@@BennJordan At 15 mins that is the tweeter voice coil heating up not heating from sound waves. Would be interesting to see a horn speaker directed at a surface you film. That would show sound impact
Next time you're looking, record the audio of that cricket on your phone, and play it back to the cricket. They get louder, and chirp much more... Not sure that's territorial or sexually influenced, but it might help you find where he's hiding.
I believe the heating you observed on your tweeter is the inductive heating of its voice coil. The woofer doesn’t heat up because the crossover blocks high frequencies. The two best ways to heat up a speaker (or coil) are either really low (DC resistive) or really high frequencies (inductive).
The tweeter is heating up because the voice-coil mass is tiny compared to the woofer, however the power being delivered to it is the same as the woofer (assuming it has roughly the same impedance response). In addition to this woofers dissipate heat more quickly as they have larger excursion and so air convection supplies some cooling.
A decade ago I worked with a special Flir camera in the petrolium industry, that could see gas in the air much like you showcase with the lighter, but much more clear and without the need of a background. This camera used a sterling motor to chill the sensor to extreme cold, and the sensor could detect temp difference within one degree in extreme saturation. You could see foot on the floor minutes after walking there. I bet this camera could have the thermal resolution to capture sound and reverse it. Thank you for the amazing video!
I'm choosing you since you seem to know about cameras. Why in the digital age do we still have rolling shutters! Why can't the sensors take a full picture at the same time. Rolling shutter seems to be an issue for so many things is all. I get why old film needs it. Not digital!
@@dianapennepacker6854 i am not an expert, but i think it has to do with the amount of data captured at once. Shooting raw you still need dedicated recorders that are essentially a stack of ssd disks, and the camera needs to process the signal. Most cameras (phones, drones, dsrl) do not have the processing power to do so real time. So you could have global shutter, but very bad bit rate and low data volume, which would turn to poor quality and artifacts. A rolling shutter can have higher bit rate and better quality real time because it is only a fraction at the time. Is my guess.
@@davidgallefoss1254I agree. It’s in the data processing. It think the cameras essentially read the way tube TVs made image frames. Starts at the top and zig zags down. YET! Those TVs had an instant response as compared to flat screens where playing Mario can feel impossible because the image is being processed, upscalled, and stretched in between me pushing jump on the controller. I do wonder how much more crisp an image would be if instead of one dot zig zagging up and down. You had multiple processors tied to the same sensor but different spot. If that makes sense. Each pixel has its own processor core. I think for once we might actually be able to delete the motion blur effect given the exposure is modified accordingly
In Yukikaze novel, they used "The Frozen Eye" device, that was some kind of super-cooled camera to detect any air disturbances. It was used to detect enemy airplanes that could otherwise hide from radars and human eyes, but still took volume in the air.
@@PeterDeGree Would've been a fitting appearance since it's way less expensive than the imagers mentioned here. But even at that price I wouldn't recommend it to a consumer. I know a DIY assemble yourself version would do well with a subset of the audience here, but then you're offering more than just an imager :)
I love how he goes into detail on uses. Definitely subbed for that even if the sound camera was not what I imagined it would look like. Thought there would be... Actual sound waves, lol. Like multiple ripples clashing. Not there yet. We will be I am sure.
I would like to say that for the heat test at 15:00 you're probably picking up the heat increase from the flexing of the rubber on the tweeter than the sound itself. the camera is even explicitly picking the ring to measure from. that doesn't mean it's not possible to measure how much heat is being transferred though sound, but this measurement was flawed.
I think that heat mostly generated by the speaker coil. For the thermal cameras, the heat in the air is invisible. Another thing, that is true the compressed air is warmer than the uncompressed, but the sound wave has a negative pressure part as well what is cooler than the normal air. The average temperature of both is equal to the silent air.
Visualizing sound via heat doesn't seem really likely. What is shown in this video seems to be more about the coils of the speaker heating up because of joules effect, that would make more sense I think. I would have to check with the right formulas but the energy carried by a sound wave isn't near enough what it takes to heat anything (at least under safe hearing levels). Plus the energy of the waves decreases proportional to r³ (sphere)
Ben, I’m so glad the YT algo intro’d you to me. I too am an adhd musician with nerdy curiosities 😂 You tend to do things I would do if I had resources and wasn’t dealing with debilitating health issues 😅
It's time for a Slo Mo Guys collab Benn. You must have some ideas in the bank for using the phantom cameras... And I'm sure Gavin Free would love your work.
A high speed camera could pick up much smaller frame differences, potentially allowing you to create a visual microphone. Considering how expensive those cameras are...
@@rhr-p7w What? Smarter evryday had massive respectand aoften works with gavin and dan of slo mo guys. They loan eachtohers camera's back and forth. Also, slo mo guys have never once claimed to be in any way scientific. At most edutainment with some of their stuff. And lastly, Respect to Benn... He puts in the work.. but in fairness, he much better fits the term you gave of ' pop-science.' I wouldn't label either channel as such, but Ben actually makes genuine big claims around different fields of science quite often, where the former doesn't. And lastly, they are very charismatic and have the tech that Ben can't access.
Looking forward to this! I got to work with aero acoustic wind tunnels in college and the microphone arrays they create to pinpoint sound are very impressive
There's an excellent combat use for this. You could mount it on a vehicle and use it to detect the location of a sniper. The US Army was trying to figure out how to do this in the Iraq war, to track down snipers who were firing from civilian locations. No idea if they ever figured it out. But this would absolutely do it.
Yeah. Already working. Really well. Uses an array of microphones arranged around the wearer's military helmet. Then on a head up display over the right eye has a crosshair which centres over the zone. And the cross hair moves with the wearer by using motion sensors. Present development is to integrate it along with many distributed wearers, each with GPS to precisely geolocate a target. Then probably the next thing would be to integrate that with a "hawk elimination device"...
As a retired tech teacher, your genius ideas resonated with me and made me a new subscriber. I was struck by the thought that animals, like dogs, insects, birds, etc., are all operating on parts of the world that humans have no perception of. That is how mosquitoes find the best blood source, or dogs sniff out drugs through walls, or even how is it that rare species find each other in order to reproduce? As a human, I OUGHT to be able to tap into the different sensing methods that animals use by using my mind and the instruments that open up those worlds. We are making good progress towards that and discoveries are being made every day. I am excited to be able to follow your journey and see what else you might be discovering!
