Corrections and FAQ in this comment! 0) Stuff I want to shoot with this camera (adding suggestions to list): Race different transmission lines vs fiber optic race light in clear tube of air vs clear tube of water refraction, bending and slowdown of wavefront patterns with lots of mirrors somehow?; double-slit interference interference with one beam delayed (really excited about this one); infinity mirror with an object in the middle that would get cloned slowly looking off into mirror-world short pulse to get "star wars blaster" effect (this might require buying a different driver!) shine the laser on a fluorescent material large lens (I have an old projection TV fresnel in my parents' attic I could pull out for this!) Electric arcs - this would have to be a macro shot, so observing actual light would be impossible at this framerate, but if I could get an arc to form extremely consistently, i bet it doesn't actually FORM at lightspeed - you have to chug some ions around through the air gap. this could be really awesome! Light solving a maze or shining into an penrose room. flash bulb in fog - like that one nebula light echo timelapse long stretch of fiber optic do a time-delay process with the camera and light in the same place to get a TOF depthmap two more different color lasers to do full-color imaging and/or refraction One of the standard computer rendering things like cubes in a mirrored room, or that one teapot or something to do “ray tracing” irl 1) This isn't a "stop motion" technique - it's more like making a 1Bfps camera that's 1x1 pixel, then tiling lots of videos next to each other. I did consider using a Pockels Cell in front of a regular camera to try to do stroboscopic stop-motion-style data capture, but I think it would have been more difficult/expensive than what I ended up with. 2) I've gotten lots of great control recommendations - a few really high-precision servo brands, and things with spinning mirrors/lead screws/stuff like that. My first concept for the camera was actually to mount my one pixel behind a static lens on a 3d printer gantry and raster it back and forth to create a "virtual" sensor. I didn't do this because In the garage (a relatively tiny space) i needed a really wide field of view and from the cyanocamera project, I knew maintaining focus at the edges would be impossible. The spinning mirror is really appealing, but "most" of the time is spent with the mirror looking the wrong way, so I'd be multiplying my scan time by 3-4x. it WOULD give a more consistent image though! I'm also about to look up a galvanometer. 3) most of the scans i scaled to take an hour or two. the more averages you take on each pixel, the better you can resolve the "darks" and get illumination in the room that isn't the primary beam. If i only wanted the primary beam and a noisy signal everywhere else, I could easily get away with one laser flash per pixel! (this gets better if I overcrank the diode, but we all saw how that ended) 4) There are a lot of comments asking about the "early" light from the laser, the fact that the beam is traveling at an angle to the camera, etc. if you look, the first half of the beam moves really fast and then it slows down as it starts moving "away" from the camera. this is a sort of doppler effect, and it's an unfortunate limitation of the size of my working area. it's a really cool effect though, and one I want to discuss more in a future video! 5) I DID use fog for the disco ball! in fact my fogger caught on fire so I got bonus fog... That's the only reason the "streaks" were visible. 6) The part about adding an aperture being similar to a "real" camera lens' aperture may be a bit muddled. in a regular camera, the resolution is set by how far apart the light sensitive pixels are on the chip, but in this case, my light sensitive area is probably more than 1 cm^2, so the aperture restricts defocused light but ALSO light that comes in off-angle. 7) How many subscribers do I need to drill a hole in my garage door? let's say 10M - that may or may not happen one day in the future xD 8) a few commenters have very accurately pointed out that the "start recording now" signal should arrive at the oscilloscope AFTER the laser light does - this is correct! thankfully the oscilloscope can back up slightly from its trigger point, and that's how I'm doing it. it really doesn't matter when i see that signal, as long as it's the same for every single flash I can use it to synchronize the data. 9) I don't think there's any simple way to make this camera faster. the scope is 1GHz sample rate, but actually 100MHz for signals, which means ~3ns rise time is all I could hope for. The laser driver ALSO is in the range of units of nanoseconds, and the PMT is also in units of nanoseconds. in order to speed this camera up, I would need a new light, a new detector, and a new data collector - basically a new everything. It'd be better to switch to a legit streak camera at that point. maybe someday. 10) a lot of people are suggesting that to see light move, the camera would need to go “faster than light” but that’s not the case - you can take a video of a supersonic jet with your phone without any part of your phone breaking the sound barrier 11) this cannot be done reasonably with mechanical shutters. If you have the laser going through a 1mm pinhole and you want to slide a shutter out of the way to turn on the light in a few nanoseconds, that shutter itself would need to be going 1mm/5ns = 200km/second = 124 mi/sec
A pulse of light shorter than the length of the room it is in? edit: Someone else has also mentioned a mirror maze. That would certainly let the path get much, much longer without taking too much space.
Laser refraction when it hits a tralie and/or the reflection of lasers among several reflective cilinders (or mirrors placed almost parallel facing eachother) to see if you can see its path occuring like a slomo lightning strike. Such a tralie could be moved reproducible using some piezo element and change the light angle in a reproducible way.
Your gimbal solution is to use an electric telescope tripod.. you can find some really nice ones with "high resolution" motion control quite affordably. And they plug into computers quite easily.
A focused white Lightbeam that passes a optical prism and gets divertet in it‘s spectrum. And maybe a bldc with a Magnet at the end and a HAL-Rotation-Sensor might nie better as a Servo? The Resolution should be much higher.
@@the_boogey_man0 He loves you too but this is the third time he's seen you speak for him out of context and he thinks you're lucky his patience is infinite.
Absolutely breathtaking! I wonder if you had a coil of bare fiber optic that was just glass with no insulation, if you could see the light looping through the coil
@@christianblack2916 Or perhaps roughing up the surface a bit could work. The total internal reflection only works when it is spotless on the outside, as the reflection happens on the interface. I think this is totally doable with a little elbow grease.
There’s a video I saw somewhere in Reddit with oodles of upvotes that was called “the speed of light” or something that showed two guys in lab coats and a very long led strip light Illuminati g from one end to the other and the fact that people were eating that up made me really sad
Really gotta get rid of all that Microsoft stuff, and more generally, avoid any subscription-based, bloated, always-online software. Free, lightweight, offline stuff is much more reliable and doesn't randomly peg the CPU or fill the memory.
@@ToyKeeperoffline stuff completly use cache memory to read and use the data you want. For the web app 100% with you for the speed, lightweight, so much time saving.
it's uploading to cloud and spying on you ... this is why i dont use any default software or setups, it always spies on you. that's why i've always used 3rd party firewall control that blocks ALL connection attempts and gives me alert first, there's at least like 100 things on windows that connect to internet without you knowing about it.
I think Gav is working on filming light over a much longer distance! (and in a single shot, not strobe-ish like I did). He mentioned it in a recent video with Adam Savage and I really hope he gets it working cause I really want to see it!
I know this is a joke, but I actually did some research on why they are so fuckin awful and it’s because they constantly record which requires a shit load of storage space so making it less detailed allows for more to be recorded before they gotta dump footage
@@TheAustinTalbert That was true in the VHS days or when a normal PC had like a 32 GB hard drive. Now you could fit months worth of 4k footage on consumer grade drives and racks. They're just lazy.
@@TheAustinTalbert "shit load of storage space" A proper full HD camera generates a video of a 5 Mbit or so. But let's step it up to 4K, maybe 10 Mbit. That's 100 GB per day. Let's say they have 5 cameras, adding up to half a TB per day. An 8 TB disk (less than $200) can store 2 weeks of video. There's really no excuse anymore for having a low resolution security camera.
The hole in the garage door and the mirror on the black screen becoming the "hole" reminded me of the simple gags in This Old Tony videos. Especially the way he doesn't acknowledge the joke but just keeps going on with his presentation.
@@eklhaft4531 Yeah, it actually appears that the "mirror" was just a green screen or maybe there is nothing there at all and it is just an inverted mask, but at any rate the "mirror" appears to be the scene behind rotoscoped onto it, since if you look at the "reflections" in it, they are not of his garage as a mirror would be. Very cool idea.
@@dirtdart81 I think it's a composite, where half the scene is an underlaid static image and then the "hole" is simply cutting a hole in said image... I unno... It makes sense for awhile, but I'm pretty sure that entire car explanation scene is still using that camera trickery, and my idea falls apart.... 🤔 Unless it's ONLY that square that's that static object... 🤷♂️
I am fairly sure that hole was already there during the entire shot, and just covered up with black until he places it. The hole he grabs from the box and places there is either nonexistent or does exist but is, either way, tracked with some footage in the shape of a circle. If it does physically exist, he might have had to paint it out when it covers up the real hole in the board until its removal from frame (possibly it's dropped when "placed"?).
Suggestion: If possible, I would love to see a when natural sunlight passing through a prism. It would be fascinating to observe the dispersion of the light spectrum at such a high frame rate!
@@ianglenn2821 He could also measure the light twice, once with an optical filter in front of it to see the blue light and once with an optical filter in front of it to see the yellow light. That way he has two colors.
@@ianglenn2821That's a clever idea to get different wavelengths of light! But do you think the light emitted by the fluorescent material will ramp up fast enough?
Unfortunately, that's not really feasible. To use natural sunlight, you would need a mechanical shutter (since you can't just turn the sun on and off.) The mechanical shutter would need to move impossibly fast to get a small enough pulse of light with a fast enough rise time for him to record.
@@hans-jurgenvogel6789 the delay between absorption and emission that happens in fluorescence is in the range of nanoseconds, so it should be a noticeable delay, but still in the timeframe of that video. Since the camera is monochrome, he should probably try out my using filters to create two videos for two different colors, like I proposed in my other comment.
I set Python to rendering that clip with all the black circles and went to watch Shane’s video while I waited. Two minutes later I saw the same graphic. I laughed so hard 😂 I guess there are only so many ways to describe a camera. I really like his ant analogy
IMHO the nanosecond camera is cooler, although his wife probably wouldn't want to stand still for an hour while being blasted a thousand times by rapid laser flashes. Mount it on a rotating arm next?
Man, Alpha Phoenix videos are so goddamn cool. Sometimes I can’t believe we have someone just hanging out, on RUclips, literally building cameras fast enough to capture the movement of light, or sensors to track the speed of atomic movement in a piece of rebar, systems for displaying electric wave propagation models, and horrifyingly powerful gerrymandering algorithms, and doing it just hanging out in his cave, with a bunch of scraps. To have someone this smart, exuberant, and good at presenting such wildly disparate topics in such understandable ways (without dumbing it completely down), with some of the coolest homemade contraptions I’ve ever seen is awesome. So thanks for putting out the best videos on this site!
"... good at presenting such wildly disparate topics in such understandable ways (without dumbing it completely down..." this is exactly my thoughts. The mentor presentation style is 11/10
Yeah, this channel, @stuffmadehere and @breakingtaps are the best at this at the moment. They have such cool ideas that take months to figure out, build and film.
