board would expose the wiring, causing light to reflect off those wires unless the object were to be viewwed in dark lighting. This is a good train of thought, though, just needs more experimenting
@@HazelnutPi you can get transparent PET flex PCBs made quite easily they're just somewhat expensive and have some mounting related difficulties. The traces would likely be thin enough to not have a considerable effect though.
Not sure if it would help, but for the LED center-alignment, they make surface mount LEDs where you put a hole in the PCB and put the pads on the rear and the LED mount from the rear and their bodies poke through the PCB. This might help put the center-line of the LEDs and the PCB in the same place.
That is actually a very good prototype! The interaction built into it at such a tiny size is awesome, and I would love to see future designs with this, higher pixel density! I do wonder if you would be able to make a frame of enameled wire so it's porous? In doing so it would allow for faster and be less power hungry along with weight and air resistance, as well allowing the LED's to be viewable at both 180° and 0° (may end up leaving a trail destroying the illusion) Or even a see through PCB board but that's probably impractical Best of luck on your future endeavours!
or, and hear me out, seal the spinning mechanism in a tube with negative pleasure, or better yet, as close to a perfect vacuum as possible, if friction is the issue here. the only issue now would be to minimize internal refraction.
@@loremaster63 I cant believe the solution to high tech display was vacuum tubes all along. But imagine that, like a 8K display rotating at 120 rotations per second in a vacuum tube. Potentially two or eight of them sandwiched like a tree. You could get some insane games from that. Or perhaps cheat a little and use holographic reflective or projection displays for HUDs or special items. Project it straight on the glass, or in their own plane. It makes me wonder though what is more practical. Fiber optic and voxel based displays, since the biggest hurdle would be clarity and wire routing in the display. Versus rotational vectors, where the hurdle is rotation speed versus clarity.
It reminds me of those cheesy toys they use to have that created a 2d image from a rotating line of LEDs. It's cool to see the concept moved into 3d space. Pretty awesome results.
The technique is the same, they are both persistence of vision. The difference of course is that rotating the plane means led timing must be handled quite differently.
You can put LED's on both sides and holes in the spinning board as you decide which light conjuration you like. I'm sure a few people have already suggested this.
I wonder if this could be done with a small lcd or oled screen for extremely high resolution compared to what you have. If you can get the high refresh rate needed that could make it a lot more detailed.
I agree. You should be able to virtualize the present display on an LCD and then extrapolate the resolution virtually and add color changes. There must be some way to generate complete 3D high resolution images.
For the leds being off center, you can try looking into alternative style LEDs that mount "reversed". They solder onto the "back" or the board, and shine thru a hole in it. I think they're sometimes referred to as "gull wing" leds
This is amazing! Very nicely done! I guess next step can be a smartwatch display, then a smartphone one… with higher resolution there will be so much more possibilities
That is Awesome! I think you might be able to use the backside, with offset diodes (between the front side), to create an interlaced so to speak resolution. ▒ So that the back side fills in the gaps between the diodes in the front side. Should also balance well. ♥ Edit: I wonder if a 2 boards in an X configuration would work? (double sided boards)
Ah great idea. I had just written my own comment about this, stating the same idea. Just to come to the comments and see that you beat me by a week. Would love to see more prototypes on this.
You can either put the lights offset and get an increased spatial resolution that's interlaced or put the lights in the same spots on the back and get an increased temporal resolution ie framerate
Very neat! For a relatively simple assembly it works surprisingly well. A double sided LED PCB with the LEDs offset on one side to double the density might be doable, though doesn't help with putting the LED die exactly on the plane of rotation.
4:25 instead of adding weight and all why don't you create another board and place it back to back with the first one it'll increase frames per second and balance as well
@@zrman96 The little rectangular bit on the side is an IR transceiver. When he puts his finger close to it, the receiver picks up the reflected IR light off his finger. Putting his finger on it wouldn't change the orientation, only the perceived framerate.
@@zrman96 I believe that it is detecting it but im not sure. There is an IR sensor on board thats probably for detecting how close your fingers are but idk if its used to detect the placement of your finger. Maybe he is only friction and im reading to much into it lmao but i didnt hear any sounds when he "touched" it so i assumed it was the IR sensor doing the work. Edit: look at 0:56 ! I actually thing the IR sensor is detecting the position! He's not touching it and the images followed his finger. Thats so cool
@@bluebaconjake405I'm guessing that the ir sensor is used so it knows when is has done a full rotation so it can stay still instead of turning if the speed is a hair off. By moving the thing the sensor detects (the finger) it changes where the display thinks forwards is.
@@masonbarber871yes, and that is the reason for the off-state if the finger is too far away: there's just no "frame" start reference (zero degree marker).
I love this idea and the fact you built a working prototype is really amazing! Thanks for sharing the results with us. I was thinking about the issue of the LED's disappearing when you are looking at the very edge of the PCB. Maybe an "L" shape (top view) would help? I suppose then it would be blank for half the rotation (which is worse). Maybe a double sided L shape or a one sided plus shape (like notch cutouts locking together)? My other pondering was that the square LED's are affected by perspective as they turn. I don't have any good ideas how to compensate for that with an image unfortunately. An LED would appear largest straight on and get less and less wide as it turns toward 90 degrees basically. It may be too difficult to change the brightness based on location of the pixel to match better (but it does seem like you enjoy a challenge).
Well done! I had a similar idea a few years back, but it was a lot bigger and it would just wobble out of control as soon as a turned it on so I gave up on it. I thought of using circular rails that wires from the spinning display would brush up against, but this design is so much better. Thanks for the inspiration!
If you could adjust the brightness of individual LEDs then you can create tonality and additionally gain the benefit of depth which in flame effects could be very realistic. The glow of the flame and then bright sparks or wisps of light. Also the use of RGB lighting could also add another layer of realism. I actually think this is a great concept and with some polish design could be a great product for candle-light like lamps. Very cool.