When you showed your camera identify a hawk’s location and which chicken clucked, I immediately thought this would be great to track down where fireworks were ignited or direction/location of a run shot. I also thought this is probably in use for spies and if not, should be. Or that it could be useful for surveillance. The ideas were just flowing fast and furious! Thanks - I learned a lot and got lots of ideas!
I think in the Netherlands they already do this. Wouldn't surprise me if they did this in many other countries as well. Edit: Lookup "Soundbrella" and read the articles by AD, NOS and gemeente. This is translated from NOS: " Leiden calls in the army to combat fireworks vandals A decent firework bomb will of course not go unnoticed, but in Leiden the chance that the police will hear your Cobra 6 is even greater. The army has placed microphones on the roof of five flats in the city, which immediately signal where it came from after the explosion. Within no time the police receive an email with the exact location of the firework bomb. "It is not the case that officers wait at the station all day until a report comes in. A lot of officers are already driving through the city and can be on the scene quickly. The chance of catching them is much greater," says mayor Henri Lenferink. Leiden war zone The army normally uses the system to measure where shots come from and where mortars fall. Although Leiden may not be a war zone, according to local residents it does seem like one in the weeks before New Year's Eve. "More and more heavy fireworks are being set off. That is not only on New Year's Eve, but also in the period before that," says the mayor in the AD. Soundbrella The soundbrella, that's the name of the system. According to Major Henk ter Kulve of Defence, a unique Dutch invention. "This sensor can measure any form of sound and knows exactly where it comes from. The microphones can hear the fireworks from three kilometres away." If it turns out that fireworks vandals are indeed caught faster by the system, other municipalities can also start using it. "
Some kids at defcon figured out you can use rolling shutter to steal secrets from badge readers. The different key steps take different amounts of power, and so the LED will dim slightly more for a SI step.
I can't wait for part two "The illegitimate sound camera" using quarter stick dynamite, multiple sound sensors, and triangulation software to stick real time 3d acoustic mapping of variable density.
Awesome stuff, especially the first section! Couple nits: 10:22 -- rolling shutter affects mechanical shutters as well (not that you said otherwise, just saying), and 13:30 -- as others have said, it seems much more likely that the thermal effect has to do with electrical resistance and conduction through the speaker, rather than the heat of actual sound waves. I have seen estimates somewhere (no links off hand, sorry) about the amount of heat-equivalent energy contained in sound waves, as well as the displacement of air molecules in a sound wave, and both were very very small.
This seems like a good tool against deep fake detection. I think Microsoft used this to determine bloodflow of the subject vs a deepfake of a subject. That said, I don't think it would be hard for models to add this lifelikeness to generative output.
depends on the type of deepfake, if it's a modified existing video then it probably wouldn't work, if it's an animated photo then it would, if it's a video made 100% by ai then the ai probably learned to mimic subtle patterns like that already would be neat to check though
Rolling Shutter is a surprise tool that will help us later. If you can figure out the exact timing between rows you can increase the temporal resolution of your recording times the number of rows, as long as you use a zoom lens to fill most of the sensor with a single object. I was almost sure you mentioned that in your laser microphone video, but I just skimmed it again and didn't find that mention
Yes. I was thinking the same thing. If you used a single line from each image and turned it into 30`000 fps (1080 lines x 30 fps-ish), there would not be much to see on each individual frame, but you'd get rid of the "time dilation" rolling shutter produces within a single frame. Possibly interpolate the rest?
Your videos are quite good for many different reasons, but I’ve really noticed the practicality of time it takes you to exhaust a topic never exceeds the attention span granted for that topic - or something like that or whatever. I’m not sure what the algorithms say about your audience engagement, but I know never feel like finishing a video early because they are too long or contain tmi, or the opposite. It’s just a bonus that I’m also about your age and have always been involved in art and music.
@@jaromir_kovar did you watch the video that was open on that tab? it was a 2 hour long lecture from a russian dude on something I don't quite remember I woldn't say it was a joke, I think it was ~research~😁
I always learn interesting things from these videos. I can't lay claim to anything close to Benn's engineering chops, but his mention of uncovering hidden layers of information reminded me of a recent experience where I was experimenting with mid-side audio processing. I was using a random audio clip, which happened to be from the 1950s production "Wonderful Town." It's a big number in which Rosalind Russell is trying to interview an unruly group of Brazilian marines who are only interested in learning the Conga dance. It's goofy and cheerful and kind of hilarious, and it's one of my favorite clips. These older recordings are really interesting, and this one seems to have been mic'ed so that the center channel contains most of what I've always thought of as "the song,"-which can mask a lot of surprises in the side channels. In this case, it seems the side mics at times were a lot closer to the men's chorus than the recording engineers accounted for. It's not obvious in the recording until the center channel is attenuated-and then, wow! As clear as a bell, you can hear all kinds of weird sounds coming from the chorus attempting to hit their marks counting off a tricky syncopated rhythm, and at other points when they are making crowd noises during which their individual voices are not intended to be discernible-It was like uncovering a secret message in a time capsule, and it revealed something of the personality of one particular member of the chorus-it's doubtful he ever knew his voice is so clearly audible, but he definitely puts the "boy" in "flamboyant." You can hear him chatting with somebody, and enthusiastically counting steps, and I'm pretty sure he playfully refers to Rosalynd using a word I didn't think people knew in the 1950s. It made me hope that they all were enjoying themselves and that he had a long and successful career onstage.
This reminds me of the video where researchers reconstruct the audio of a speech by filming the vibrations on a potato-chip bag through a soundproof glass. There was also a technique that was used during the Cold War where a country managed to have the windows of an embassy replaced by a specific glass, to allow them spying on the conversations by measuring the tiny vibrations on the glass, from the outside (although it was not by using cameras but rather a laser microphone)
When the tweeter got warmer, you were just seing the heat from electrical losses in the voice coil, not heat from the sound wave itself. A better test would be film a material that absorbs the sound waves from the tweeter, e.g. acoustic foam placed opposite of the tweeter. But frankly, I doubt it would be measureable in practice. Anyway, thanks for the great video!
The "acoustic" heat from your speaker was more likely the lossy component of the speakers' materials' elastic/viscoelatic structure (flex something vigotously enough and it gets hot). Potato/potahto...this is inspiring and very well done!!