Better project than 95% of our Master students finish with. And without guidance.. Awesome! And results are just amazing! As an alternative to PMT you could buy a cheaper silicon PMT, or Si-PMT for short. They are not as good as but way cheaper(~50$), far less dangerous voltages and smaller change to kill coupled electronics. Also funny how you darkened the room. Reminds me how we did the same for some long time measurement with actually photon counting... Every LED, every possible light from outside. All were sticked with a black tape. It is amasing how sensitive eyes are after being an hour in complete darkness :)
Small hint for the LIDAR sensors: Most cheap sensors don't actually measure the time the light takes directly and instead the shift in phase of intensity modulated light (which takes longer to measure but is way easier)
@@AlphaPhoenixChannel also - possibly dont control the sensor movement, esp not stepped - give it constant movement (with rotating surface coated mirrors) and measure it like LIDAR sensors/oldschool lissajous lasers - could even overlap the smeared pixels for better resolution via interpolation if it moves fast enough
And LIDAR doesn't measure the intensity of the reflected light. Just the phase shift. It can do this by measuring how much constructive/destructive interference there is with the reference signal. Using the phase shifts for multiple different modulation frequencies allows it to estimate a distance
@@bornach trying to understand this :D so the phase shift occurs due to light's travel time? for different frequencies the phase shift differs because each frequency is at a different point in it's respective phase when it gets reflected? wouldn't that give you a resolution on the scale of the wavelength, ie. nm? how do you derive distances on the scale of meters from that?
Digging the hole out of the box was so good. Didn't give the joke away at all, just played it off like it was a real thing, and the effect was really well executed. Loved it. Cool little nudge to Corridor later on too.
Frankly one of the best RUclips videos I've seen in a long time! The enthusiasm, the explanations, the outcome, freaking amazing. I just feel like going to my office and start tinkering with stuff right now!!!
Smooth brain question here: So does this video finally prove that I should start cleaning out the photons from my android phone camera to keep it from getting clogged up over time or that I we all need to start finding a way to manufacture the ultimate surf board that can lets us start catching light waves? All jokes aside, does this video prove one or the other in any way??
If you actually understood light youd understand that the camera is limited to the reflection and process itself physically, the light moved far before the camera actually picked it up, your basically looking at a 1billion fps reflection, the light moved before that
@@romankozlovskiy7899 If you are replying to me, then I will start by asking you to name me 10 people who actually "understands" light, and then I want you to compare that to all the people that dont. Then consider that your perception of who "understands" light is very "light". Also, great of you to follow up with extra info about the behavior between the camera lens and "light".
Couple of ideas: 1. Any chance you could switch to a white light laser and repeat the experiment with R/G/B filters in front of the detector? Having this shot in color would be amazing! 2. How about a light maze with multiple mirrors just allowing enough spill on the side of a background plate (that the mirrors are fixed to) to have the progress of the light travelling visible? 3. Could do a split beam perhaps with one half travelling through another medium (glass? water?) and have it race against the light of the other (half) beam. Amazing video, mad props!
@@jdr91 Yeah, what do you not trust? You believe it happened, right? And the maths are correct enough? (that's me pluralizing "maths" to make myself sound European and, somehow, smarter)
Just a humble advice: considering the impulse response of your detector is deterministic (which it seems roughly to be), you can apply a temporal deconvolution to recreate a much faster response time sensor. A basically instantaneous light sensor detector with a ramp-up time of ~0ns. In the end your video will be just your laser ramp-up. If your laser ramp up signal is deterministic too, you can deconvoluate your laser too ! A double Wiener deconvolution can in theory bring you back to a
I'm always impressed by your demo setups. They intuitively show advanced concepts in a way that is achievable by a normal person. What I would like to see: - White light getting refracted through a prism - Light passing through a fresnel lens (or any lens really)
That video was mind-blowing! Whouah!!! ...!!! Great practical science! ❤😀😀 Suggestion: round trips between almost parallel mirrors. With a bit of fog to "see" a fraction of the the light. The laser would enter the cavity between the mirrors with an angle. One issue will be the to ensure a constant density of fog during the duration of the scan. Maybe without any fog, and just the humidity of the room will suffice. 🤞
@@kid_missive If you want to prove it's a quantum effect, yes. But you still get the same interference pattern regardless of if you use single photons or a lot of them.
You should definitely try to focus the beam and make it as thin as possible without loosing detail on camera, then bounces it between multiple small mirrors around the room to see it zig zag between the mirrors! Would be super cool seeing the lights travel from a to b like that, never seen this type of footage before!
Great work. Can you get video of electricity arcing from a negative electrode to a positive electrode? That is something I would love to see (or actual lightening if it can be done safely). You should place several different solid color squares against the wall you are recording to see the differences. Under water, the deeper a diver goes, the more colors all end up looking grey (red is the first color to go grey).
I think that this is unfortunately not possible with this technique. You need exatly the same illumination thousands of times to composite this footage.
Oh you genius bastard. It’s a rotating lens connected to an oscilloscope into the inverse of a Matrix bullet time. Have you invented a whole new camera? Give this man an award.
Alas, deconvolution is ambiguous. The motion blur has decreased the SNR in the high-temporal-frequency content of the video, and there's no way to get that back other than guessing.
@@krisztianfekete3277 Alas, convolution->deconvolution is only a round-trip with infinite precision. With finite precision it irreversibly loses information, as I stated. (Note: there is a much simpler argument that _almost_ works here - consider a 2-bit color depth with a 3 frame box blur, and the input signals [..., 3, 0, 0, 3, 0, 0, ...] versus [..., 1, 1, 1, 1, 1, 1, ...], both of which result in the exact same output signal. However, this only fails in the infinite-sequence case. So I'm going for a different more convoluted (hah) counterexample instead.) Consider a very simple example: a single pixel; 2 bit color depth (0, 1, 2, 3) with a five frame box blur. That is, a blur function of [1/5, 1/5, 1/5, 1/5, 1/5] in the temporal dimension. So e.g. [0, 5, 5, 5, 5, 5, 0, 0, 0, 0, 0] becomes [0, 1, 2, 3, 4, 5, 4, 3, 2, 1, 0]. So far so good. Now consider the input signals: [0, 0, 0, 0, 0, 0, 0] [0, 1, 0, 0, 0, 0, 0] [0, 2, 0, 0, 0, 0, 0] What happens? Well, in infinite precision the answer is easy: [0, 0, 0, 0, 0, 0, 0] [0, 1/5, 1/5, 1/5, 1/5, 1/5, 0] [0, 2/5, 2/5, 2/5, 2/5, 2/5, 0] In 2-bit color depth? If we truncate, then these three both map to exactly zero, and we've lost information. If we round, then these three still both map to exactly zero, and we've still lost information. If we round stochastically (so e.g. 6/5 means "1 with a probability of 20% and 2 with a probability of 80%")? Well, I just did that ([random.random() < PROB for x in range(5)]) and got [0, 1, 0, 1, 0, 0] So what's PROB? Answer: it could be any value 0 < PROB < 1. So from 1/5 or 2/5: both work with finite probability. And so we've lost information.
Thank you so much! Really appreciate seeing someone just genuinely interested in scientific experiments with an engineers heart and so much curiosity. And you show and explain all this with so much passion and focus on what you want to communicate without this - I don't know how to explain it.. this feeling that I get from some other RUclipsrs that somehow manage to make it feel it's more about showing off and making it about themselves rather than science. Like without any "narcissistic vibe" if you want. Best wishes from Germany, looking forward to seeing more like this! :)
@@entertainme121 Well, I posted that comment before watching the video. So, thanks for the summary, I guess ? If you had insane rolling shutter (in that case on both axis) while filming a way too fast periodic thing (in that case a multiple frequency of the framerate). Would it make it not "slow motion" ? When did it stop being "slow motion" ? (I get your argument, I just want to be a pedant.)
this is so great! i remember seeing that coke bottle light and researching how they actually did it. i love that you have basically redone that with cheap components
This was my thought as well. Being able to visibly see different speeds of light through different mediums would be neat. I wonder if you could work in more, like air, water, and glass. Or are there other fluids which slow down light even more than water/glass?
He'd have to improve his time resolution to the sub-femtosecond range. His setup can resolve propagation differences on the order of a meter, but optical interference happens due to path differences on the order of 100 nanometers. You need attosecond streaking or some fancy optical cross-correlation methods to visualize the temporal phase of a light wave.
You could probably film the light passing through a diffraction grating wit the setup, but I don't really know if you would see anything interesting (but it would still be cool!)
@YSPACElabs it's monochromatic, so no. Even if it was white light, the speed of the different colors is virtually identical in air. It's a different story in water, though.
Such a cool demo! 🤩 So happy you got it working, that footage is just mesmerizing. Seeing the rays bounce off the disco ball was amazing. I would watch literally anything and everything filmed with this setup, please keep them coming!
Oooh! Okay imagine this. You have a clear tank with a hole in the side and a laminar stream of water going down into a cup. You shine your light through the tank and the light follows the stream of laminar flow water into the cup, so you get all the refractions of light from the container and the cup, and you can see the light working its way down the stream of water. Now, you'd need to have some sort of a pump configuration to take water from the cup and constantly replenish it in the tank in order to keep the levels stable so that your slow scan imaging isn't thrown off by changes in flow rate. But if you could pull that off, that would be the most epic video ever. And I know SmarterEveryDay would love it. Destin might even be willing to make a collaboration video with you to hammer out all of the engineering problems dealing with the tank, and improving the precision of your scanning equipment.
This is possibly the best science channel on youtube. Putting together real experiments to show unintuitive things in a practical way. The stuff like showing how electricity literally moves through a wire, when all the creators were bickering over the theory of light waves. That was an awesome video, and this is another amazing banger. Ever since I was a child, I'd imagined what it would take to literally see light move across the room, and I'd tried to mentally understand what happened when I flipped a light switch. Seeing the experiments gives me such an amazing visual that I wish I'd been able to see as a kid. It's an answer to all the questions I had, and more. Seeing the disco ball lights bounce around was amazing. Even though I tutored college physics, and I understand all the theory and formulas, it's an entirely different thing to see it with my own eyes. What's even more impressive, is that you're able to build this setup with a reasonable budget and tons of knowledge. Definitely the best science channel on youtube. The only one that's doing something different and something genuinely impressive. You're probably the closest we'll ever get to a modern Richard Feynman, with your ability to so clearly explain complex topics. You're an inspiration.
as a nice round number: 30cm is just 207.5 micrometers shorter than a light-nanosecond. In comparison, 1 ft is 5008 micrometers longer than a light-nanosecond.
You put the anti clickbait video at the beginning and then give us the full show at the end. Oh, you tease us! And we like it. I do know that every time I skip to the money shot at the end of AP videos, I end up jumping back just a bit for context of the things. And then a bit more. And then jump back to the beginning to make sure I understand the interaction between different parts. And then look for the test setup. And now I know it's actually faster to wait until the end of the video than to skio to the end and watch backwards and then forwards again. But seriously, we love how technically dense and well ordered your scripting is while walking is through concepts we have a basic understanding of or potentially none at all. And getting us to an experiment and conclusion in less time than a single classroom session. Keep up the great work.