I am absolutely captivated by this prototype, I would love to have something like this in the form of the bracelet to show off ideas or to analyze my 3D models in my free time.
I have two ideas to improve the work, you can use the Tesla tower to give energy to the device, and since the coil is self-similar, the problem of instability will disappear. And the second point is that you can put this inside a tube without air to eliminate the air resistance.🤔
Awesome! Having the LEDs randomly scattered on the board could help deal with the pixelated look. Although it will probably add complexity to programming the animations.
Subbed! Would love to see your progress on this beautiful device. In a world where people expect every new technology to feature some kind of “AI” it’s refreshing to see someone simply come up with a new practical physical concept.
When you're tweaking the candle flame parameters, please keep in mind that candle flames only flicker constantly if they are in a draft. Most of the time they are quite steady, perhaps with a slight lean to one side or another.
very cool and interesting stuff ,as for centering the led's i suggest adding a clear layer over them that weighs the same as the board that is on the other side thus having the led's sandwithed between the two sheets and having a good centered mass.
I think using a higher pitch led matrix would be perfect for such an amazing concept, and I would love to see it more developed into a final product... love what you're doing and definetly you gained a new subscriber :D
One of the nicest small and relatively cheapo electronics projects to get going on, also would be an amazing addition to a number of larger projects. Great work and thank you for the video :)
to eliminate the central bar (the axe) that is too visible, you should have two system face to face and have leds on both side for brighter display. The display would be between the two axes. Great idea and realization you did, go on !!!
I wonder if it might work to wire up a grid of LEDs not on a PCB at all, using as thin of wires as possible for the current needs, then cast the entire matrix in resin. This would give you a mostly clear "screen" that is able to shine in both direction.
Problem with the grid idea is the LEDs in the middle may be hidden by those in front of them that are turned off. But I think it has been tried and works ok for low resolution bc then the front don't block as much of the ones behind but high res won't work.
I had a similar idea but I'd spin a special semi transparent film that could still display projector light, something akin to holloween nets or a film I have not discovered yet. Then use dual projectors to project on to the spinning screen. The images projected when match the rpm of the screen, the projectors would be 90 degrees from one another for full time coverage.
Thanks for sharing! I've been really interested in Holographic displays as I seem to be noticing more show up. The website is very well put together and had a good time reading the steps you took on how you assembled your first Candle Hologram Prototype. You rock! Keep up the great work, I look forward to the next higher-resolution prototype!
You could also consider a mechanical shutter effect. 35mm film projected the image at 24 frames a second , twice on the screen every second with a timed shutter to clear up the image.
@@ferrumignis the question will be both how big and how fast you are spinning the display we have displays that can go up to 300+ hertz but the problem is cooling as they can make a lot of heat (although the spinning might help…)
@@piman13_71 Can you really get tiny displays with a 300Hz refresh? I'm very interested, what interface do they use? (presumably SPI?) Could you point me in the direction of a manufacturer? The usual tiny LCD or OLED displays can't get anywhere near that.
3:09 : Can I suggest, as an artist 'perspective' is bleeding here. So if you can move one half of the LED on the other side of the board, thus it makes one complete rotation as single frame. Or build a board that has LED on both the sides, where the gaps from one side the board is filled in by LED from the other side of the board. 😊 hope it's not confusing 😬.
Only watched about 1:30 into this video, and its the first one from you I have seen. I get the feeling I am listening to someone who could teach me a lot.
Wow, truly amazing work showcased in this video! Your approach to integrating technology and practical applications is impressive. I was wondering, have you considered incorporating an Inertial Measurement Unit (IMU) into your setup? The addition of an IMU could provide precise positional tracking, which would be particularly interesting when you're demonstrating the fluid dynamics inside the container.
some really simple animations might look great, like animating a ring up and down and changing its diameter. The resolution seems a bit low for things like "flame" or "rotating cube" but I bet some slightly simpler animations would look fantastic on it
@@gw6667 yeah I liked the fluid one too! I guess it's one big object moving so the detail doesn't matter as much compared to a flame where it's all about the details
If on the middle an led could be on the very middle top of the display it would really add to the candle effect but the a hollow middle row or one slighter thicker cylinder below what I’ve suggested could also have LEDs all the way down to make the axis’s shape more active but I guess becoming thicker but shouldn’t mess with balance if centred well. May even improve stability. Great effort as usual and a wonderful prof of concept.
This is incredible! I hope you're able to improve it and create larger-scale concepts. I'm just imagining sculptures you could display in your home and change on a whim. Well done!
This is incredible! Thank you so much for sharing and the writeup. Very detailed and interesting stuff! I can definitely see a future version using addressable LED matrix, something like the high density screens adafruit offers.
couldn't you inset the LEDs upside-down into the board, so they're on the same plane as the board and centerline, without needing additional balancing?
Hard to do with regular leds, but there are leds packaged for exactly this with their legs poking out. Might prevent loading any components on the back though.
you could balance it with a plastic transparent sheet of approximately the same weight as the LEDs support, glued on top of the LEDs, and then place the LEDs centered on the rotation axis
This is super interesting, I would be curious to see what this would look like with an old cellphone screen or any small pixel based screen in the place of that LED array!
Add a thin piece of transparent (translucent?) plastic to offset the PCB weight, to allow the LED's to be moved to the midplane; also make the PCB single sided and sand the PCB down to make it thinner/lighter, or print it on flexible kapton with cutouts for the LED's with two rigid wire towers for support. Maybe have flexible tabs for each end of the LED that fold down perpendicular, and use a tiny jig for soldering the LED's to the bent tabs.
I like this project a lot! I liked the idea of spinning all components and the IR sensor to sense the orientation. I'm curious to know if it is possible to use a mini LCD / OLED display instead of the LED matrix, in order to get much higher resolution? Great work.