I remember leaving a comment suggesting this idea a few months ago! Don't know if you read it or not, but either way, I'm super excited to watch the video! I think this sort of technology is absolutely awesome, and never gets enough attention.
This was such a cool video man! I'll be honest, I don't understand much of your detailed music videos but I still like watching them. All the other videos though, they are so engaging and you've done just such a wonderful job! I truly don't understand how you don't have more subscribers but you got at least one more from me. I found you from the ADHD meds video and spiraled from there. Also, 15:33 that ADHD hyperfocus is real lol. Keep making just fantastic videos!
Wow, this blows my mind. I made an algorithm nearly 20 years ago that took an image stream in, then converted it to monochrome, then compared the previous frame with the current frame, and was able to outline movement with pixel precision in an overlay over the original colored image. I also worked with sonar arrays to achieve the same thing for sound detection overlay. With all that said, I never considered if it would be possible to visually see the actual sound wave. I am now going to Google the hell out of it, lol
I don't know if you will see this Benn, but I have been learning from you since like 2013, I was a kid and came across your music on youtube haha, just wanted to say thank you man! We appreciate you bro
Hey! We followed the same path of thought / information search ^^ After watching Steve Mould's video on motion amplification, I found some of the Github repositories you mentioned, and a few weeks later the Posy video! But you followed the path way longer than me ^^ It's a bit useless as information but it made me laugh :)
Ever since I saw that Abe Davis TED talk in 2015 I"ve had an idea for using acoustic imaging to find anoamlies in a room, and even for finding, say, studs or other framing members behind drywall during a renovation or DIY so that the screw goes in the right place the first time. I never pursued it and thanks to this video I think it might be time to revisit the idea!
Ive screenshotted your message and forwarded it (together with this video) to a former colleague. She worked at the company that makes acoustic imagers that I work at and recently started a job at a company that does deepfake detections. If I receive any news then Ill follow up.
Interesting idea, but the way you usually build AI video generator is to train it on real videos so if this pattern is present in them there is a decent chance the AI picks it up.
These kind of things are already used by identity verification software as liveliness proof required by financial regulators. Meaning, when you open an online account for banking or stock trading, it is very likely that a software is used to check on the image of your webcam.
that posy video took my glitches to the next level. I made apurple carpet out of static by iterating his difference layers in premier. it is so so cool
Regarding measuring the energy content of sound: The sound intensity (sound power per area) of a 1kHz tone at 90dB SPL is a hair short of 1mW/m^2. For reference, sunlight is about 1 million times stronger (1kW/m^2). Maybe you could stick a really precise thermometer probe into some sound-absorbing foam? You'd have to pulse the sound and correlate it to changes in temperature.
This is extremely closely related to my PhD and I can tell you with sound that quiet it is not possible with current technology. The cooling capacity of the material means any slight increase in temperature would be almost instantly transported away via convection, conduction or emission. With higher pressure sound it is certainly possible.
Ben, you freaking rule. Thanks for being Ben. I wouldn't want you to be anyone else. Well, maybe a happier version of Ben. Everyone deserves to be a little happier.
you sure it is the sound generating the heat and not the speaker coils heating up from having electricity pumped through them at an ascending frequency?
I have just found your channel so this is my 2nd video of yours , ( the first was how animals perceive sound ) and i just wanted to say thanks . your videos are amazing and the way you explain things are perfect for a layman . THANK YOU FOR EDUCATING ME AND GIVING ME SOME GREAT STARTING POINTS FOR FURTHER EXPANDING MY UNDERSTANDING AND KNOWLEDGE . much love from Scotland
Finally I'll be able to see in real time the energy that's coming out of my hand and the beautiful geometric energy forms I'm projecting 😊 Thank you Benn, you're awesome.
What does "being a nerd" actually mean? That's not a rethorical question. Everytime I hear someone say or write it, it feels like "being a nerd" is somehow self-deprecating and a not-so-humble brag at the same time. But what does it *mean* ? Wearing glasses and short-sleeved button-down-shirts with a bow-tie and suspenders and carrying around a pocket calculator?
@@michaelgraflmusic to me it's being ultra curious about the inner working in life and how they interconnect and influence each other in a spiraling way
A nerd is someone who spends their time and energy to study something in great detail or experiment for no reason other than the enjoyment of it. Usually having to do with science. For a simple example, I like to think being a nerd plays out just like the Revenge of the Nerds movies. Being a nerd is originally a put down until the nerds use their smarts to become the dominant group and get the babes. Maybe not exactly like that, but being a nerd is a compliment usually because they are so desperately needed in our advanced world.
You should consider making a laser microphone. You probably already know what they are, but in case you don't: they shine a laser on a target and detect the reflected light. Vibrations in the target will ever so slightly alter the phase of the reflected light (much like how a laser interferometer works). By observing the interference pattern made when the source beam and the reflection are combined (and with a lot of math) you can extract frequency information and convert it into audio. I'll admit this is probably a lot more complex than I think it is. But I've spent hours playing with lasers and I've noticed that by reflecting a laser off of a surface that scatters the light, the speckle pattern gets magnified and projected onto the ceiling (or wall, depending on how you aim it). This projected and magnified speckle is extremely sensitive to changes in either the laser's position or the position of the surface off of which it is reflecting. So much so that I could see my heartbeat even when I was sitting in my chair on the opposite side of the room from where the laser and the reflecting surface was. Using a shorter wavelength (i.e. a 532nm green laser instead of a red laser) increases the "resolution" of the speckle, and thus the sensitivity. You could probably utilize the code used for motion extraction to do this.
This is so cool! I was born dear on one side and cannot localize sounds. Do you know how cool this would be as an AR assistive technology that could help someone like me localize a sound? It would be amazing!
Back in I think, the 1970's, there was an article in Popular Science Magazine about how GM was using sound and photography to find hairline cracks in Corvette engine parts. When I saw this title, I was hoping it would be about that type of tech, as I was recently trying to remember how they did that again because I forgot and am interested in casting metal items and it would be great for analyzing there quality.
Just a quick shout to say thanks Benn for continuing to not only make excellent videos covering a variety of music related subjects but just for plainly still being here, period. Too many YTbr's give there all, in one explosive season and then they disappear. Not you. Not you. I love your tenacity and grit. Of course I love your music that goes without saying but it's really your honest approach to transparency that I admire. Keep it up man. Cheers from Nova Scotia Canada .