Maybe just try watching the video from beginning to end as intended. A decent creator knows what they're trying to convey and puts a lot of effort into structuring the video for the best result
i think a large(2m ish) circle of acrylic ligth from the edge to show caustics would be really cool. also if the sensitivity allows it watching a double slit has to happen.
Would be cool to have a few different media with different refractive indices, and watch the light slow down inside! Other ideas are: Retroreflectors Making a big photomultplier tube out of a fluorescent light tube (just power with a capacitor without the ballast or starter). Diffraction grating Polarization filters Vapor/dust Fluorescent dye
I've done some work on super fast timing electronics for PMTs and a common technique for making even higher precision timing measurements is using the repeatable shape of the PMT response. Sure it takes a couple of ns to rise, but that behavior is close to identical every time, so the actual uncertainty is just the jitter in the measurement. That jitter can easily be as low as a couple hundred ps. With a good PMT type device and good electronics it can be as low as 50 ps. It does mean more careful fitting of the voltage curve on every run.
Suggestions: - Get optic fiber filament and demo how total internal reflection breaks at a certain angle (laser pointing into fiber) - To get better resolution, make a fine/coarse adjustment system with some high precision low res servos - Put a pane of glass in front of it at an angle - Maybe try to demo some interference? That'd probably require much higher resolution but would be sick
This is so so cool! I would love to see the light travel in water. I've been trying to understand why light slows down in water for a while now and I think that this demo could shed some light on it. The best answer I could come up with is that while original light still travels through at speed c, the light made by the electrons jiggling opposes that light and eats away at the front of the original beam. I'm really not sure about that explanation though. But if it were the case, you'd might have a detectable amount of light that hasn't been eroded away yet, and is traveling at the speed c. Probably not, but it'd be awesome to look at the results and see! Thanks for the awesome video!
As a side note to this and the other video about electricity propagation, the 1nsec/foot rule is pretty accurate for electrical signal transmission in a wire as well. Anyway, it is a close enough first order approximation. For those that have not thought about it, this is a huge reason why small equals fast in computers. Large computers eventually end up limited by their ability to move information from say the top of a 6 foot rack to the bottom. The only way to fix that is to locate processing elements close to each other and transmit distance between complete calculations. And to move this to frequency, a 1 nsec period = 1 GHz. My computer here (which is a bit older) has a clock speed of 3.4 GHz. That means the 18 feet in the video here is 61.2 clock cycles of my computer. I hope that gives everyone a bit of perspective on speed and size.
At first I thought you should fill the room with smoke to make the light rays from the disco ball visible as they propagate in all directions. But of course having the smoke be exactly identical over time might be difficult. Filling a water tank with slightly opaque liquid would probably be better to enable recording a focused light beam in various interesting configurations.
I love the solution of reverse time slicing to stack a high speed camera. I had similar fun with my first PMT, the treasure for hours eliminating any source of photons entering a room even through small gaps and cracks. Have you put the PMT in it's own self contained light proof box and taken it outside at night and pointed it around at the night sky? The 50/60Hz hum of street lamps used to swamp other light sources so adding a 200Hz high pass filter and connecting the output to an audio amplifier makes for some interesting listening. There are some really fun light sources to listen to.
Seen a couple of direct capture light-moving-across-a-room videos but this is by far the cleanest and most impressive I've seen. The explanation of how it was done was great and... illuminating. (Couldn't resist.) Amazingly done!
The best setup for scanning a scene is using an equilateral polygon mirror like a hexagon, the rotating polygon mirror mechanism is also known as a polygon scanner, it gives you a precise continuous scan with no jerking motions, as you don't stop or accelerate the mirror, as it spins, your line of sight pans the scene, and when it reach one edge to the next your perspective resets, like teleporting the mirror to its original starting angle.
9:34 you just need to spin the mirror constantly with a very precise and high rpm and shift the phase compared to the laser pulses slightly on each rotation. This way you don't record a single pixel in time but swipe through the scene at a specific time of start with a specific delay for each location. So you can match each location depending on the start time and delay compared to the start time of the laser. This should be much faster than a stepper motor, and you record a much higher resolution, as your resolution is now just determined by the rotation speed of mirror and the resolution of your Oscilloscope.
A stepper motor is what u might want to use for better resolution of movement compared to the servo. They are specifically designed to have exact control over how much the shaft spins, repeatedly at a high resolution. The motor itself can spin at precise fractions of a degree, and you can always use gear ratios to take it even further but then you have physical tolerances and slop to deal with. But if you managed to make something work at all with a servo, the built in resolution of a stepper motor alone should be plenty to get the accuracy you are after. After typing this out i noticed that u had microstepping as an option written on the screen hahaha. But ya thats what id think to look into first!
Problem is the imager is being scanned, not the laser. You'd have to move the imaging path but it has a huge aperture - you'd need a mirror the size of the camera, which is exactly what he already has! (Kind of funny to think of that setup as the world's worst laser galvo.)
The problem is the laser galvo input aperture is going to be a magnitude or more smaller. Though maybe that removes the need for the secondary aperture?? I think closed loop servos for the existing mirror are a better incremental idea.
Yes this is a great idea. They are typically used for laser light shows. I in fact have already been using one for a similar project (not light TOF, but instead other light properties) and it works great. The aperture is a little smaller, but I don't think it should be too much of an issue if you average frames. I'm curious if you can but a large lens in front of it if you really need more light, but I would recommend trying just the galvo first.
26:01 'Oh, ('ehh') I don't see the laser - I guess the glasses really work...' (I'da flipped them up to find the laser, lol); I don't know of anything cooler than this experiment, butt others' suggestions are great!
Suggestion: I would love to see single/multiple slit interference of light over time! Not just at the detector/wall but also the path it takes visualised with some medium like smoke. Amazing video as always, keep up the amazing work
Holy shit!!! This is incredibe! I was gobsmacked by the exact same video that inspired your project. Up until that moment I only thought of cameras as taking the whole scene at once, and I have been a fan of high speed video since I saw a book by Harold Edgerton back in the 80s. I couldn't understand how they had a sensor that was orders of magnitude better than the state of the art, after watching the video five or six times eventually pieced their description together to arrive at the implementation. Your explanation is amazing and is SOOOO much better and more accessible. I had an existential realization during that original video. I realized that we weren't seeing the laser pulse as it passed through the scene. We were seeing the laser pulse as it passed through the scene and travelled to the sensor. The sensor was blind to what was happening in realtime. This made me realize *I* was blind to what was happening in realtime. That my eyes weren't actually 'looking into space', but that they were only seeing what was hitting my retina...like my eyes were tastebuds for light...and that I was actually in a way blind to the world. Not sure if that makes sense, but that exact same oscilloscope is sitting behind me. I also have a cheap ebay 2 axis galvanometer in the basement (could be a recommendation instead of the hobby servos). Might have to actually try this!!!
I think this raises an interesting philosophical question about the objectivity of the human point of view vs a theoretical "objective" view of a scene
Interesting video. I actually worked on a commercial LIDAR system. So I found it interesting when you showed the IR modules. One does need to be careful with exposing your eyes to too much IR light. We had protective glasses that we used when in the same room when the IR source is on.
11:08 You could have kept a focused laser beam, added mirrors so the light goes from side to side multiple times and finally added some smoke to see the beam travelling.
08:03 - careful, your bottleneck quickly becomes the oscilloscope, showing a sample rate of 1 x 10^9 samples / second = 1 sample every ns, and you are operating in that domain. also reflections and impedance matching (47 Ohm + blue tape) :-)
Awesome again! Love the DoE. For more accurate motion, try alt-azimuth telescope mounts (fork or not). Most are USB serial interfaces and very high angular resolution. Older models are cheap used.
hello, im a new subscriber. i enjoy the videos that you are making. your easy to understand and i have learned alot from you. i am going to binge watch your content now. thanks
Amazing setup, congrats! 1. Shoot lazer from the "camera" position as short pulse as posible, I think you then make depth picture, each next frame will show farther objects 2. Two hole experiment, maybe some more complex variations 3. bunch of mirrors, also laser shoud be focused to parallel beam 4. fog machine and focused laser
@@emovard3n You kinda need the exact opposite: neutrally buoyant particles and completely still air (I don't think this is practical). Have you ever seen a fog machine? You can see when that stuff moves. A fan would just... move it more. Fog consists of heavier-than-air particles. It doesn't want to stay evenly mixed with the air.
@@antonliakhovitch8306 I think turbulent flow from far away or low resolution will show perfect picture. Main issue is overall fog density changes over time as "exposure" of such camera is long
You had mentioned sparks, and I wanted to say how crazy it is that on a rotary engine running at 9000RPM, each coil is firing a spark 150 times per second, and my dwell timing is at about 4-5 milliseconds. (Dwell=time spent charging the coil before it fires off). But my coil is firing a spark every 6.66 milliseconds. Meaning it has about 1.5 milliseconds of rest. This is insanely fast to me. But after watching your video, your tests showing milliseconds and nano/micro seconds REALLY put things into perspective for me, for things I never really had a reference for before. Thank you for that! I now have an actual reference point for how ungodly fast this actually is, in a way that I can actually SEE. Love it!
Milisecond range isn't really that fast, if you think about sounds. 9kRPM engine just produces a buzz, as 150 times a second is still just 150 Hz, the bass end of the audio spectrum. Now, a speaker outputting just 10kHz (which is still only halfway to the top of the audible range, for humans that is) is vibrating 10000 times a second, which takes 0.1ms time for the ENTIRE back and forth movement, so in one direction it is 50 microseconds. And we can admit that the audio range is still almost DC compared to really high-frequency / high-speed things.
Its possible the most effective way to step the mirror would be not to - go super old school and use a pair of rotating mirrors like lissajous lasers and read the position instead of controlling it - no acceleration might mean less shake
That research team that made the bottle video featured in this video as inspiration, they have a video of a single photon bouncing in a mirror box. It’s amazing! Check it out!
This is amazing! Regarding the "motion blur", it might be possible to apply deconvolution in the temporal dimension of every pixel. The PSF could be generated from a reflection free single laser pulse observation to model the ramp up and ramp down of the system response.
If he could somehow make a shutter operate with nanosecond precision (even if it takes many nanoseconds to open and close) then maybe he take full scene photos instead. Maybe Electronic shutters are precise enough, if he could somehow write some code to read specific pixels at a certain timing.