Hi, love your work. Would it be possible to use a LCD/OLED screen as the light source ? They are often quite thin, and if you use a 3D-printed jig to align it, then it may be possible to a chieve a very nice volumetric cube. (But battery could be heavy, maybe use a low friction slipring ? small solar pannel and laser ?)
I wonder if you can get one with a high enough refresh rate. A pixel has to update for each voxel it passes through, per frame, assuming there's one frame per rotation (there might be a minimum RPM required for persistence of vision idk) There are some clear OLED screens, like the ones Sean Hodgins used for his TENEX volumetric display, that would fix the issue of half of the updates facing away from the viewer, which might decrease the necessary RPM? which would in turn increase the spatial resolution for a given temporal resolution, I think. Couldn't find the refresh rate for those displays in the specs, though. It might not be high enough to make up for the double visibility.
This is absolutely insane, I don't think people get just how cool it is considering you can give this to a professional brand and they can iterate and make it high res + rgb.
You’re the best man. I love everything electronic, circuit boards and their components and what you can do with them. I thought you led ear ring you made. That tiny flexible pcb and those tiny leds were so crazy cool
This is incredible! I'd love to try making something similar. I love the use of an infrared proximity sensor as both a switch and feedback for the rotational speed, absolutely brilliant! Keep up the great work!
the weight of the offset led mounting panel can be balanced by a clear panel in front of the leds... more ideal but harder to prototype would be a clear panel with leds and wiring embedded in a clear panel, possibly a poured resin that then picks up the leds and leads. in any case a several-way beautiful project :)) thank you for showing us.
I've seen that now companies are putting displays directly over processors which makes HQ display over a tiny surface take that and replace the circuit, the receiver and display with that one of the many problems would be how you deliver sufficient power to a rotating disc
Such a lovely design. I would paint the back with the blackest most matte paint you can find. Some new black blacks are amazing. Get rid of that sheen on the back. Great work! I want one.
Expectation: virtual candle. Reality: Star Wars communicator.
I can't wait until we have SW communicators. Really seems like you could just up the resolution/LED count to get a high quality image.
It need to be blue and have Princess Leia in it
@sam8404 Here’s the main roadblock as of now: real-time rendering. We currently don’t have the capacity to render something as complex as a human real-time. The other thing being that we don’t really have a "convenient" way of scanning someone’s entire body
"radios evolved so the user could change stations with the wave of the hand. which ment that if users wanted to stay om the same station they would have to hold a steady hand awkwardly and uncomfortably in the same position"
mixtela is one of those timeless youtube creators, who upload every now and again, but always produces something interesting.
it's mitxela apparently
i just noticed this
that or he sneakily changed the name to that
Never heard of him before but I'm interested
@@Psyden5757 No, it's always been mitxela (It's just his name backwards)
IMO more fascinating than interesting 😊🤝🏻👍🏻🇳🇱
Reminds of the early UA-cam days, when it’s was just regular or cleaver people doing things, now billion dollar entertainment corporations, treating UA-cam like the next cable channel.
Theoretically you can double your pixel density by putting another LED panel behind the first facing the opposite direction with the pixels of the second panel being placed in between the pixels of the first LED panel. So when it spins the second panel can fill in the gaps of the first panel.
This has quite the potential I must say.
Think this would also add the effect of whatever angle you view the cube at, you would see the same result!
"This has quite the potential" ...It's an inferior version of a monitor.
@@grabble7605 monitor has a pseudo depth to it, this is an actual 3D illustration that doesn’t involve VR technology.
@@grabble7605 wow, if we'd all think like you. We'd never have invented/discovered anything. " A bow and arrow you say? How is this better than a spear?" " A car you say? How is this better than a horse?" Maybe think a little more on the key word of Potential not on what it will be, but what it could be. Dosnt mean it's better now or will even replace the thing you think it will. Just that it's interesting, innovative and moving tech forward not stagnating...
@grabble7605 all the fancy tech we use today started like this. Remember that.
I'm kind of surprised how balanced it seems to be. It would be pretty brilliant to see this in a glass tube replicating the look of a vacuum tube or Nixie tube. Very impressive as always!
Or a snow globe...
I actually thought this was a Nixie tube at first. I definitely like this trend of modernizing the function but keeping vintage design
I love how this device obviously supports tracking yet no mention of this feature is made, as if that's trivial.
Its inherent with the design as the sensor needs something to measure a complete rotation. It's really cool and intuitive to use
@@alextasarov1341 The tracker also acts as on/off button.
Why would your orientation matter? It's spinning so one of the lights is always on. You'll just be seeing it in a different position in the rotation no matter where you're standing.
@@NarattoRadians two words: frame timing. the display needs to spin at a good RPM relative to the framerate of the display's refreshing.
ninja edit: well, "needs." it doesn't NEED it, it just looks almost immeasurably more... correct is the word i'll chose.
@@NarattoRadians because you'd be seeing a different angle
What I love is the speed control is the "rotation" control to rotate the "image". The slower it goes or the faster it goes, it rotates the "image". Such a simple design but absolutely brilliant!
you could possibly balance out the offset of the leds in the rotation center by sandwiching the leds between their PCB and a bit of acrylic. That way, no weight is needed, but the leds are still easily visible. Amazing project :)
Cool..
Frosted acrylic might look great, to smooth together the pixels, if you want that.
@@rich1051414I was just going to say that, add a diffuser as a counter weight
Exactly, yes, perfect Comment
Ideally with a Material with the same density should give the best balance and could also be used to protect the led s in between
In a bigger version you could also drill out holes in between the led s to reduce atleast somewhat air resistance for better battery performance
Or use a double-sided board (preferably with slightly offset columns, so you get higher resolution).
The fact that it spins so silently and stably is really impressive. Well done!
Does it, or is that just cause he put a voice over in post? Still epic though!