Brother we have a new subscriber today! I have learned more about three topics that are important in my life and you're one video today and I have in 20 something years of sound engineering and musical producing as far as recording songs and making beats and all that jazz digitally and analog and I'm very impressed and I'm very appreciative thank you so much!
Thank you! This is like the old YT video "The Visual Microphone: Passive Recovery of Sound from Video" where they record on a full hd canon dslr and extract audio source from the images only. Mindblowing ruclips.net/video/FKXOucXB4a8/видео.html
This could've easily been a four part video series! You always cover such interesting topics, thank you for tickling my brain in regions I didn't even know were ticklish
This is by far the most interesting and informative video I've seen in quite a while. Such a compelling rabbit-hole to explore! Many Thanks for making this!
3:43 My frustration with Python exactly. I find a Python script has a shelf life of like 1 year, and if I revisit it after that I'm in for a whole mess of digging into why it no longer works and finding old versions of x, y, z....
Heh. We never figured out reproducible builds in the python packaging world. It never used to matter either, because breaking backward compatibility was considered an enormous faux pas. I've had some success with nix for repeatable python, but there is definitely a learning curve.
I dipped my toes in Python, and got frustrated by this. Well, this, and the fact that most documentation is written as a flowery tribute to how great Python is, and how easy it is to read. Of course it's easy to read. Nobody ever writes error-checking in Python, they just let the whole thing collapse with a stack trace. Turns out, C and ASM are easy to read when you are unreasonably optimistic, too! TBH, it was never going to work out. I just can't with languages that use line endings as syntax. Gimme my { braces and semicolons; }
This is where you want to be using venv and a package manager that supports semantic versioning, but yeah it's not trivial. NodeJS is a nice example of baking this in by default with NPM.
One of the physics labs I did in college was optics, and one of the cooler demos we set up was a laser interferometer. Similar in principle to what you're describing here. We were able to tune it to visualize thermal layers and flow around e.g. a small airfoil (finger sized).
nice practical brainstorming! Since some time I have searched for people who managed to invent ways to locate mosquitos, and it seems your video is a very nice source of inspiration. And no, it is not really needed to shoot down mosquitos with a laser, actually to make mosquito trap it is enough to know when they pass by to either trap them mechanically by a net or one can use either vacuum , pressurized air or combination of to remove the bug.
AFAIK the software for an audio camera thing using multiple microphones is just a 2D Fourier decomposition. It's the same problem the long baseline radio telescopes face. The optimum sensor layout isn't actually a square array, it's putting them at weird locations, as far apart as possible, that are not multiple wavelengths of each other because you get more resolution. When you don't have a fully filled array, you get diffraction pattern effects but most sounds are multispectral so the bits you don't want won't usually line up in the frequency domain.
I have no use for any of this, but your videos are so interesting and well thought out, I love learning all these seemingly magical audio things. Thanks!!
Visual reflection detection of audio is super useful for treating studio and listening rooms. Being able to see the reflections and treat them where necessary is a super time saver with no real trial and error required. And it is recommended that you rent a very high end camera for this because you really only have to do it a couple of times to map out the rooms, provided you're not the kind of person who moves things around frequently. It's hard surface reflections that cause the most trouble. Of course, the "golden ear" club will try and tell you they can do it with just their ears so they can seem like they have ultimate hearing but, the eyes are better than the ears in most cases, especially this one.
Great video, you've definitely become one of my favorite channels over the past year. Just one question: how do you know the heat wasn't generated through the speaker itself rather than the audio waves coming out of it?
I already spot the UMA-16 v2 USB mic array from miniDsp. It‘s a nice one and I‘d love to have one that even has more channels. Looking forward to your implementation…
I think Steve Mould, or someone in that same sphere, made a video about acoustic imaging a few years ago. The other day I was thinking about it again and marveling about how since we typically record sound at around 44 khz, when we use acoustic imaging we are essentially capturing the data at 44,000 fps, meaning it is inherently a high speed camera.
This concept is actually a really great way for a human presence sensor. I've seen other projects that map out bird calls to the specfic species, but to measure a heart beat could even be useful in a non invasive medical setting.
The amount of research for your videos is astounding, so many things in one video, pretty amazing. Thanks for the shoutout! I'm a fan of your channel.
Hope to meet and go down a rabbit hole some day! ❤️
Posy x BennJordan collab when
I hope you get a nice big bump from the shoutout! You're one of the most under-rated youtubers.
@@BennJordan
At 15 mins that is the tweeter voice coil heating up not heating from sound waves.
Would be interesting to see a horn speaker directed at a surface you film. That would show sound impact
this would be a dream come true... i love both of your channels SO much this would be perfect.
FINALLY I'll be able to find that dang cricket.
Next time you're looking, record the audio of that cricket on your phone, and play it back to the cricket. They get louder, and chirp much more... Not sure that's territorial or sexually influenced, but it might help you find where he's hiding.
@@BWP-u3y You can also spray WD-40 towards them and it shuts them up for about 20 minutes.
@@StubbyPhillips I guess WD-40 can fix anything
@@AverageMichaelJordans They must think they've gone deaf.
Hell yeah, amazing project!
I believe the heating you observed on your tweeter is the inductive heating of its voice coil. The woofer doesn’t heat up because the crossover blocks high frequencies. The two best ways to heat up a speaker (or coil) are either really low (DC resistive) or really high frequencies (inductive).
I agree, it's likely heat from an unideal impedance match when the tweeter is was beyond the desired operating freq.
Woofer is heating up as well, you just can't see it as the voice coil is deeper inside and heat doesn't travel that fast.
The tweeter is heating up because the voice-coil mass is tiny compared to the woofer, however the power being delivered to it is the same as the woofer (assuming it has roughly the same impedance response). In addition to this woofers dissipate heat more quickly as they have larger excursion and so air convection supplies some cooling.
Came to check the comments for this very hypothesis.
I was thinking, wouldn't a better test be to point the sound at like a wall or piece of paper and see if THAT heats up?
A decade ago I worked with a special Flir camera in the petrolium industry, that could see gas in the air much like you showcase with the lighter, but much more clear and without the need of a background. This camera used a sterling motor to chill the sensor to extreme cold, and the sensor could detect temp difference within one degree in extreme saturation. You could see foot on the floor minutes after walking there. I bet this camera could have the thermal resolution to capture sound and reverse it. Thank you for the amazing video!