I’ve always been very close to, but not quite impressed by your videos. This project, however, is the most impressive I’ve ever seen on RUclips. Absolutely fantastic work
He pulls out a green card from the box, and they edited it to make it seem like a 'hole'. Then when he puts it on the black board it's shot of a real hole form there onwards.
could you make a "mirror maze" from the side where you can just see the path that the light takes when bouncing back and forth to many different mirrors? that would be fun. you could use gearing to reduce the problems with the servos, if they are 360 degree servos. better yet, stepper motors. more importantly: your setup needs to be wobble-proof. make it heavy, no limp connections, etc. that's what's limiting your resolution as well IMO.
Dude absolutely amazing , id love to see a larger coil of some fiber optic wire in front of the laser , so that when the light hits it loops around and around , by the time the laser hits the wall, it would still be like halfway in the cable.
you should use a 0.9° stepper with a cycloidal reduction to increase resolution and reduce backlash as much as possible. That could give you smoother movement and as good of a resolution as precision you can put into that cycloidal reduction, which id recommend using bushings and thin oil since you want precision and dont have a huge dynamic or static load.
Or just use a much smaller mirror. That was a lot of mass being swung around. Waiting a little longer after stepping for the mirror to stabilize just means the image capture is a bit slower, so not that big a deal. Just convenience.
A slow, steady movement would probably be more precise than stepping. At nanosecond frame rates, even a fast moving mirror will appear to be stationary.
9:36 I have spent my fair share of time on this problem and what I would suggest is stepper motors with a belt reduction. The good thing about steppers is, that they have a very predictible motion and are very cheap. I would not run them without a reduction, since they tend to vibrate a lot if they are driven at a low speed. Microstepping, even with the best drivers (TMC last time I checked) isn't enough if you want truly smooth motion. The good thing about belts, is, that they barely have backlash, which makes them also very good for precise motion. If you want extra accuracy, I would suggest mounting encoders to the axles. Alternatively, you could use an ODrive with a brushless motor, but that is way more expensive and probably wouldn't be that much of a benefit for such a low load application.
I just finished building something with a decent tmc driver, and I had debug issues because the test motor was moving so slowly, so quietly and completely without vibration. Not particularly fast, but I think they'd be a huge improvement on hobby servos. I'm a bit amazed they worked so well at all! For this level of precision
I also suggested an ODrive in another comment. It would probably end up being ~$400 for the whole setup, but that would be a very useful item to have for future projects that need any kind of highly precise motion that won’t be affected by the vibration inherent to steppers. Another thing that comes to mind is that it might be possible to use a VESC, which I believe can be had for less money than an ODrive (but would also probably be more difficult to get setup). I am also wondering if the mirror could just be spun instead of moved on an arc. It’s pretty easy to set up brushless motor controllers to spin a motor at a specific speed, so if that will work for this application, it could reduce the costs significantly.
For such low power application, I would suggest the SimpleFOC library for arduino. Lots of example and good documentation. It allows to use gimbal motors with FOC, but also stepper motors. Really easy to use and way cheaper!
@ oh cool! Somehow I hadn't heard of that before. Yeah, that probably would be a very good way to go. One other thing that came to mind, a while ago I came across a TMC chip (I think it's the TMC4671) that integrates all the functionality in hardware to drive brushless motors with FOC, which also could be a good way to go. I have the dev board, but haven't gotten around to doing anything with it yet.
Gimbal motors with high phase resistance can be easily driven with a SimpleFOCMini (around 4-5 USD) and an Arduino (can be less than 10 USD depending on which model) running SimpleFOC. If you need more capability than most Arduinos, you could opt for an ESP32-based device, such as those based on the ESP32-S3 (3-6 USD).
I just have a doubt, you said when the laser turns on it sends a signal to the oscilloscope and it starts recording, but the current (electrons) travel with a speed closer to the light (still
the signal to the laser is also delayed, it doesn't matter what the delay to one or the other is, as long as they're roughly (within a few nanoseconds) synced
@@fuby6065yes, but the laser turns on first and then it sends signal to the oscilloscope, there is definitely a delay gap, maybe it's only a few nano seconds and negligible.
Digital scopes can capture some number of samples before the trigger point as well as after, so it's just a matter of extracting the appropriate part of the data.
Light travels 29.98cm in a nanosecond. It's already closer to a metric value. Like, just redefine a foot to 30cm, it's only 5mm shorter and you'll all become "taller"
The meter is as long as it because the polar circumference of the earth was defined to be 40,000 km. Some error propagation later, they made a platinum-iridium alloy rod the length to match and made it the definition of meter. And then they had to match that with every successive definition including the constant definition. The speed of light is now the definition of a meter by way of a precise clock.
30cm: ~0.2 mm error 1 ft: ~5mm error just to make it more obvious WolframAlpha queries: "1 ft - light-nanosecond in mm" and "30cm - light-nanosecond in mm"
This is a FUN place to hang out!!! Thanks AlphaPhoenix!! I'd like to seethe double slit interference pattern from above at a billion frames per second!
This is honestly the first video I’ve seen from you. Based on what I’m seeing, I am definitely here to stay. BUT I have some fun RUclips algorithm happenstances that seem too obvious not to point out. First off, this video did not pop up on my recommended page, it popped up as a recommended video to watch AFTER watching The Thought Emporium’s video on the XJ-9, an antenna scanner that reads wifi waves. The interesting thing about it is what both videos have in common: the clip from the Iron Man movie “TONY STARK WAS ABLE TO MAKE THIS IN A CAVE WITH A BOX OF SCRAPS.” And also, besides the fact that I obviously watch this nerdy side of RUclips on a regular basis and am far from being ashamed about it, I also have been watching a lot of Shorts involving clips from the MCU. Just something I noticed. Maybe someone should look into it and see if it’s worth capitalizing on.
Corrections and FAQ in this comment!
0) Stuff I want to shoot with this camera (adding suggestions to list):
Race different transmission lines vs fiber optic
race light in clear tube of air vs clear tube of water
refraction, bending and slowdown of wavefront
patterns with lots of mirrors somehow?; double-slit interference
interference with one beam delayed (really excited about this one); infinity mirror with an object in the middle that would get cloned slowly looking off into mirror-world
short pulse to get "star wars blaster" effect (this might require buying a different driver!)
shine the laser on a fluorescent material
large lens (I have an old projection TV fresnel in my parents' attic I could pull out for this!)
Electric arcs - this would have to be a macro shot, so observing actual light would be impossible at this framerate, but if I could get an arc to form extremely consistently, i bet it doesn't actually FORM at lightspeed - you have to chug some ions around through the air gap. this could be really awesome!
Light solving a maze or shining into an penrose room. flash bulb in fog - like that one nebula light echo timelapse
long stretch of fiber optic
do a time-delay process with the camera and light in the same place to get a TOF depthmap
two more different color lasers to do full-color imaging and/or refraction
One of the standard computer rendering things like cubes in a mirrored room, or that one teapot or something to do “ray tracing” irl
1) This isn't a "stop motion" technique - it's more like making a 1Bfps camera that's 1x1 pixel, then tiling lots of videos next to each other. I did consider using a Pockels Cell in front of a regular camera to try to do stroboscopic stop-motion-style data capture, but I think it would have been more difficult/expensive than what I ended up with.
2) I've gotten lots of great control recommendations - a few really high-precision servo brands, and things with spinning mirrors/lead screws/stuff like that. My first concept for the camera was actually to mount my one pixel behind a static lens on a 3d printer gantry and raster it back and forth to create a "virtual" sensor. I didn't do this because In the garage (a relatively tiny space) i needed a really wide field of view and from the cyanocamera project, I knew maintaining focus at the edges would be impossible. The spinning mirror is really appealing, but "most" of the time is spent with the mirror looking the wrong way, so I'd be multiplying my scan time by 3-4x. it WOULD give a more consistent image though! I'm also about to look up a galvanometer.
3) most of the scans i scaled to take an hour or two. the more averages you take on each pixel, the better you can resolve the "darks" and get illumination in the room that isn't the primary beam. If i only wanted the primary beam and a noisy signal everywhere else, I could easily get away with one laser flash per pixel! (this gets better if I overcrank the diode, but we all saw how that ended)
4) There are a lot of comments asking about the "early" light from the laser, the fact that the beam is traveling at an angle to the camera, etc. if you look, the first half of the beam moves really fast and then it slows down as it starts moving "away" from the camera. this is a sort of doppler effect, and it's an unfortunate limitation of the size of my working area. it's a really cool effect though, and one I want to discuss more in a future video!
5) I DID use fog for the disco ball! in fact my fogger caught on fire so I got bonus fog... That's the only reason the "streaks" were visible.
6) The part about adding an aperture being similar to a "real" camera lens' aperture may be a bit muddled. in a regular camera, the resolution is set by how far apart the light sensitive pixels are on the chip, but in this case, my light sensitive area is probably more than 1 cm^2, so the aperture restricts defocused light but ALSO light that comes in off-angle.
7) How many subscribers do I need to drill a hole in my garage door? let's say 10M - that may or may not happen one day in the future xD
8) a few commenters have very accurately pointed out that the "start recording now" signal should arrive at the oscilloscope AFTER the laser light does - this is correct! thankfully the oscilloscope can back up slightly from its trigger point, and that's how I'm doing it. it really doesn't matter when i see that signal, as long as it's the same for every single flash I can use it to synchronize the data.
9) I don't think there's any simple way to make this camera faster. the scope is 1GHz sample rate, but actually 100MHz for signals, which means ~3ns rise time is all I could hope for. The laser driver ALSO is in the range of units of nanoseconds, and the PMT is also in units of nanoseconds. in order to speed this camera up, I would need a new light, a new detector, and a new data collector - basically a new everything. It'd be better to switch to a legit streak camera at that point. maybe someday.
10) a lot of people are suggesting that to see light move, the camera would need to go “faster than light” but that’s not the case - you can take a video of a supersonic jet with your phone without any part of your phone breaking the sound barrier
11) this cannot be done reasonably with mechanical shutters. If you have the laser going through a 1mm pinhole and you want to slide a shutter out of the way to turn on the light in a few nanoseconds, that shutter itself would need to be going 1mm/5ns = 200km/second = 124 mi/sec
A pulse of light shorter than the length of the room it is in?
edit: Someone else has also mentioned a mirror maze. That would certainly let the path get much, much longer without taking too much space.
Race light in a vacuum vs in normal pressure vs higher pressure, try filming lenses work. mirrors bouncing light between each other.
Laser refraction when it hits a tralie and/or the reflection of lasers among several reflective cilinders (or mirrors placed almost parallel facing eachother) to see if you can see its path occuring like a slomo lightning strike.
Such a tralie could be moved reproducible using some piezo element and change the light angle in a reproducible way.
Your gimbal solution is to use an electric telescope tripod.. you can find some really nice ones with "high resolution" motion control quite affordably. And they plug into computers quite easily.
A focused white Lightbeam that passes a optical prism and gets divertet in it‘s spectrum.
And maybe a bldc with a Magnet at the end and a HAL-Rotation-Sensor might nie better as a Servo? The Resolution should be much higher.