@@BloodAspit does sound like he did the voice-over live though
It's amazing what good quality drivers will do for motor whine
@@BloodAspin my experience, even the cheap brushless dc motors of that size are usually silent
I took a listen again, it does have feint sound, so I believe this is a live video, not a voice over, just a quiet motor.
Please don’t give up on this. This is amazing, and would love to see one day a higher solution, bigger version of it.
Reminds me of a Tamagotchi. Low pixel, but very clever and novel
China's on it don't worry
i doubt it will ever happen. the amount of movement you need is a safety concern.
@@genshinF2Play if its getting your finger caught in it you could just put it in a case, if its possible to make parts lighter that'll also help with making it shake less? or maybe improve the motor itself
There's one at the ATL airport. I think the point of this is that it has the potential to be small & affordable.
I love the fact you use a proximity sensor(?) as a way to determine the frame timing. Seeing it follow your hand as it animates is so cool!
TCRT5000 IR sensor :)
Wow. I thought the sensor was more of a trigger. Didn't strike me it helps with frame timing!!
@@MrVijayMadhavan Yeah, probably used as a sync reference, because the DC motor doesn't seem to have any connections that would allow the microcontroller to know when it completes a full revolution, so instead it probably measures the intervals between something being close to the IR sensor to determine how long is one revolution and to have a reference.
@@hundredfireify yep! Seems so
I wonder if a Hall effect sensor could piggy back on the magnetic field of the motor. But I guess this IR sensor is a satisfying on-off switch too.
the way it follows your finger while spinning looks so ergonomic, like controlling a hologram with hand-tracking (well, I guess it is!)
It's a truly digital interface.
@@RFC3514 heh
i have no idea whats happening lol. can someone explain
@@pvic6959I think he’s manipulating the distance from his finger and the ir sensors to do different things.
@@pvic6959 There's an infrared sensor on the side. The device uses the infrared sensor to check where his finger is many times a second, and then uses that information + some math to make the image face towards his finger
the added feature of being able to spin this a little with your finger tip is next level
I love how you implemented the ability to rotate the image using that IR sensor, makes for a very cool result imo
Well, since everything is spinning and that sensor has a definite location, it would automatically rotate the image while using the right speed as a function of the sensor location.
Yeah, the sensor is required to calibrate the timing of the image display. Moving that calibration point therefore moves the image "by default"
Adding a few degree tilt to the display should let you double the number of apparent rows that are swept through on each rotation, which might help improve the resolution 🤩
Smart!
I would love to see this with a full colour, tilted transparent OLED on each side.
That might help with the balencing problem on centering the LEDs as well. Smart design.
This is the right of of comment section. Some great ideas in here!
Maybe even have another display on the back face and stagger the LEDs both horisontally and vertically.
For the dumb people in the back, are you saying tilt it like a box onto its corner, so that e.g. the leftmost pixel in a row spins higher than the rightmost pixel and it's like it's an extra row?
1:22 does that look like that irl too (transparent / not noticeable rotation) or it's cause the camera fps?
it's because the lights aren't on i think
I quite like the "bonfire" effect over if it would be just a flame of a candle. But I suppose it might be worth trying to get that one done as well, for the practice at just getting better at programming and tweaking. This is super impressive. Imagine a set of 20 of these, lined up in a grid fashion, and flying on a drone. It's almost like a flying hologram(ish) moving picture sign!
Well, there is already some presentations using hundreds of drones and it is pretty much what you just described! Just search it up.
Yeah but the drones are not static, and they don't allow viewers to rotate the image, and the image density is severely restricted by the drone's bodies. Something like this we could probably scale up to 4k/8k image quality over time. We already have pretty high quality 2D spinning displays like this available on Amazon. By 90's standards this tech is already futuristic as hell.
@@bobriquardo5317 At first I thought even with one display like this over the body of the drone it still wouldn't be high resolution because the wings ocupy some space... but then I came to a conclusion, the wings are literally rotating motors! If the wings themselves were volumetric displays, then we would see the future my friends
@@brunogarbin6305problem is they always have to be facing up
@Roach_Dogg_JR they don't though, it's just a prototype, but flipping the image doesn't seem too much of a problem in the future
Recommend the each row contain an odd count of pixel elements with the center row right on the primary spin axis. Or maybe even better, put an odd number of elements on one side of the board with an even number of elements on the backside. This could be two boards glued back to back of course. Bottom line is that as it spins, it would appear as if there were no gaps in the pixel density. (note that the even board's rows should also be slightly off from the odd number to make sure the horizontal gaps disappear along with the vertical gaps. BRILLIANT display. I love it!
Overall adding offsets to this would allow it to fill space much more fully. It's interesring how such a great thing still has the potential to get even better
Thats really clever! From making pixel art i see tge advantage of odd and even number of pixels/voxels but putting them back to back so they fill the gaps and eachothers weaknesses is so smart
Now im wondering if its possible to add color by having 3 sides with different color LEDs. Of course thatd require faster rotation and break your idea, but itd still be cool and interesting
Also ruins the beautiful minimalism i think. But itis a direction this display could be taken in
No need to glue two boards, just have a double sided PCB made. Also being double sided guarantees alignment. He should also do RGB and as small as possible for maximum density.
@dougcox835 RGB ruins the minimalism but yeah
Another thing thatd be interesting is doing 3 sided display with red green and blue LEDs on each side, i feel thatd be neat. Itd have a side effect of slowing down rotation being a way to animate color change
i would pay to see this in real life. I watched some modern tech art galleries and this kind of stuff is amazing to see. Would be very cool also to see a double sized or tripple sized version, maybe smaller leds? I respect your work on this a lot.
ur profile pic got me, kudos to you :D
@@reezek3956 hah. One person per year!