I'm choosing you since you seem to know about cameras.
Why in the digital age do we still have rolling shutters! Why can't the sensors take a full picture at the same time.
Rolling shutter seems to be an issue for so many things is all. I get why old film needs it. Not digital!
@@dianapennepacker6854 i am not an expert, but i think it has to do with the amount of data captured at once. Shooting raw you still need dedicated recorders that are essentially a stack of ssd disks, and the camera needs to process the signal. Most cameras (phones, drones, dsrl) do not have the processing power to do so real time. So you could have global shutter, but very bad bit rate and low data volume, which would turn to poor quality and artifacts. A rolling shutter can have higher bit rate and better quality real time because it is only a fraction at the time. Is my guess.
@@davidgallefoss1254I agree. It’s in the data processing. It think the cameras essentially read the way tube TVs made image frames. Starts at the top and zig zags down. YET! Those TVs had an instant response as compared to flat screens where playing Mario can feel impossible because the image is being processed, upscalled, and stretched in between me pushing jump on the controller.
I do wonder how much more crisp an image would be if instead of one dot zig zagging up and down. You had multiple processors tied to the same sensor but different spot. If that makes sense. Each pixel has its own processor core.
I think for once we might actually be able to delete the motion blur effect given the exposure is modified accordingly
@@dianapennepacker6854the real reason this happens is talked about by TechnologyConnections
In Yukikaze novel, they used "The Frozen Eye" device, that was some kind of super-cooled camera to detect any air disturbances. It was used to detect enemy airplanes that could otherwise hide from radars and human eyes, but still took volume in the air.
ffmpeg implementation for Posy's motion extraction
ffmpeg -i -filter_complex "[0:v]split=2[main][inverted];[inverted]negate[inv_neg];[inv_neg]format=rgba,colorchannelmixer=aa=0.5[inv_neg_opacity];[inv_neg_opacity]setpts=PTS+(1/FR/TB)[delayed];[main][delayed]overlay=eof_action=pass[v]" -fps_mode cfr -map "[v]" -map 0:a? output.mp4
Ooohhh... Thank you! I'll take a copy of that. Glad to see so many Posy lovers :)
Wow, the power of ffmpg never stops to surprise me. Maybe should read up on it some day
This video is technically sponsored by an acoustic imager company, since I'm a Patreon subscriber that works at one
I was actually expecting to see a CAM64 make a cameo in that segment.
Thanks for that, now he's going to shoot the hawk 🙄🤣(5:23)
@@PeterDeGree Would've been a fitting appearance since it's way less expensive than the imagers mentioned here. But even at that price I wouldn't recommend it to a consumer. I know a DIY assemble yourself version would do well with a subset of the audience here, but then you're offering more than just an imager :)
@@MooshYT Luckily for that hawk Im not on the 250k tier of Patreon!
y’all got any soup over there?
I really appreciate a RUclipsr who spends time researching before filming. This is worth a sub for sure!
I love how he goes into detail on uses. Definitely subbed for that even if the sound camera was not what I imagined it would look like.
Thought there would be... Actual sound waves, lol. Like multiple ripples clashing.
Not there yet. We will be I am sure.
8:04 No matches found in the PH search, ya got some niche tastes huh
Did you try to type that URL into your browser?
There's a series of physics videos on PH in Russian 😅
try the URL...it's literally a Russian scientist posting physics lectures
@@dengyun846 Exactly! 😁I wrote the same thing 😅
"phase based motion amplification" porn is pretty niche, yeh
you don't need ph when the flashbulb has already come to make the girls horny
I would like to say that for the heat test at 15:00 you're probably picking up the heat increase from the flexing of the rubber on the tweeter than the sound itself. the camera is even explicitly picking the ring to measure from. that doesn't mean it's not possible to measure how much heat is being transferred though sound, but this measurement was flawed.
I think that heat mostly generated by the speaker coil.
For the thermal cameras, the heat in the air is invisible.
Another thing, that is true the compressed air is warmer than the uncompressed, but the sound wave has a negative pressure part as well what is cooler than the normal air. The average temperature of both is equal to the silent air.
@@raytry69speaker coil 🔉🔉🔉🔉🔉⚡️
Visualizing sound via heat doesn't seem really likely. What is shown in this video seems to be more about the coils of the speaker heating up because of joules effect, that would make more sense I think. I would have to check with the right formulas but the energy carried by a sound wave isn't near enough what it takes to heat anything (at least under safe hearing levels). Plus the energy of the waves decreases proportional to r³ (sphere)
Ben, I’m so glad the YT algo intro’d you to me. I too am an adhd musician with nerdy curiosities 😂 You tend to do things I would do if I had resources and wasn’t dealing with debilitating health issues 😅
I had no intention of watching this entire video, but I really could not stop watching it. You've got incredibly good pacing and storytelling
Same here!
It's time for a Slo Mo Guys collab Benn. You must have some ideas in the bank for using the phantom cameras... And I'm sure Gavin Free would love your work.
Nah, keep this channel clean from the dirty hands of those pop-science content-making machines. Smarter Everyday or Jeremy Fielding are the real deal
A high speed camera could pick up much smaller frame differences, potentially allowing you to create a visual microphone. Considering how expensive those cameras are...
I'm aware you can already kind of do this with a regular camera, but the fidelity is not high
@@rhr-p7w What? Smarter evryday had massive respectand aoften works with gavin and dan of slo mo guys. They loan eachtohers camera's back and forth.
Also, slo mo guys have never once claimed to be in any way scientific. At most edutainment with some of their stuff.
And lastly, Respect to Benn... He puts in the work.. but in fairness, he much better fits the term you gave of ' pop-science.' I wouldn't label either channel as such, but Ben actually makes genuine big claims around different fields of science quite often, where the former doesn't.
And lastly, they are very charismatic and have the tech that Ben can't access.
Looking forward to this!
I got to work with aero acoustic wind tunnels in college and the microphone arrays they create to pinpoint sound are very impressive
Too freaking cool. I can't wait to mute specific sounds before they even get to head.
posy mentioned! we love posy
we do!