Suggestion: I always wanted to see a laser bouncing between two mirrors back and forth zigzagging at a slight angle
seconded
thirded.
Like an infinite mirror, maybe you'd see the rooms appearing one at a time.
Please this
Was gonna say the same thing
Yo that was a beautifully complex project!! Well done!
Build one, Mehdi!! 😂🙏♥
8 million sub special when? :D
For sure!
@@the_boogey_man0 He loves you too but this is the third time he's seen you speak for him out of context and he thinks you're lucky his patience is infinite.
fuuulll bridge rectifier!
I've come to the conclusion that blue painters tape is an essential piece of scientific equipment.
you built a what
Lol
@@보보코코와 "IN A CAVE! WITH A BOX OF SCRAPS!" nice Obadiah Stain meme 😅
Sounded so... penny from bbt
he built a 1,000,000,000 fps video camera
"Did I stutter?"
Absolutely breathtaking! I wonder if you had a coil of bare fiber optic that was just glass with no insulation, if you could see the light looping through the coil
that would be insane
that'd be sick
No doubt Telecoms fibre would do near total internal reflection but there are light pipes made for other applications that would leak satisfactorily
@@christianblack2916 Or perhaps roughing up the surface a bit could work. The total internal reflection only works when it is spotless on the outside, as the reflection happens on the interface.
I think this is totally doable with a little elbow grease.
There’s a video I saw somewhere in Reddit with oodles of upvotes that was called “the speed of light” or something that showed two guys in lab coats and a very long led strip light Illuminati g from one end to the other and the fact that people were eating that up made me really sad
Thanks!
23:16 "Why is onenote using 22% of my CPU?" Most real thing ever said in any video ever
Really gotta get rid of all that Microsoft stuff, and more generally, avoid any subscription-based, bloated, always-online software. Free, lightweight, offline stuff is much more reliable and doesn't randomly peg the CPU or fill the memory.
@@ToyKeeperoffline stuff completly use cache memory to read and use the data you want.
For the web app 100% with you for the speed, lightweight, so much time saving.
@@ToyKeeper Instructions unclear. Why is Emacs using 22% of my CPU?
tell me why ublock is not working
it's uploading to cloud and spying on you ... this is why i dont use any default software or setups, it always spies on you. that's why i've always used 3rd party firewall control that blocks ALL connection attempts and gives me alert first, there's at least like 100 things on windows that connect to internet without you knowing about it.
slo mo guys have been real quiet since this one
I think Gav is working on filming light over a much longer distance! (and in a single shot, not strobe-ish like I did). He mentioned it in a recent video with Adam Savage and I really hope he gets it working cause I really want to see it!
@@AlphaPhoenixChannelsmells like an awesome collab opportunity to me
@@AlphaPhoenixChannel wow, that's awesome, I'm excited!!
The video was up for only minutes when you commented. So how fast are they supposed to acknowledge it.
yeah, they only filmed it 10 thousand times better 5 years ago
This guy: builds a 1000000000 fps camera from scrap
*Bank security cameras:*
I know this is a joke, but I actually did some research on why they are so fuckin awful and it’s because they constantly record which requires a shit load of storage space so making it less detailed allows for more to be recorded before they gotta dump footage
@@TheAustinTalbert That was true in the VHS days or when a normal PC had like a 32 GB hard drive. Now you could fit months worth of 4k footage on consumer grade drives and racks. They're just lazy.
Well then you should make the bad guy walk into the bank a couple of thousand times
@@TheAustinTalbert "shit load of storage space" A proper full HD camera generates a video of a 5 Mbit or so. But let's step it up to 4K, maybe 10 Mbit. That's 100 GB per day. Let's say they have 5 cameras, adding up to half a TB per day. An 8 TB disk (less than $200) can store 2 weeks of video. There's really no excuse anymore for having a low resolution security camera.
The garage door really got me !!!
me after the spring fails
That was funny. I thought, there's no way he's drilling through his door unless he's getting a new one soon for some reason.
Me too but that edit with the mirror on a black screen was also quite good.
The hole in the garage door and the mirror on the black screen becoming the "hole" reminded me of the simple gags in This Old Tony videos. Especially the way he doesn't acknowledge the joke but just keeps going on with his presentation.
@@eklhaft4531 Yeah, it actually appears that the "mirror" was just a green screen or maybe there is nothing there at all and it is just an inverted mask, but at any rate the "mirror" appears to be the scene behind rotoscoped onto it, since if you look at the "reflections" in it, they are not of his garage as a mirror would be. Very cool idea.
The hole in the board at around 4:20 is breaking my brain. Also 19:15 blue tape is ALWAYS a great idea. This is so sick, great work!!
You should teach him how to integrate a condom. ;-)
I was like "wait, what...?" and then I noticed the box as well as something else 😂
It got me good, still puzzling it out
@@dirtdart81 I think it's a composite, where half the scene is an underlaid static image and then the "hole" is simply cutting a hole in said image... I unno...
It makes sense for awhile, but I'm pretty sure that entire car explanation scene is still using that camera trickery, and my idea falls apart....
🤔 Unless it's ONLY that square that's that static object... 🤷♂️
I am fairly sure that hole was already there during the entire shot, and just covered up with black until he places it. The hole he grabs from the box and places there is either nonexistent or does exist but is, either way, tracked with some footage in the shape of a circle. If it does physically exist, he might have had to paint it out when it covers up the real hole in the board until its removal from frame (possibly it's dropped when "placed"?).
Suggestion: If possible, I would love to see a when natural sunlight passing through a prism. It would be fascinating to observe the dispersion of the light spectrum at such a high frame rate!
I second this except just use the UV laser on a strongly fluorescing material, then put that light through a prism
@@ianglenn2821 He could also measure the light twice, once with an optical filter in front of it to see the blue light and once with an optical filter in front of it to see the yellow light. That way he has two colors.
@@ianglenn2821That's a clever idea to get different wavelengths of light! But do you think the light emitted by the fluorescent material will ramp up fast enough?
Unfortunately, that's not really feasible. To use natural sunlight, you would need a mechanical shutter (since you can't just turn the sun on and off.) The mechanical shutter would need to move impossibly fast to get a small enough pulse of light with a fast enough rise time for him to record.
@@hans-jurgenvogel6789 the delay between absorption and emission that happens in fluorescence is in the range of nanoseconds, so it should be a noticeable delay, but still in the timeframe of that video.
Since the camera is monochrome, he should probably try out my using filters to create two videos for two different colors, like I proposed in my other comment.
First StuffMadeHere doing "fake custom lenses", now this! Love seeing all this optical physics content!
Yeah there were a lot of similarities in the setup between the 2 videos, super interesting coincidence
I set Python to rendering that clip with all the black circles and went to watch Shane’s video while I waited. Two minutes later I saw the same graphic. I laughed so hard 😂
I guess there are only so many ways to describe a camera. I really like his ant analogy
IMHO the nanosecond camera is cooler, although his wife probably wouldn't want to stand still for an hour while being blasted a thousand times by rapid laser flashes. Mount it on a rotating arm next?
I like how alpha phoenix made it look much easier to build a bn fps video camera than a photo camera on a spinning stick
Both these guys are insane. I'm constantly mindblowed.
Man, Alpha Phoenix videos are so goddamn cool.
Sometimes I can’t believe we have someone just hanging out, on RUclips, literally building cameras fast enough to capture the movement of light, or sensors to track the speed of atomic movement in a piece of rebar, systems for displaying electric wave propagation models, and horrifyingly powerful gerrymandering algorithms, and doing it just hanging out in his cave, with a bunch of scraps.
To have someone this smart, exuberant, and good at presenting such wildly disparate topics in such understandable ways (without dumbing it completely down), with some of the coolest homemade contraptions I’ve ever seen is awesome.
So thanks for putting out the best videos on this site!
You are SO right! The guy is amazing.
Yeah. This guy is the [vaguely science-related subject of choice] teacher you wanted as a kid.
Completely agree to everything.
For me, the voltage through the cable and this one are absolutely top level videos.
"... good at presenting such wildly disparate topics in such understandable ways (without dumbing it completely down..." this is exactly my thoughts. The mentor presentation style is 11/10
Yeah, this channel, @stuffmadehere and @breakingtaps are the best at this at the moment. They have such cool ideas that take months to figure out, build and film.
Better project than 95% of our Master students finish with. And without guidance.. Awesome! And results are just amazing!
As an alternative to PMT you could buy a cheaper silicon PMT, or Si-PMT for short. They are not as good as but way cheaper(~50$), far less dangerous voltages and smaller change to kill coupled electronics.
Also funny how you darkened the room. Reminds me how we did the same for some long time measurement with actually photon counting... Every LED, every possible light from outside. All were sticked with a black tape. It is amasing how sensitive eyes are after being an hour in complete darkness :)
Small hint for the LIDAR sensors: Most cheap sensors don't actually measure the time the light takes directly and instead the shift in phase of intensity modulated light (which takes longer to measure but is way easier)
🤔
@@AlphaPhoenixChannel also - possibly dont control the sensor movement, esp not stepped - give it constant movement (with rotating surface coated mirrors) and measure it like LIDAR sensors/oldschool lissajous lasers - could even overlap the smeared pixels for better resolution via interpolation if it moves fast enough
And LIDAR doesn't measure the intensity of the reflected light. Just the phase shift. It can do this by measuring how much constructive/destructive interference there is with the reference signal. Using the phase shifts for multiple different modulation frequencies allows it to estimate a distance
@@bornach trying to understand this :D
so the phase shift occurs due to light's travel time? for different frequencies the phase shift differs because each frequency is at a different point in it's respective phase when it gets reflected? wouldn't that give you a resolution on the scale of the wavelength, ie. nm? how do you derive distances on the scale of meters from that?
@@DackelDelay You can send a coded signal (AM signal) and use those two points to get distance very precisely.
Digging the hole out of the box was so good. Didn't give the joke away at all, just played it off like it was a real thing, and the effect was really well executed. Loved it. Cool little nudge to Corridor later on too.
In a project about cameras it felt appropriate 😁
i thought i was trippin😂😂😂
For real both jokes got me 😂 I love that the second wasn't referenced at all!
Frankly one of the best RUclips videos I've seen in a long time! The enthusiasm, the explanations, the outcome, freaking amazing. I just feel like going to my office and start tinkering with stuff right now!!!
4:29 oi, don’t make me watch this 25 times. How dare you make such a smooth edit dammit
It's such a great bit lol
So casual haha
Yeah. That broke my brain for a second or two
he didn't move the camera so he could cut out the mirror with the underlay of the background without the black poster board.
did you even see what box he pulled that out of ?!
I just want to thank you for your videos.