@@reezek3956i hope that for a milisecond your face was like the face when you google just "oh you!". =)
I reckon you get a really high resolution hologram by basically doing this but using one of those SPI OLED displays (maybe a transparent one?) instead of the LED matrix board. Although you'd have to cover/paint it to make it yellow if you wanted it to look like a candle
I looked into this great idea, but unfortunately for this application its not super viable :( let's say you wanted your pov display to spin at 20 rps (20fps basically). These spi displays are usually 64x128. if you wanted reasonable pixel density, i would use the radial pixel count x π to find how many times per rotation you would need to update the display. In this case, the desired circumference resolution would be 201 screen refreshes per rotation (a refresh for each "pixel"). This resolution at 20 fps would require a screen framerate of around 4020 fps, and the theoretical max speed of these spi oleds is around 200fps at 8mhz. This math is for if the display was either offset or in landscape. If you centered the display vertical, you would only need 2000fps, but thats still unfortunately out of range:( I love the idea though!
A guy here on yt made a 3D volumetric display by stacking 10 of these slimer but longer oled displays they are $26 each and transparent it could work
I found a thing called Voxon, and it looks like it's already sophisticated volumetric display.
I really need to be careful what videos I click on.
This is one of the coolest things I have ever seen and I'm afraid it's going to re-ignite my electronics hyperfocus until I build one of these
I want one so bad but I'm also 14 so its gonna take me a couple years until I can figure out how I would make one💀
@@USureAbtThat I bet you're more capable than you know; start by finding a tutorial for a diy led display and you'll move up from there
@USureAbtThat You can do it! Research is key. Also, not being afraid to fail and try again. That is probably the most important part.
@@USureAbtThat 14 is the perfect age to start such a project! Don't wait for others, just dive into it.
@@USureAbtThat Study electronics in your spare time, if you have access to the internet then you have the knowledge of millions of people at your fingertips. Leonardo Da Vinci never had that.
0:34 other leds behind and trigger every 180 other way- and thin dark layer cover so we see only lights not circuits
Having done LED POV with the whole shebang - slip rings, motors, etc etc this is a very fresh and minimalistic take on it, very lovely.
This has to be one of the coolest things i've ever seen ! This little piece of tech is so cute yet so impressive i love it
0:57 can you get transparent circuit board?
That would look so much better I think... I hope!
The change would be only marginal.
Genius. Especially the fluid simulation, if you can eventually make it "real." Would sell like hotcakes.
Put McDonald's outa business 😂😂😂
Even this has me reaching for my wallet
@@davey2k12sure. everyone on diets will want to look at 3d versions of McDonald's food instead of being able to eat it and get fat. Lol
This is not a new idea. I made such a display in 8th grade in a science contest. Here is a video from like 8y ago from another youtuber:
ua-cam.com/video/8MFhoGbUYPY/v-deo.html&pp=ygUYbWljcm8gbGVkIHZveGVsIHNwaW5uaW5n
@@unorevers7160 did you miss the part where it’s the size of a thumb and running a fluid sim based on accelerometer data?
How about a super thin PCB and leds on both sides, would fix the balancing problem and give you higher resolution of you mapped the actual points the LEDs end up at. Really cool project, feeling inspired!
Maybe offset the reverse side to fill in the blank areas of the first. The wiring is a challenge I’d imagine, because a ribbon would probably be easier, but I’d imagine the current is not inconsequential.
@@silverXnoise Was thinking the same thing; like 2-axis interlacing.
Yes, stagger the rows for higher voxel resolution
You could make it 2 colour
@@andrewmacphail813 Ideally RGB LEDs. They do make very small ones these days.
This is so cool!!! Thanks so much for explaining can’t wait to see the next iteration
Cool stuff. For balancing, Consider adding weight 'strips' to the left and right edges all the way to the top, pointing outwards to the led side. Or conside adding a metal arc. If you attach it from bottom left to top right, it might average out the amount of light blocking, making it invisible. Alternatively, it could end up with significant interference.
Edit: the acrylic layer cover is a much better idea :)
Maybe another idea: if you change the board color gradient from the centre to the outside, to reduce the difference in color of the board due to the local speed.
For prototype I’d add 4 screws to the corners of led board. Rotating changes weight distribution, fine tune then fix with a drop of glue.
I might recommend having the LEDs be flat and reflected on a half-silvered mirror that's vertical, to mitigate the thing where the center of the volume obstructs voxels in the rear (creating an "opaque central column" illusion in the current version)
I'm trying to understand; where exactly would the half silvered mirror be?
Put the mirror where the leds are and put the leds flat on top of the pcb facing up at the spinning mirror
if i understand it right (and that's an ask) i think the mirror can be placed vertically where the led board is currently, the led grid being placed horizontally at the bottom facing upwards and the mirror's half silvering will create the 45 degree angle necessary to remove the columnar effect?@lolithighs
youre right, i cant make it work in my head either!@lolithighs
Impressive work! It's amazing to see how you've managed to bypass traditional challenges like power transmission in spinning volumetric displays. The rotating cube and fluid simulation look really cool, especially considering the low resolution. I can imagine how satisfying it must be to see it come to life after all the hard work. The next steps, especially with centering the LEDs and balancing the motor, will be exciting to watch. Keep it up-looking forward to seeing how this evolves!
written with GPT
Woah, impressive !
To get a more "candle" effect, the "flame" needs to be thin, and to dance around it's center on the lower part, letting the higher part going off the "center of mass". It will need your whole LED on the upper parts but not on the lower parts. And moves are generally erratic and quick, with phases of more stable moments.
Thanks for sharing this, this is awesome!
You seem to have assumed he hasn’t seen a candle before.
@@piefliesshut up if youre not here to contribute
I can envision many possibilities including RGB LEDs more tightly packed together, or even a high res video display instead of LEDs. Very cool video!!
yeah, in fact, half the time the screen is facing away from us, so why not put another screen on the opposite side showing a mirrored image of the first screen.
@@khenricxGpio limit might be a issue
could you do that by spinning a smartphone. use the smartphone accelerometer for orientation & program the display on the screen. maybe even spin it by hand? so the only hardware you would need is some kind of pivot for the phone
Nice! a rotary transformer is another option for coupling power, like they used in VCR head drums, though can get mechanically complicated combining it with the motor.