There's an excellent combat use for this. You could mount it on a vehicle and use it to detect the location of a sniper. The US Army was trying to figure out how to do this in the Iraq war, to track down snipers who were firing from civilian locations. No idea if they ever figured it out. But this would absolutely do it.
Yeah. Already working. Really well.
Uses an array of microphones arranged around the wearer's military helmet.
Then on a head up display over the right eye has a crosshair which centres over the zone. And the cross hair moves with the wearer by using motion sensors.
Present development is to integrate it along with many distributed wearers, each with GPS to precisely geolocate a target.
Then probably the next thing would be to integrate that with a "hawk elimination device"...
As a retired tech teacher, your genius ideas resonated with me and made me a new subscriber.
I was struck by the thought that animals, like dogs, insects, birds, etc., are all operating on parts of the world that humans have no perception of. That is how mosquitoes find the best blood source, or dogs sniff out drugs through walls, or even how is it that rare species find each other in order to reproduce?
As a human, I OUGHT to be able to tap into the different sensing methods that animals use by using my mind and the instruments that open up those worlds.
We are making good progress towards that and discoveries are being made every day. I am excited to be able to follow your journey and see what else you might be discovering!
When you showed your camera identify a hawk’s location and which chicken clucked, I immediately thought this would be great to track down where fireworks were ignited or direction/location of a run shot. I also thought this is probably in use for spies and if not, should be. Or that it could be useful for surveillance. The ideas were just flowing fast and furious! Thanks - I learned a lot and got lots of ideas!
I think in the Netherlands they already do this. Wouldn't surprise me if they did this in many other countries as well.
Edit:
Lookup "Soundbrella" and read the articles by AD, NOS and gemeente.
This is translated from NOS:
"
Leiden calls in the army to combat fireworks vandals
A decent firework bomb will of course not go unnoticed, but in Leiden the chance that the police will hear your Cobra 6 is even greater. The army has placed microphones on the roof of five flats in the city, which immediately signal where it came from after the explosion.
Within no time the police receive an email with the exact location of the firework bomb. "It is not the case that officers wait at the station all day until a report comes in. A lot of officers are already driving through the city and can be on the scene quickly. The chance of catching them is much greater," says mayor Henri Lenferink.
Leiden war zone
The army normally uses the system to measure where shots come from and where mortars fall. Although Leiden may not be a war zone, according to local residents it does seem like one in the weeks before New Year's Eve. "More and more heavy fireworks are being set off. That is not only on New Year's Eve, but also in the period before that," says the mayor in the AD.
Soundbrella
The soundbrella, that's the name of the system. According to Major Henk ter Kulve of Defence, a unique Dutch invention. "This sensor can measure any form of sound and knows exactly where it comes from. The microphones can hear the fireworks from three kilometres away." If it turns out that fireworks vandals are indeed caught faster by the system, other municipalities can also start using it.
"
Some kids at defcon figured out you can use rolling shutter to steal secrets from badge readers. The different key steps take different amounts of power, and so the LED will dim slightly more for a SI step.
I can't wait for part two "The illegitimate sound camera" using quarter stick dynamite, multiple sound sensors, and triangulation software to stick real time 3d acoustic mapping of variable density.
I heard he is already recording it. (there are explosions near my house that are making my day miserable)
so a gun
Seems like the kinda tech the military would be very interested in lol... Who am I kidding they probably have it
@kans4629 the gov can see in the dark with their alien tech, spaceships //puts on foil hat// 😂😂😂😂
Ok, yup, I am doing this project. It's like peering into a whole new world. My son is going to love this, fantastic video!
Awesome stuff, especially the first section! Couple nits: 10:22 -- rolling shutter affects mechanical shutters as well (not that you said otherwise, just saying), and 13:30 -- as others have said, it seems much more likely that the thermal effect has to do with electrical resistance and conduction through the speaker, rather than the heat of actual sound waves. I have seen estimates somewhere (no links off hand, sorry) about the amount of heat-equivalent energy contained in sound waves, as well as the displacement of air molecules in a sound wave, and both were very very small.
This seems like a good tool against deep fake detection. I think Microsoft used this to determine bloodflow of the subject vs a deepfake of a subject. That said, I don't think it would be hard for models to add this lifelikeness to generative output.
depends on the type of deepfake, if it's a modified existing video then it probably wouldn't work, if it's an animated photo then it would, if it's a video made 100% by ai then the ai probably learned to mimic subtle patterns like that already
would be neat to check though
Posy is awesome. I got hooked on the 7 segment series, and his unearthly beautiful voice.
Yeah! Personally, One of my favorite works on on the internet
Rolling Shutter is a surprise tool that will help us later.
If you can figure out the exact timing between rows you can increase the temporal resolution of your recording times the number of rows, as long as you use a zoom lens to fill most of the sensor with a single object. I was almost sure you mentioned that in your laser microphone video, but I just skimmed it again and didn't find that mention
Yes. I was thinking the same thing. If you used a single line from each image and turned it into 30`000 fps (1080 lines x 30 fps-ish), there would not be much to see on each individual frame, but you'd get rid of the "time dilation" rolling shutter produces within a single frame. Possibly interpolate the rest?
Your videos are quite good for many different reasons, but I’ve really noticed the practicality of time it takes you to exhaust a topic never exceeds the attention span granted for that topic - or something like that or whatever. I’m not sure what the algorithms say about your audience engagement, but I know never feel like finishing a video early because they are too long or contain tmi, or the opposite. It’s just a bonus that I’m also about your age and have always been involved in art and music.
8:06 the porn hub tab 😂😂😂
Man, I would never look up "phased base motion amplification" on pоrn hub, but now I kinda want to
How did I not notice that 😂
It's real... After the porn ad, it's a 2 hour lecture of a paper by a Russian scientist.