I have a PhD in physics and worked on a combination of electrical measurements and ultrafast lasers (
Smooth brain question here:
So does this video finally prove that I should start cleaning out the photons from my android phone camera to keep it from getting clogged up over time or that I we all need to start finding a way to manufacture the ultimate surf board that can lets us start catching light waves?
All jokes aside, does this video prove one or the other in any way??
If you actually understood light youd understand that the camera is limited to the reflection and process itself physically, the light moved far before the camera actually picked it up, your basically looking at a 1billion fps reflection, the light moved before that
The sensor is slower then the light moving itself
@@romankozlovskiy7899 If you are replying to me, then I will start by asking you to name me 10 people who actually "understands" light, and then I want you to compare that to all the people that dont.
Then consider that your perception of who "understands" light is very "light".
Also, great of you to follow up with extra info about the behavior between the camera lens and "light".
Couple of ideas:
1. Any chance you could switch to a white light laser and repeat the experiment with R/G/B filters in front of the detector? Having this shot in color would be amazing!
2. How about a light maze with multiple mirrors just allowing enough spill on the side of a background plate (that the mirrors are fixed to) to have the progress of the light travelling visible?
3. Could do a split beam perhaps with one half travelling through another medium (glass? water?) and have it race against the light of the other (half) beam.
Amazing video, mad props!
This is incredible. The fact that we can see light move isn't something I thought would ever be possible. And you're doing it with scraps in a cave!
I'm not entirely sure I trust the results but interesting video
@@TranceSFXwhy not? It's been done before
@@TranceSFX it's literally on camera what more proof do you want haha
@@jdr91 Yeah, what do you not trust? You believe it happened, right? And the maths are correct enough? (that's me pluralizing "maths" to make myself sound European and, somehow, smarter)
That is because you can't see light move, lmao.
The video guy doesn't know what he is doing.
As a lighting control designer, I found watching the second and third order reflections fascinating
He could do an experiment with the Cornell Box and compare the result with a pathtracing simulation where every path length is bounded.
Yep, that's equally interesting to me as the initial beam travelling across the room
Just a humble advice: considering the impulse response of your detector is deterministic (which it seems roughly to be), you can apply a temporal deconvolution to recreate a much faster response time sensor. A basically instantaneous light sensor detector with a ramp-up time of ~0ns. In the end your video will be just your laser ramp-up.
If your laser ramp up signal is deterministic too, you can deconvoluate your laser too ! A double Wiener deconvolution can in theory bring you back to a
Can you show a double-slit experiment with that camera?
will that work or change the result
@@Meilk27 You mean like how observation can affect quantum experiments?
@@fluffsquirrel I have only a layman's understanding of the experiment, but yes that's what I was referring to (I think)
@@Meilk27 Thank you! That would be very interesting to see the effect
@@the_boogey_man0 He loves you too but he would like to point out that many of the other posters have already seen you speak for him in other replies.
I'm always impressed by your demo setups. They intuitively show advanced concepts in a way that is achievable by a normal person.
What I would like to see:
- White light getting refracted through a prism
- Light passing through a fresnel lens (or any lens really)
That video was mind-blowing! Whouah!!! ...!!! Great practical science! ❤😀😀
Suggestion: round trips between almost parallel mirrors. With a bit of fog to "see" a fraction of the the light. The laser would enter the cavity between the mirrors with an angle.
One issue will be the to ensure a constant density of fog during the duration of the scan. Maybe without any fog, and just the humidity of the room will suffice. 🤞
That might be the best pitch for Brilliant I've heard. Take note, Brilliant, pay this man more.
Totaly agree whit you.
It might not be feasible with the current setup, but I have a romanticized idea about what we would see with a double-slit experiment in slow motion.
I was thinking the exact same thing!
Oh yes!
Yes please!
The double slit experiment requires single photons interacting with themselves.
@@kid_missive If you want to prove it's a quantum effect, yes. But you still get the same interference pattern regardless of if you use single photons or a lot of them.
You should definitely try to focus the beam and make it as thin as possible without loosing detail on camera, then bounces it between multiple small mirrors around the room to see it zig zag between the mirrors! Would be super cool seeing the lights travel from a to b like that, never seen this type of footage before!
The reflection wave off the garage door further illuminating the entire room was great!!
Great work. Can you get video of electricity arcing from a negative electrode to a positive electrode? That is something I would love to see (or actual lightening if it can be done safely).
You should place several different solid color squares against the wall you are recording to see the differences. Under water, the deeper a diver goes, the more colors all end up looking grey (red is the first color to go grey).
I think that this is unfortunately not possible with this technique. You need exatly the same illumination thousands of times to composite this footage.
Oh you genius bastard. It’s a rotating lens connected to an oscilloscope into the inverse of a Matrix bullet time. Have you invented a whole new camera? Give this man an award.
mostly turned LIDAR up to 11
@@mycosys It’s weird but true, how the best inventions reuse something else but for way better purposes. The steam engine was totally that.
If you watch the video he tells you why it's not an original idea. The original inventors did win awards many years ago
15:18 If you can estimate the motion blur, you could try to deconvolve it to get approximate ground truth images without blur.
Dammmm this is TRUE! We already have this technology and could bolt that on.
That means you could get higher resolution and higher framerate?
Alas, deconvolution is ambiguous. The motion blur has decreased the SNR in the high-temporal-frequency content of the video, and there's no way to get that back other than guessing.
@@TheLoneWolflingexcept that he knows (or can estimate with reasonable accuracy) the parameters of the original convolution
@@krisztianfekete3277 Alas, convolution->deconvolution is only a round-trip with infinite precision. With finite precision it irreversibly loses information, as I stated.
(Note: there is a much simpler argument that _almost_ works here - consider a 2-bit color depth with a 3 frame box blur, and the input signals [..., 3, 0, 0, 3, 0, 0, ...] versus [..., 1, 1, 1, 1, 1, 1, ...], both of which result in the exact same output signal. However, this only fails in the infinite-sequence case. So I'm going for a different more convoluted (hah) counterexample instead.)
Consider a very simple example: a single pixel; 2 bit color depth (0, 1, 2, 3) with a five frame box blur. That is, a blur function of [1/5, 1/5, 1/5, 1/5, 1/5] in the temporal dimension.
So e.g. [0, 5, 5, 5, 5, 5, 0, 0, 0, 0, 0] becomes [0, 1, 2, 3, 4, 5, 4, 3, 2, 1, 0]. So far so good.
Now consider the input signals:
[0, 0, 0, 0, 0, 0, 0]
[0, 1, 0, 0, 0, 0, 0]
[0, 2, 0, 0, 0, 0, 0]
What happens?
Well, in infinite precision the answer is easy:
[0, 0, 0, 0, 0, 0, 0]
[0, 1/5, 1/5, 1/5, 1/5, 1/5, 0]
[0, 2/5, 2/5, 2/5, 2/5, 2/5, 0]
In 2-bit color depth?
If we truncate, then these three both map to exactly zero, and we've lost information.
If we round, then these three still both map to exactly zero, and we've still lost information.
If we round stochastically (so e.g. 6/5 means "1 with a probability of 20% and 2 with a probability of 80%")?
Well, I just did that ([random.random() < PROB for x in range(5)]) and got
[0, 1, 0, 1, 0, 0]
So what's PROB? Answer: it could be any value 0 < PROB < 1. So from 1/5 or 2/5: both work with finite probability. And so we've lost information.
Thank you so much! Really appreciate seeing someone just genuinely interested in scientific experiments with an engineers heart and so much curiosity. And you show and explain all this with so much passion and focus on what you want to communicate without this - I don't know how to explain it.. this feeling that I get from some other RUclipsrs that somehow manage to make it feel it's more about showing off and making it about themselves rather than science. Like without any "narcissistic vibe" if you want.
Best wishes from Germany, looking forward to seeing more like this! :)
You fully got me. I was like OK you just actually drilled a whole in your garage door.
"... And it would have worked if you two hadn't stopped me!" -- Egon, Ghostbusters
Yeah I was like that's a whole ass hole in that door.
Because drilling a hole in a garage door is the most rediculous part of this video, obviously
If someone else used that title, I would think that it's bullshit.
But now I'm intrigued.
At least he didn't claim it can see around opaque walls - Acme portable holes, notwithstanding
@@entertainme121 Well, I posted that comment before watching the video.
So, thanks for the summary, I guess ?
If you had insane rolling shutter (in that case on both axis) while filming a way too fast periodic thing (in that case a multiple frequency of the framerate).
Would it make it not "slow motion" ?
When did it stop being "slow motion" ?
(I get your argument, I just want to be a pedant.)
It was bullshit, just it's deceiving perception of reality not actually capturing the speed of light.
@@chazlabreck Seeing light propagating through space, even if it's a bit indirectly, is still really impressive.
this is so great! i remember seeing that coke bottle light and researching how they actually did it.
i love that you have basically redone that with cheap components
Put a fish tank in the shot, or like a long tube of water, so you can look at the propagation rate in two different mediums at once.
This was my thought as well. Being able to visibly see different speeds of light through different mediums would be neat. I wonder if you could work in more, like air, water, and glass. Or are there other fluids which slow down light even more than water/glass?
You said that on porpoise!
Wow. You might even be able to capture the rotation of polarised light through a stereoactive solution
I would love to see this as well!
I feel like a double slit experiment or some other interference demonstration is a *requirement* for this.
filming the double slit will probably break the universe
He'd have to improve his time resolution to the sub-femtosecond range. His setup can resolve propagation differences on the order of a meter, but optical interference happens due to path differences on the order of 100 nanometers. You need attosecond streaking or some fancy optical cross-correlation methods to visualize the temporal phase of a light wave.
You could probably film the light passing through a diffraction grating wit the setup, but I don't really know if you would see anything interesting (but it would still be cool!)
Why not point the camera at the pattern the light makes on the target as as the interference pattern collapses?
@YSPACElabs it's monochromatic, so no. Even if it was white light, the speed of the different colors is virtually identical in air. It's a different story in water, though.
Getting the hole out of the box and putting on that wall are really EPIC!! Nice video!! Keep going, we like your channel!
The video title seems something from the future
The future is now!
Because it was! And now its not!
"The future is already here, it's just not evenly distributed."
@@The18107j Not to be rude or something, but WHAT THE FUCK does that mean?
Such a cool demo! 🤩 So happy you got it working, that footage is just mesmerizing. Seeing the rays bounce off the disco ball was amazing. I would watch literally anything and everything filmed with this setup, please keep them coming!
Oooh! Okay imagine this. You have a clear tank with a hole in the side and a laminar stream of water going down into a cup. You shine your light through the tank and the light follows the stream of laminar flow water into the cup, so you get all the refractions of light from the container and the cup, and you can see the light working its way down the stream of water. Now, you'd need to have some sort of a pump configuration to take water from the cup and constantly replenish it in the tank in order to keep the levels stable so that your slow scan imaging isn't thrown off by changes in flow rate. But if you could pull that off, that would be the most epic video ever. And I know SmarterEveryDay would love it. Destin might even be willing to make a collaboration video with you to hammer out all of the engineering problems dealing with the tank, and improving the precision of your scanning equipment.