I suspect balancing at the bottom may cause some imbalance about a tilted axis. Maybe a thin PCB with some heavy inductors/ferrite beads between the LEDs could balance it fully.
Similar approach would be to add a larger solderpad above/around the leds and put enough solder to balance it.
This is amazing. I am no expert so discard my ideas if useless, but perhaps a clear circuit board to eliminate the black "fog", and a clear resin over the LEDs to act as the counterweight for the board after shifting it so that it can still be higher for balance? Getting that right would still be tricky, though. Also heat may become an issue if they are encapsulated. Really cool project! ❤
i wonder if a clear board and resin might transfer light much easier. you might have a full form image when spun with this method or it might just completely glow like with light throughout.
Would be really cool but glass PCBs are quite difficult and expensive to manufacture.
A clear board would expose the wiring, causing light to reflect off those wires unless the object were to be viewwed in dark lighting. This is a good train of thought, though, just needs more experimenting
@@HazelnutPi you can get transparent PET flex PCBs made quite easily they're just somewhat expensive and have some mounting related difficulties. The traces would likely be thin enough to not have a considerable effect though.
Not sure if it would help, but for the LED center-alignment, they make surface mount LEDs where you put a hole in the PCB and put the pads on the rear and the LED mount from the rear and their bodies poke through the PCB. This might help put the center-line of the LEDs and the PCB in the same place.
That is actually a very good prototype! The interaction built into it at such a tiny size is awesome, and I would love to see future designs with this, higher pixel density!
I do wonder if you would be able to make a frame of enameled wire so it's porous? In doing so it would allow for faster and be less power hungry along with weight and air resistance, as well allowing the LED's to be viewable at both 180° and 0° (may end up leaving a trail destroying the illusion)
Or even a see through PCB board but that's probably impractical
Best of luck on your future endeavours!
or, and hear me out, seal the spinning mechanism in a tube with negative pleasure, or better yet, as close to a perfect vacuum as possible, if friction is the issue here. the only issue now would be to minimize internal refraction.
@@loremaster63 I cant believe the solution to high tech display was vacuum tubes all along.
But imagine that, like a 8K display rotating at 120 rotations per second in a vacuum tube. Potentially two or eight of them sandwiched like a tree. You could get some insane games from that.
Or perhaps cheat a little and use holographic reflective or projection displays for HUDs or special items. Project it straight on the glass, or in their own plane.
It makes me wonder though what is more practical. Fiber optic and voxel based displays, since the biggest hurdle would be clarity and wire routing in the display. Versus rotational vectors, where the hurdle is rotation speed versus clarity.
It reminds me of those cheesy toys they use to have that created a 2d image from a rotating line of LEDs. It's cool to see the concept moved into 3d space. Pretty awesome results.
That tech is getting better and better. I don't think I'd call it cheesy anymore haha
I still think the old ones are cool.
It's yellow tho
any reason this cant work with a small LCD screen?
The technique is the same, they are both persistence of vision. The difference of course is that rotating the plane means led timing must be handled quite differently.
You can put LED's on both sides and holes in the spinning board as you decide which light conjuration you like. I'm sure a few people have already suggested this.
I wonder if this could be done with a small lcd or oled screen for extremely high resolution compared to what you have. If you can get the high refresh rate needed that could make it a lot more detailed.
I was just in the middle of typing the same thing when I saw that you had already said it. 😅
I doubt you could update it fast enough
@@mikeselectricstuff why? Im sure there are some 1000hz+ specialized small panels
How about adding more screens with an lower offset frame rate, and use a mirror to display it on the top...
I agree. You should be able to virtualize the present display on an LCD and then extrapolate the resolution virtually and add color changes. There must be some way to generate complete 3D high resolution images.
For the leds being off center, you can try looking into alternative style LEDs that mount "reversed". They solder onto the "back" or the board, and shine thru a hole in it. I think they're sometimes referred to as "gull wing" leds
This is amazing! Very nicely done! I guess next step can be a smartwatch display, then a smartphone one… with higher resolution there will be so much more possibilities
That is Awesome!
I think you might be able to use the backside, with offset diodes (between the front side), to create an interlaced so to speak resolution. ▒ So that the back side fills in the gaps between the diodes in the front side. Should also balance well. ♥
Edit: I wonder if a 2 boards in an X configuration would work? (double sided boards)
Oh great we're going back to interlacing 😂
Ah great idea. I had just written my own comment about this, stating the same idea. Just to come to the comments and see that you beat me by a week. Would love to see more prototypes on this.
You can either put the lights offset and get an increased spatial resolution that's interlaced or put the lights in the same spots on the back and get an increased temporal resolution ie framerate
Lol. At least 3 of us are on the same page. Yes i would love to see him use the back side too. Btw you beat my comment by 1 day
You intelligent boys xx
Very neat! For a relatively simple assembly it works surprisingly well. A double sided LED PCB with the LEDs offset on one side to double the density might be doable, though doesn't help with putting the LED die exactly on the plane of rotation.
4:25 instead of adding weight and all why don't you create another board and place it back to back with the first one it'll increase frames per second and balance as well
This is so fucking awesome. The fact that it detects your finger and rotates with it adds more realism for some reason. Its amazing
Does it detect or is he just adding friction that slows it down for a little bit and changes the orientation?
@@zrman96 The little rectangular bit on the side is an IR transceiver. When he puts his finger close to it, the receiver picks up the reflected IR light off his finger. Putting his finger on it wouldn't change the orientation, only the perceived framerate.
@@zrman96 I believe that it is detecting it but im not sure. There is an IR sensor on board thats probably for detecting how close your fingers are but idk if its used to detect the placement of your finger. Maybe he is only friction and im reading to much into it lmao but i didnt hear any sounds when he "touched" it so i assumed it was the IR sensor doing the work.