With a search query for a fetish only people with extremely niche kink would consume 😁
Well spotted, friend, and what a nice hidden joke by Benn! ♥
@@jaromir_kovar did you watch the video that was open on that tab? it was a 2 hour long lecture from a russian dude on something I don't quite remember
I woldn't say it was a joke, I think it was ~research~😁
I always learn interesting things from these videos. I can't lay claim to anything close to Benn's engineering chops, but his mention of uncovering hidden layers of information reminded me of a recent experience where I was experimenting with mid-side audio processing. I was using a random audio clip, which happened to be from the 1950s production "Wonderful Town." It's a big number in which Rosalind Russell is trying to interview an unruly group of Brazilian marines who are only interested in learning the Conga dance. It's goofy and cheerful and kind of hilarious, and it's one of my favorite clips. These older recordings are really interesting, and this one seems to have been mic'ed so that the center channel contains most of what I've always thought of as "the song,"-which can mask a lot of surprises in the side channels. In this case, it seems the side mics at times were a lot closer to the men's chorus than the recording engineers accounted for. It's not obvious in the recording until the center channel is attenuated-and then, wow! As clear as a bell, you can hear all kinds of weird sounds coming from the chorus attempting to hit their marks counting off a tricky syncopated rhythm, and at other points when they are making crowd noises during which their individual voices are not intended to be discernible-It was like uncovering a secret message in a time capsule, and it revealed something of the personality of one particular member of the chorus-it's doubtful he ever knew his voice is so clearly audible, but he definitely puts the "boy" in "flamboyant." You can hear him chatting with somebody, and enthusiastically counting steps, and I'm pretty sure he playfully refers to Rosalynd using a word I didn't think people knew in the 1950s. It made me hope that they all were enjoying themselves and that he had a long and successful career onstage.
Posy got shoutout by Benn? this makes me so happy for some reason
This reminds me of the video where researchers reconstruct the audio of a speech by filming the vibrations on a potato-chip bag through a soundproof glass. There was also a technique that was used during the Cold War where a country managed to have the windows of an embassy replaced by a specific glass, to allow them spying on the conversations by measuring the tiny vibrations on the glass, from the outside (although it was not by using cameras but rather a laser microphone)
When the tweeter got warmer, you were just seing the heat from electrical losses in the voice coil, not heat from the sound wave itself. A better test would be film a material that absorbs the sound waves from the tweeter, e.g. acoustic foam placed opposite of the tweeter. But frankly, I doubt it would be measureable in practice.
Anyway, thanks for the great video!
The "acoustic" heat from your speaker was more likely the lossy component of the speakers' materials' elastic/viscoelatic structure (flex something vigotously enough and it gets hot). Potato/potahto...this is inspiring and very well done!!
I bet you could also turn some of this kit into a really great ANC headphone setup...if not particularly portable.
I remember leaving a comment suggesting this idea a few months ago! Don't know if you read it or not, but either way, I'm super excited to watch the video! I think this sort of technology is absolutely awesome, and never gets enough attention.
This was such a cool video man! I'll be honest, I don't understand much of your detailed music videos but I still like watching them. All the other videos though, they are so engaging and you've done just such a wonderful job! I truly don't understand how you don't have more subscribers but you got at least one more from me. I found you from the ADHD meds video and spiraled from there. Also, 15:33 that ADHD hyperfocus is real lol. Keep making just fantastic videos!
Wow, this blows my mind. I made an algorithm nearly 20 years ago that took an image stream in, then converted it to monochrome, then compared the previous frame with the current frame, and was able to outline movement with pixel precision in an overlay over the original colored image. I also worked with sonar arrays to achieve the same thing for sound detection overlay. With all that said, I never considered if it would be possible to visually see the actual sound wave. I am now going to Google the hell out of it, lol
That P*rnhub frame got me in tears. Also amazing content like always! You’re such a blessing
I don't know if you will see this Benn, but I have been learning from you since like 2013, I was a kid and came across your music on youtube haha, just wanted to say thank you man! We appreciate you bro
Hey! We followed the same path of thought / information search ^^
After watching Steve Mould's video on motion amplification, I found some of the Github repositories you mentioned, and a few weeks later the Posy video!
But you followed the path way longer than me ^^
It's a bit useless as information but it made me laugh :)
Ever since I saw that Abe Davis TED talk in 2015 I"ve had an idea for using acoustic imaging to find anoamlies in a room, and even for finding, say, studs or other framing members behind drywall during a renovation or DIY so that the screw goes in the right place the first time. I never pursued it and thanks to this video I think it might be time to revisit the idea!
I'm wondering if the biometrics stuff could be used to detect AI generated fake content.
Had the same idea. Would work till AI is being thought how to fake that as well.
Ive screenshotted your message and forwarded it (together with this video) to a former colleague. She worked at the company that makes acoustic imagers that I work at and recently started a job at a company that does deepfake detections. If I receive any news then Ill follow up.
Interesting idea, but the way you usually build AI video generator is to train it on real videos so if this pattern is present in them there is a decent chance the AI picks it up.
@@petergerdes1094 i presume it would be fairly inconsistent throughout the video though
These kind of things are already used by identity verification software as liveliness proof required by financial regulators. Meaning, when you open an online account for banking or stock trading, it is very likely that a software is used to check on the image of your webcam.
that posy video took my glitches to the next level. I made apurple carpet out of static by iterating his difference layers in premier. it is so so cool
We love Posy!
I'm not an audiophile; but, You, sir (and Posey), with every video, give me a greater respect for my sense of hearing. Thank you.
Hey it's finally done, congrats! So just how long IS this one? 😅
Over a month due to all the trial and error with various ideas and devices. But at least I will no longer be kept up at night thinking about it. 😂
I respect you highly. It wouldn't be as high if I didn't think you were a good person. Ty for everything.
0:15 casually green screening yourself in front of a quantum computer was pretty funny
Regarding measuring the energy content of sound:
The sound intensity (sound power per area) of a 1kHz tone at 90dB SPL is a hair short of 1mW/m^2. For reference, sunlight is about 1 million times stronger (1kW/m^2).
Maybe you could stick a really precise thermometer probe into some sound-absorbing foam? You'd have to pulse the sound and correlate it to changes in temperature.
This is extremely closely related to my PhD and I can tell you with sound that quiet it is not possible with current technology. The cooling capacity of the material means any slight increase in temperature would be almost instantly transported away via convection, conduction or emission. With higher pressure sound it is certainly possible.
Love the Posy plug
The first part was the best part. More technical is better ! :)
This is crazy. Great job
That was awesome! I've always wondered about making a real life 'sonar' visualization... might have to give it a shot after seeing this!
You are missing those links you refer to in the description of the video... Thanks.
Ben, you freaking rule. Thanks for being Ben. I wouldn't want you to be anyone else. Well, maybe a happier version of Ben. Everyone deserves to be a little happier.
you sure it is the sound generating the heat and not the speaker coils heating up from having electricity pumped through them at an ascending frequency?