This is possibly the best science channel on youtube. Putting together real experiments to show unintuitive things in a practical way. The stuff like showing how electricity literally moves through a wire, when all the creators were bickering over the theory of light waves. That was an awesome video, and this is another amazing banger. Ever since I was a child, I'd imagined what it would take to literally see light move across the room, and I'd tried to mentally understand what happened when I flipped a light switch.
Seeing the experiments gives me such an amazing visual that I wish I'd been able to see as a kid. It's an answer to all the questions I had, and more. Seeing the disco ball lights bounce around was amazing. Even though I tutored college physics, and I understand all the theory and formulas, it's an entirely different thing to see it with my own eyes.
What's even more impressive, is that you're able to build this setup with a reasonable budget and tons of knowledge. Definitely the best science channel on youtube. The only one that's doing something different and something genuinely impressive. You're probably the closest we'll ever get to a modern Richard Feynman, with your ability to so clearly explain complex topics. You're an inspiration.
Well said.
so true
Feet are REALLY convenient for working in audio too, 1 foot = ~1ms - i convert to feet to make acoustic delay 'visible' XD
as a nice round number: 30cm is just 207.5 micrometers shorter than a light-nanosecond. In comparison, 1 ft is 5008 micrometers longer than a light-nanosecond.
@@laundmo okay but half a meter would be 50cm 😅
@@Mezgrman So you're telling me I'm not 3m tall???
@@Mezgrman oop, brain fart. fixed it, thanks
I need to make a video about universal units. It bothers me that the speed of light is within rounding error of a simple number.
25:13 C'est incroyable de voir la lumière ainsi se propager. Très intéressant, merci.
You put the anti clickbait video at the beginning and then give us the full show at the end. Oh, you tease us! And we like it. I do know that every time I skip to the money shot at the end of AP videos, I end up jumping back just a bit for context of the things. And then a bit more. And then jump back to the beginning to make sure I understand the interaction between different parts. And then look for the test setup. And now I know it's actually faster to wait until the end of the video than to skio to the end and watch backwards and then forwards again.
But seriously, we love how technically dense and well ordered your scripting is while walking is through concepts we have a basic understanding of or potentially none at all. And getting us to an experiment and conclusion in less time than a single classroom session. Keep up the great work.
Maybe just try watching the video from beginning to end as intended. A decent creator knows what they're trying to convey and puts a lot of effort into structuring the video for the best result
i think a large(2m ish) circle of acrylic ligth from the edge to show caustics would be really cool. also if the sensitivity allows it watching a double slit has to happen.
Would be cool to have a few different media with different refractive indices, and watch the light slow down inside!
Other ideas are:
Retroreflectors
Making a big photomultplier tube out of a fluorescent light tube (just power with a capacitor without the ballast or starter).
Diffraction grating
Polarization filters
Vapor/dust
Fluorescent dye
I've done some work on super fast timing electronics for PMTs and a common technique for making even higher precision timing measurements is using the repeatable shape of the PMT response. Sure it takes a couple of ns to rise, but that behavior is close to identical every time, so the actual uncertainty is just the jitter in the measurement. That jitter can easily be as low as a couple hundred ps. With a good PMT type device and good electronics it can be as low as 50 ps. It does mean more careful fitting of the voltage curve on every run.
Not fast enough.
A string of christmas lights to see if they light up one at a time
Remember the part where he has shown that the LEDs in his garage took way too long to turn on?
@@markusa3803 A string of Christmas lasers?
@@DevinBaillie 😭 that would probably work just fine even
@@DevinBaillie Every laser would require it's own controller board then, that would probably be pretty expensive.
Suggestions:
- Get optic fiber filament and demo how total internal reflection breaks at a certain angle (laser pointing into fiber)
- To get better resolution, make a fine/coarse adjustment system with some high precision low res servos
- Put a pane of glass in front of it at an angle
- Maybe try to demo some interference? That'd probably require much higher resolution but would be sick
This is so so cool! I would love to see the light travel in water. I've been trying to understand why light slows down in water for a while now and I think that this demo could shed some light on it. The best answer I could come up with is that while original light still travels through at speed c, the light made by the electrons jiggling opposes that light and eats away at the front of the original beam. I'm really not sure about that explanation though. But if it were the case, you'd might have a detectable amount of light that hasn't been eroded away yet, and is traveling at the speed c. Probably not, but it'd be awesome to look at the results and see! Thanks for the awesome video!
3blue1brown has best video on that topic I think. (I.e. what actually happens to the waves of light, rather than just analogies)
As a side note to this and the other video about electricity propagation, the 1nsec/foot rule is pretty accurate for electrical signal transmission in a wire as well. Anyway, it is a close enough first order approximation. For those that have not thought about it, this is a huge reason why small equals fast in computers. Large computers eventually end up limited by their ability to move information from say the top of a 6 foot rack to the bottom. The only way to fix that is to locate processing elements close to each other and transmit distance between complete calculations. And to move this to frequency, a 1 nsec period = 1 GHz. My computer here (which is a bit older) has a clock speed of 3.4 GHz. That means the 18 feet in the video here is 61.2 clock cycles of my computer. I hope that gives everyone a bit of perspective on speed and size.
Amazing, I admire your energy and enthusiasm. Beautifully described.
At first I thought you should fill the room with smoke to make the light rays from the disco ball visible as they propagate in all directions. But of course having the smoke be exactly identical over time might be difficult. Filling a water tank with slightly opaque liquid would probably be better to enable recording a focused light beam in various interesting configurations.
I did use smoke! And yeah by the top of frame it was unfortunately thinner…
A few drops of milk in a tank of water should work
Nanoseconds? Welcome to the 60s. Call back when you have attoseconds to see the lightwave waving.
I love the solution of reverse time slicing to stack a high speed camera.
I had similar fun with my first PMT, the treasure for hours eliminating any source of photons entering a room even through small gaps and cracks.
Have you put the PMT in it's own self contained light proof box and taken it outside at night and pointed it around at the night sky? The 50/60Hz hum of street lamps used to swamp other light sources so adding a 200Hz high pass filter and connecting the output to an audio amplifier makes for some interesting listening. There are some really fun light sources to listen to.
Seen a couple of direct capture light-moving-across-a-room videos but this is by far the cleanest and most impressive I've seen. The explanation of how it was done was great and... illuminating. (Couldn't resist.) Amazingly done!
The best setup for scanning a scene is using an equilateral polygon mirror like a hexagon, the rotating polygon mirror mechanism is also known as a polygon scanner, it gives you a precise continuous scan with no jerking motions, as you don't stop or accelerate the mirror, as it spins, your line of sight pans the scene, and when it reach one edge to the next your perspective resets, like teleporting the mirror to its original starting angle.
Also, another way to 'cheat' is using the high frequency laser in a material that fluoresces with a lower frequency light
This is what I came here to say
Aside from an insane project. The editing in this video is fantastic! Well done!
7:55
"Hey, what's your camera's resolution?"
"10 kilopixels per second"
We hitting them scrub-low figures fr fr
9:34 you just need to spin the mirror constantly with a very precise and high rpm and shift the phase compared to the laser pulses slightly on each rotation. This way you don't record a single pixel in time but swipe through the scene at a specific time of start with a specific delay for each location.
So you can match each location depending on the start time and delay compared to the start time of the laser.
This should be much faster than a stepper motor, and you record a much higher resolution, as your resolution is now just determined by the rotation speed of mirror and the resolution of your Oscilloscope.
So, basicaly, actually make a LIDAR.
A stepper motor is what u might want to use for better resolution of movement compared to the servo. They are specifically designed to have exact control over how much the shaft spins, repeatedly at a high resolution. The motor itself can spin at precise fractions of a degree, and you can always use gear ratios to take it even further but then you have physical tolerances and slop to deal with. But if you managed to make something work at all with a servo, the built in resolution of a stepper motor alone should be plenty to get the accuracy you are after.
After typing this out i noticed that u had microstepping as an option written on the screen hahaha. But ya thats what id think to look into first!
A laser galvo would probably be a good idea - moves very fast, so should be able to get a decent imaging time
Problem is the imager is being scanned, not the laser. You'd have to move the imaging path but it has a huge aperture - you'd need a mirror the size of the camera, which is exactly what he already has! (Kind of funny to think of that setup as the world's worst laser galvo.)
The problem is the laser galvo input aperture is going to be a magnitude or more smaller. Though maybe that removes the need for the secondary aperture?? I think closed loop servos for the existing mirror are a better incremental idea.
Yes this is a great idea. They are typically used for laser light shows. I in fact have already been using one for a similar project (not light TOF, but instead other light properties) and it works great. The aperture is a little smaller, but I don't think it should be too much of an issue if you average frames. I'm curious if you can but a large lens in front of it if you really need more light, but I would recommend trying just the galvo first.
26:01 'Oh, ('ehh') I don't see the laser - I guess the glasses really work...' (I'da flipped them up to find the laser, lol); I don't know of anything cooler than this experiment, butt others' suggestions are great!
Suggestion: I would love to see single/multiple slit interference of light over time! Not just at the detector/wall but also the path it takes visualised with some medium like smoke. Amazing video as always, keep up the amazing work
Holy shit!!! This is incredibe!
I was gobsmacked by the exact same video that inspired your project. Up until that moment I only thought of cameras as taking the whole scene at once, and I have been a fan of high speed video since I saw a book by Harold Edgerton back in the 80s. I couldn't understand how they had a sensor that was orders of magnitude better than the state of the art, after watching the video five or six times eventually pieced their description together to arrive at the implementation. Your explanation is amazing and is SOOOO much better and more accessible.
I had an existential realization during that original video. I realized that we weren't seeing the laser pulse as it passed through the scene. We were seeing the laser pulse as it passed through the scene and travelled to the sensor. The sensor was blind to what was happening in realtime. This made me realize *I* was blind to what was happening in realtime. That my eyes weren't actually 'looking into space', but that they were only seeing what was hitting my retina...like my eyes were tastebuds for light...and that I was actually in a way blind to the world.
Not sure if that makes sense, but that exact same oscilloscope is sitting behind me. I also have a cheap ebay 2 axis galvanometer in the basement (could be a recommendation instead of the hobby servos). Might have to actually try this!!!
I think this raises an interesting philosophical question about the objectivity of the human point of view vs a theoretical "objective" view of a scene
18:28 I was thinking "You need photomultiplier for that" throughout the whole video and here i was relieved!
Interesting video. I actually worked on a commercial LIDAR system. So I found it interesting when you showed the IR modules. One does need to be careful with exposing your eyes to too much IR light. We had protective glasses that we used when in the same room when the IR source is on.