Edit: look at 0:56 ! I actually thing the IR sensor is detecting the position! He's not touching it and the images followed his finger. Thats so cool
@@bluebaconjake405I'm guessing that the ir sensor is used so it knows when is has done a full rotation so it can stay still instead of turning if the speed is a hair off. By moving the thing the sensor detects (the finger) it changes where the display thinks forwards is.
@@masonbarber871yes, and that is the reason for the off-state if the finger is too far away: there's just no "frame" start reference (zero degree marker).
I love this idea and the fact you built a working prototype is really amazing! Thanks for sharing the results with us. I was thinking about the issue of the LED's disappearing when you are looking at the very edge of the PCB. Maybe an "L" shape (top view) would help? I suppose then it would be blank for half the rotation (which is worse). Maybe a double sided L shape or a one sided plus shape (like notch cutouts locking together)? My other pondering was that the square LED's are affected by perspective as they turn. I don't have any good ideas how to compensate for that with an image unfortunately. An LED would appear largest straight on and get less and less wide as it turns toward 90 degrees basically. It may be too difficult to change the brightness based on location of the pixel to match better (but it does seem like you enjoy a challenge).
Well done! I had a similar idea a few years back, but it was a lot bigger and it would just wobble out of control as soon as a turned it on so I gave up on it. I thought of using circular rails that wires from the spinning display would brush up against, but this design is so much better. Thanks for the inspiration!
If you could adjust the brightness of individual LEDs then you can create tonality and additionally gain the benefit of depth which in flame effects could be very realistic. The glow of the flame and then bright sparks or wisps of light.
Also the use of RGB lighting could also add another layer of realism. I actually think this is a great concept and with some polish design could be a great product for candle-light like lamps.
Very cool.
Love the design, honestly great work. It looks difficult to scale up, but I think an array of similar devices could produce some very unique effects.
I'm filled with so much envy for people like yourself who make these beautiful electronic artifacts. Well done on making it so nice!
I am absolutely captivated by this prototype, I would love to have something like this in the form of the bracelet to show off ideas or to analyze my 3D models in my free time.
I love projects like this! Great job! Can’t wait to see where it goes.
I have two ideas to improve the work, you can use the Tesla tower to give energy to the device, and since the coil is self-similar, the problem of instability will disappear.
And the second point is that you can put this inside a tube without air to eliminate the air resistance.🤔
Awesome! Having the LEDs randomly scattered on the board could help deal with the pixelated look. Although it will probably add complexity to programming the animations.
Subbed! Would love to see your progress on this beautiful device. In a world where people expect every new technology to feature some kind of “AI” it’s refreshing to see someone simply come up with a new practical physical concept.
When you're tweaking the candle flame parameters, please keep in mind that candle flames only flicker constantly if they are in a draft. Most of the time they are quite steady, perhaps with a slight lean to one side or another.
very cool and interesting stuff ,as for centering the led's i suggest adding a clear layer over them that weighs the same as the board that is on the other side thus having the led's sandwithed between the two sheets and having a good centered mass.
I think using a higher pitch led matrix would be perfect for such an amazing concept, and I would love to see it more developed into a final product... love what you're doing and definetly you gained a new subscriber :D
One of the nicest small and relatively cheapo electronics projects to get going on, also would be an amazing addition to a number of larger projects. Great work and thank you for the video :)
to eliminate the central bar (the axe) that is too visible, you should have two system face to face and have leds on both side for brighter display. The display would be between the two axes. Great idea and realization you did, go on !!!
I wonder if it might work to wire up a grid of LEDs not on a PCB at all, using as thin of wires as possible for the current needs, then cast the entire matrix in resin. This would give you a mostly clear "screen" that is able to shine in both direction.
That's a sick idea. I wonder if the dispersion of the light through the resin would add to the effect or blur everything together?
Problem with the grid idea is the LEDs in the middle may be hidden by those in front of them that are turned off. But I think it has been tried and works ok for low resolution bc then the front don't block as much of the ones behind but high res won't work.
I had a similar idea but I'd spin a special semi transparent film that could still display projector light, something akin to holloween nets or a film I have not discovered yet. Then use dual projectors to project on to the spinning screen. The images projected when match the rpm of the screen, the projectors would be 90 degrees from one another for full time coverage.
Absolutely phenomenal - these are the sorts of creative endeavours that push technology to new frontiers. Massive thumbs up!
Thanks for sharing! I've been really interested in Holographic displays as I seem to be noticing more show up.
The website is very well put together and had a good time reading the steps you took on how you assembled your first Candle Hologram Prototype.
You rock! Keep up the great work, I look forward to the next higher-resolution prototype!
You could also consider a mechanical shutter effect. 35mm film projected the image at 24 frames a second , twice on the screen every second with a timed shutter to clear up the image.
Imagine this with an lcd or better yet transparent display at a larger scale (might need to be in a vacuum for less resistance)
I was literally about to type this, how amazing this indeed!
Not sure you'd ever be able to get the update speed required for a decent rotational speed.
@@ferrumignis the question will be both how big and how fast you are spinning the display we have displays that can go up to 300+ hertz but the problem is cooling as they can make a lot of heat (although the spinning might help…)
@@ferrumignis I mean if you optimize it for airflow maybe right?
@@piman13_71 Can you really get tiny displays with a 300Hz refresh? I'm very interested, what interface do they use? (presumably SPI?) Could you point me in the direction of a manufacturer? The usual tiny LCD or OLED displays can't get anywhere near that.
3:09 : Can I suggest, as an artist 'perspective' is bleeding here. So if you can move one half of the LED on the other side of the board, thus it makes one complete rotation as single frame. Or build a board that has LED on both the sides, where the gaps from one side the board is filled in by LED from the other side of the board. 😊 hope it's not confusing 😬.