25:13 the moment
11:08 You could have kept a focused laser beam, added mirrors so the light goes from side to side multiple times and finally added some smoke to see the beam travelling.
Dude your a genius and the way you explain things even the layman can understand
Absolutely incredible
I enjoyed your video so much thank you!
08:03 - careful, your bottleneck quickly becomes the oscilloscope, showing a sample rate of 1 x 10^9 samples / second = 1 sample every ns, and you are operating in that domain. also reflections and impedance matching (47 Ohm + blue tape) :-)
The the real friend was the 100MHz scope limit we tried to bludgeon to death with sensors along the way
@@AlphaPhoenixChannelYour probes probably aren't doing you any favors either. You may need an actual high speed scope with high speed probes.
Awesome again! Love the DoE. For more accurate motion, try alt-azimuth telescope mounts (fork or not). Most are USB serial interfaces and very high angular resolution. Older models are cheap used.
hello, im a new subscriber. i enjoy the videos that you are making. your easy to understand and i have learned alot from you. i am going to binge watch your content now. thanks
Amazing setup, congrats!
1. Shoot lazer from the "camera" position as short pulse as posible, I think you then make depth picture, each next frame will show farther objects
2. Two hole experiment, maybe some more complex variations
3. bunch of mirrors, also laser shoud be focused to parallel beam
4. fog machine and focused laser
The fog machine probably wouldn't work well, as there's no way to stop the fog from moving.
Four greatest words in history: Can we weaponize this?
@@antonliakhovitch8306 make it consistent over whole room with fan
@@emovard3n You kinda need the exact opposite: neutrally buoyant particles and completely still air (I don't think this is practical).
Have you ever seen a fog machine? You can see when that stuff moves. A fan would just... move it more. Fog consists of heavier-than-air particles. It doesn't want to stay evenly mixed with the air.
@@antonliakhovitch8306 I think turbulent flow from far away or low resolution will show perfect picture. Main issue is overall fog density changes over time as "exposure" of such camera is long
You had mentioned sparks, and I wanted to say how crazy it is that on a rotary engine running at 9000RPM, each coil is firing a spark 150 times per second, and my dwell timing is at about 4-5 milliseconds. (Dwell=time spent charging the coil before it fires off). But my coil is firing a spark every 6.66 milliseconds. Meaning it has about 1.5 milliseconds of rest.
This is insanely fast to me. But after watching your video, your tests showing milliseconds and nano/micro seconds REALLY put things into perspective for me, for things I never really had a reference for before. Thank you for that! I now have an actual reference point for how ungodly fast this actually is, in a way that I can actually SEE. Love it!
Milisecond range isn't really that fast, if you think about sounds. 9kRPM engine just produces a buzz, as 150 times a second is still just 150 Hz, the bass end of the audio spectrum.
Now, a speaker outputting just 10kHz (which is still only halfway to the top of the audible range, for humans that is) is vibrating 10000 times a second, which takes 0.1ms time for the ENTIRE back and forth movement, so in one direction it is 50 microseconds.
And we can admit that the audio range is still almost DC compared to really high-frequency / high-speed things.
Awesome work! Suggestions:
-Light splitting into a spectrum through a prism
-Superfluid Helium Flow
-Cymatics
Its possible the most effective way to step the mirror would be not to - go super old school and use a pair of rotating mirrors like lissajous lasers and read the position instead of controlling it - no acceleration might mean less shake
I’ve been fascinated by the idea of trapping photons in a mirror box-using a two-sided mirror to let light in, but preventing it from escaping.
That research team that made the bottle video featured in this video as inspiration, they have a video of a single photon bouncing in a mirror box. It’s amazing! Check it out!
this guy deserves more views and subscribers. Thanks, i have understood so many things, that i knew, but never gave a proper thought on it
4:50 this guy is pulling a @ZachKing on us here
This is amazing! Regarding the "motion blur", it might be possible to apply deconvolution in the temporal dimension of every pixel. The PSF could be generated from a reflection free single laser pulse observation to model the ramp up and ramp down of the system response.
If he could somehow make a shutter operate with nanosecond precision (even if it takes many nanoseconds to open and close) then maybe he take full scene photos instead.
Maybe Electronic shutters are precise enough, if he could somehow write some code to read specific pixels at a certain timing.
I’ve always been very close to, but not quite impressed by your videos. This project, however, is the most impressive I’ve ever seen on RUclips. Absolutely fantastic work
I would want to see how you put that hole in the board on a second channel video, looks awesome. I liked the video.
He pulls out a green card from the box, and they edited it to make it seem like a 'hole'. Then when he puts it on the black board it's shot of a real hole form there onwards.
You can see it in the previous shot behind me at the table
could you make a "mirror maze" from the side where you can just see the path that the light takes when bouncing back and forth to many different mirrors? that would be fun.
you could use gearing to reduce the problems with the servos, if they are 360 degree servos. better yet, stepper motors. more importantly: your setup needs to be wobble-proof. make it heavy, no limp connections, etc. that's what's limiting your resolution as well IMO.
this!
prolly need lenses to keep the light focused
@@mymemeplex nowhere did I say he cannot use the lenses. Idk exactly what you're talking about.
@@JustinKoenigSilica Sometimes a comment is just a comment, not a criticism. I'm wondering how my comment makes you feel I'm attacking you?
@@JustinKoenigSilicaJesus chill lmao
Dude absolutely amazing ,
id love to see a larger coil of some fiber optic wire in front of the laser , so that when the light hits it loops around and around , by the time the laser hits the wall, it would still be like halfway in the cable.
you should use a 0.9° stepper with a cycloidal reduction to increase resolution and reduce backlash as much as possible. That could give you smoother movement and as good of a resolution as precision you can put into that cycloidal reduction, which id recommend using bushings and thin oil since you want precision and dont have a huge dynamic or static load.
Or just use a much smaller mirror. That was a lot of mass being swung around. Waiting a little longer after stepping for the mirror to stabilize just means the image capture is a bit slower, so not that big a deal. Just convenience.
A slow, steady movement would probably be more precise than stepping. At nanosecond frame rates, even a fast moving mirror will appear to be stationary.
9:36 I have spent my fair share of time on this problem and what I would suggest is stepper motors with a belt reduction. The good thing about steppers is, that they have a very predictible motion and are very cheap. I would not run them without a reduction, since they tend to vibrate a lot if they are driven at a low speed. Microstepping, even with the best drivers (TMC last time I checked) isn't enough if you want truly smooth motion. The good thing about belts, is, that they barely have backlash, which makes them also very good for precise motion. If you want extra accuracy, I would suggest mounting encoders to the axles. Alternatively, you could use an ODrive with a brushless motor, but that is way more expensive and probably wouldn't be that much of a benefit for such a low load application.
I just finished building something with a decent tmc driver, and I had debug issues because the test motor was moving so slowly, so quietly and completely without vibration. Not particularly fast, but I think they'd be a huge improvement on hobby servos. I'm a bit amazed they worked so well at all! For this level of precision
I also suggested an ODrive in another comment. It would probably end up being ~$400 for the whole setup, but that would be a very useful item to have for future projects that need any kind of highly precise motion that won’t be affected by the vibration inherent to steppers.
Another thing that comes to mind is that it might be possible to use a VESC, which I believe can be had for less money than an ODrive (but would also probably be more difficult to get setup).
I am also wondering if the mirror could just be spun instead of moved on an arc. It’s pretty easy to set up brushless motor controllers to spin a motor at a specific speed, so if that will work for this application, it could reduce the costs significantly.
For such low power application, I would suggest the SimpleFOC library for arduino. Lots of example and good documentation. It allows to use gimbal motors with FOC, but also stepper motors. Really easy to use and way cheaper!
@ oh cool! Somehow I hadn't heard of that before. Yeah, that probably would be a very good way to go.
One other thing that came to mind, a while ago I came across a TMC chip (I think it's the TMC4671) that integrates all the functionality in hardware to drive brushless motors with FOC, which also could be a good way to go. I have the dev board, but haven't gotten around to doing anything with it yet.
Gimbal motors with high phase resistance can be easily driven with a SimpleFOCMini (around 4-5 USD) and an Arduino (can be less than 10 USD depending on which model) running SimpleFOC. If you need more capability than most Arduinos, you could opt for an ESP32-based device, such as those based on the ESP32-S3 (3-6 USD).
Love the ACME Prints on the Box
I just have a doubt, you said when the laser turns on it sends a signal to the oscilloscope and it starts recording, but the current (electrons) travel with a speed closer to the light (still
the signal to the laser is also delayed, it doesn't matter what the delay to one or the other is, as long as they're roughly (within a few nanoseconds) synced
@@fuby6065yes, but the laser turns on first and then it sends signal to the oscilloscope, there is definitely a delay gap, maybe it's only a few nano seconds and negligible.
the delay is consistent and can be compensated
Digital scopes can capture some number of samples before the trigger point as well as after, so it's just a matter of extracting the appropriate part of the data.
Gotcha
Light travels 29.98cm in a nanosecond. It's already closer to a metric value. Like, just redefine a foot to 30cm, it's only 5mm shorter and you'll all become "taller"
The meter is as long as it because the polar circumference of the earth was defined to be 40,000 km. Some error propagation later, they made a platinum-iridium alloy rod the length to match and made it the definition of meter. And then they had to match that with every successive definition including the constant definition. The speed of light is now the definition of a meter by way of a precise clock.
30cm: ~0.2 mm error
1 ft: ~5mm error
just to make it more obvious
WolframAlpha queries: "1 ft - light-nanosecond in mm" and "30cm - light-nanosecond in mm"
Whose foot some King's foot? Not my foot my foot I can't find shoes in my country. I have a 48 or an American 13
@@IsYitzach
One centimeter is a side length of one Kg of water.
You guys sure do get worked up over something that doesn't affect you at all.
This is a FUN place to hang out!!! Thanks AlphaPhoenix!! I'd like to seethe double slit interference pattern from above at a billion frames per second!
2:00 "less than a thousand dollars"
Literally constraint optimized that equation!
Not quite the fastest anything ever moves unless you are evacuating your garage of all air😂
This is honestly the first video I’ve seen from you. Based on what I’m seeing, I am definitely here to stay. BUT I have some fun RUclips algorithm happenstances that seem too obvious not to point out. First off, this video did not pop up on my recommended page, it popped up as a recommended video to watch AFTER watching The Thought Emporium’s video on the XJ-9, an antenna scanner that reads wifi waves. The interesting thing about it is what both videos have in common: the clip from the Iron Man movie “TONY STARK WAS ABLE TO MAKE THIS IN A CAVE WITH A BOX OF SCRAPS.” And also, besides the fact that I obviously watch this nerdy side of RUclips on a regular basis and am far from being ashamed about it, I also have been watching a lot of Shorts involving clips from the MCU. Just something I noticed. Maybe someone should look into it and see if it’s worth capitalizing on.
0:37 "I like using feet"
I see. I see...