Only watched about 1:30 into this video, and its the first one from you I have seen. I get the feeling I am listening to someone who could teach me a lot.
Wow, truly amazing work showcased in this video! Your approach to integrating technology and practical applications is impressive. I was wondering, have you considered incorporating an Inertial Measurement Unit (IMU) into your setup? The addition of an IMU could provide precise positional tracking, which would be particularly interesting when you're demonstrating the fluid dynamics inside the container.
some really simple animations might look great, like animating a ring up and down and changing its diameter. The resolution seems a bit low for things like "flame" or "rotating cube" but I bet some slightly simpler animations would look fantastic on it
Haha, low resolution for a "flame" or "rotating cube" as he proceeds to model sloshing liquid in a cylindrical container
@@gw6667 yeah I liked the fluid one too! I guess it's one big object moving so the detail doesn't matter as much compared to a flame where it's all about the details
If on the middle an led could be on the very middle top of the display it would really add to the candle effect but the a hollow middle row or one slighter thicker cylinder below what I’ve suggested could also have LEDs all the way down to make the axis’s shape more active but I guess becoming thicker but shouldn’t mess with balance if centred well. May even improve stability. Great effort as usual and a wonderful prof of concept.
This is incredible! I hope you're able to improve it and create larger-scale concepts. I'm just imagining sculptures you could display in your home and change on a whim. Well done!
You forgot the fact that anyone who walks into a larger one of these dies.
@@RZ302 a glass/plexiglass enclosure would fix that pretty easily
@@RZ302 They earned it walking into a large spinning sculpture.
i cant wait for someone to run DOOM on this thing
Amazing idea. Maybe add a layer of transparent material over the LEDs to balance the weight? On the corner might cause the axis to bend overtime.
This is incredible! Thank you so much for sharing and the writeup. Very detailed and interesting stuff! I can definitely see a future version using addressable LED matrix, something like the high density screens adafruit offers.
couldn't you inset the LEDs upside-down into the board, so they're on the same plane as the board and centerline, without needing additional balancing?
Hard to do with regular leds, but there are leds packaged for exactly this with their legs poking out. Might prevent loading any components on the back though.
you could balance it with a plastic transparent sheet of approximately the same weight as the LEDs support, glued on top of the LEDs, and then place the LEDs centered on the rotation axis
3:31 is it following your finger ?
This is super interesting, I would be curious to see what this would look like with an old cellphone screen or any small pixel based screen in the place of that LED array!
I watch this video so often. This is so awesome.
Damn, that's a neat display. Cyberpunk "Nixie Tubes", here we come!
edit: finished the video, I see we had the same idea!
pretty genius, especially i like the part you did the starting process and 3D handling, well thought.
Add a thin piece of transparent (translucent?) plastic to offset the PCB weight, to allow the LED's to be moved to the midplane; also make the PCB single sided and sand the PCB down to make it thinner/lighter, or print it on flexible kapton with cutouts for the LED's with two rigid wire towers for support. Maybe have flexible tabs for each end of the LED that fold down perpendicular, and use a tiny jig for soldering the LED's to the bent tabs.
I like this project a lot! I liked the idea of spinning all components and the IR sensor to sense the orientation. I'm curious to know if it is possible to use a mini LCD / OLED display instead of the LED matrix, in order to get much higher resolution? Great work.
The pow persistence of OLED might be preferable here, would give you a less fuzzy result on the outside especially if you spin it faster.
Hi, love your work. Would it be possible to use a LCD/OLED screen as the light source ? They are often quite thin, and if you use a 3D-printed jig to align it, then it may be possible to a chieve a very nice volumetric cube. (But battery could be heavy, maybe use a low friction slipring ? small solar pannel and laser ?)
I wonder if you can get one with a high enough refresh rate. A pixel has to update for each voxel it passes through, per frame, assuming there's one frame per rotation (there might be a minimum RPM required for persistence of vision idk)
There are some clear OLED screens, like the ones Sean Hodgins used for his TENEX volumetric display, that would fix the issue of half of the updates facing away from the viewer, which might decrease the necessary RPM? which would in turn increase the spatial resolution for a given temporal resolution, I think. Couldn't find the refresh rate for those displays in the specs, though. It might not be high enough to make up for the double visibility.
Do you need to center the LEDs? Couldn't you use it to your advantage and do some sort of sub-pixel rendering with it?
This is absolutely insane, I don't think people get just how cool it is considering you can give this to a professional brand and they can iterate and make it high res + rgb.
and put it on a transparent board too
You’re the best man. I love everything electronic, circuit boards and their components and what you can do with them. I thought you led ear ring you made. That tiny flexible pcb and those tiny leds were so crazy cool
ironman vibes
This is incredible! I'd love to try making something similar. I love the use of an infrared proximity sensor as both a switch and feedback for the rotational speed, absolutely brilliant! Keep up the great work!
the weight of the offset led mounting panel can be balanced by a clear panel in front of the leds... more ideal but harder to prototype would be a clear panel with leds and wiring embedded in a clear panel, possibly a poured resin that then picks up the leds and leads. in any case a several-way beautiful project :)) thank you for showing us.
Shave the top corners off too. If your going for candle flame you dont need them and it could help with your stability
Maybe have leds on the other side?
I've seen that now companies are putting displays directly over processors which makes HQ display over a tiny surface
take that and replace the circuit, the receiver and display with that
one of the many problems would be how you deliver sufficient power to a rotating disc
Such a lovely design. I would paint the back with the blackest most matte paint you can find. Some new black blacks are amazing. Get rid of that sheen on the back. Great work! I want one.
I think I speak for everyone when I say
Y U NO Bad Apple?
It's pretty much the law!
Amazing! Great result!
I bet you spent a lot of time with rotation matrices or quaternions to program this thing!
should put this on Something like PCB way so people can buy it already assembled and soldered for those who have less experience with stuff like this
This man created cephalon simaris before GTA 6