I have just found your channel so this is my 2nd video of yours , ( the first was how animals perceive sound ) and i just wanted to say thanks . your videos are amazing and the way you explain things are perfect for a layman . THANK YOU FOR EDUCATING ME AND GIVING ME SOME GREAT STARTING POINTS FOR FURTHER EXPANDING MY UNDERSTANDING AND KNOWLEDGE . much love from Scotland
that shirt goes crazy
Always a good time when Benn Jordan nerds out a bit. Great video.
13:08 This is really good to learn to animate fluids, please release more of those experiments
As a programmer & producer, this channel is an absolute treat.
Thanks for saving me 30k on a camera I didn't know I needed 😂
Finally I'll be able to see in real time the energy that's coming out of my hand and the beautiful geometric energy forms I'm projecting 😊 Thank you Benn, you're awesome.
Being a nerd is one of the best things in life
What does "being a nerd" actually mean? That's not a rethorical question. Everytime I hear someone say or write it, it feels like "being a nerd" is somehow self-deprecating and a not-so-humble brag at the same time. But what does it *mean* ? Wearing glasses and short-sleeved button-down-shirts with a bow-tie and suspenders and carrying around a pocket calculator?
@@michaelgraflmusic to me it's being ultra curious about the inner working in life and how they interconnect and influence each other in a spiraling way
Ok nerd
A nerd is someone who spends their time and energy to study something in great detail or experiment for no reason other than the enjoyment of it. Usually having to do with science. For a simple example, I like to think being a nerd plays out just like the Revenge of the Nerds movies. Being a nerd is originally a put down until the nerds use their smarts to become the dominant group and get the babes. Maybe not exactly like that, but being a nerd is a compliment usually because they are so desperately needed in our advanced world.
@@lucasgraeff5391this you nailed it. The one thing I feel all different types of nerds have in common is a relentless curiosity
You should consider making a laser microphone. You probably already know what they are, but in case you don't: they shine a laser on a target and detect the reflected light. Vibrations in the target will ever so slightly alter the phase of the reflected light (much like how a laser interferometer works). By observing the interference pattern made when the source beam and the reflection are combined (and with a lot of math) you can extract frequency information and convert it into audio.
I'll admit this is probably a lot more complex than I think it is. But I've spent hours playing with lasers and I've noticed that by reflecting a laser off of a surface that scatters the light, the speckle pattern gets magnified and projected onto the ceiling (or wall, depending on how you aim it). This projected and magnified speckle is extremely sensitive to changes in either the laser's position or the position of the surface off of which it is reflecting. So much so that I could see my heartbeat even when I was sitting in my chair on the opposite side of the room from where the laser and the reflecting surface was. Using a shorter wavelength (i.e. a 532nm green laser instead of a red laser) increases the "resolution" of the speckle, and thus the sensitivity.
You could probably utilize the code used for motion extraction to do this.
This is so cool! I was born dear on one side and cannot localize sounds. Do you know how cool this would be as an AR assistive technology that could help someone like me localize a sound? It would be amazing!
You beautiful nerds out there need to get on this so homie here can localize sound!
This should be like 5 videos, each gadget is so interesting!
Back in I think, the 1970's, there was an article in Popular Science Magazine about how GM was using sound and photography to find hairline cracks in Corvette engine parts. When I saw this title, I was hoping it would be about that type of tech, as I was recently trying to remember how they did that again because I forgot and am interested in casting metal items and it would be great for analyzing there quality.
The creativity in this video is definitely refreshing for RUclips
Your Low cost Schlieren Photography trick is Absolutely Brilliant!!
Just a quick shout to say thanks Benn for continuing to not only make excellent videos covering a variety of music related subjects but just for plainly still being here, period.
Too many YTbr's give there all, in one explosive season and then they disappear. Not you. Not you. I love your tenacity and grit. Of course I love your music that goes without saying but it's really your honest approach to transparency that I admire. Keep it up man. Cheers from Nova Scotia Canada .
One can only imagine the toys like this developed by the department of defense.
You are amazing man 400k and all your accomplishments are incredible but I think this man is very underrated in the yt department
You and I hyperfixate on the same damn things
This is absolutely amazing, the amount of ideas you covered in this video alone deserves a whole series of videos. Thank you for all your hard work.
I'm glad I found Henry Rollins' youtube channel.
Mace Griffin Bounty Hunter YT channel....FTFY :)
Brother we have a new subscriber today! I have learned more about three topics that are important in my life and you're one video today and I have in 20 something years of sound engineering and musical producing as far as recording songs and making beats and all that jazz digitally and analog and I'm very impressed and I'm very appreciative thank you so much!
Thank you! This is like the old YT video "The Visual Microphone: Passive Recovery of Sound from Video" where they record on a full hd canon dslr and extract audio source from the images only. Mindblowing ruclips.net/video/FKXOucXB4a8/видео.html
Great video, love the usage of rolling shutter to achieve higher virtual framerate!
This could've easily been a four part video series! You always cover such interesting topics, thank you for tickling my brain in regions I didn't even know were ticklish
11:19 Because you put it there!
Hello fellow creep
If this is the amazing quality we get when you miss a video... I look forward to your next missed video! Brilliant work.
14:09 (HEADPHONE ‼️ WARNING)
Ouch. 😣
This is by far the most interesting and informative video I've seen in quite a while. Such a compelling rabbit-hole to explore! Many Thanks for making this!
3:43 My frustration with Python exactly. I find a Python script has a shelf life of like 1 year, and if I revisit it after that I'm in for a whole mess of digging into why it no longer works and finding old versions of x, y, z....
Heh. We never figured out reproducible builds in the python packaging world. It never used to matter either, because breaking backward compatibility was considered an enormous faux pas. I've had some success with nix for repeatable python, but there is definitely a learning curve.
I dipped my toes in Python, and got frustrated by this. Well, this, and the fact that most documentation is written as a flowery tribute to how great Python is, and how easy it is to read. Of course it's easy to read. Nobody ever writes error-checking in Python, they just let the whole thing collapse with a stack trace. Turns out, C and ASM are easy to read when you are unreasonably optimistic, too!
TBH, it was never going to work out. I just can't with languages that use line endings as syntax. Gimme my { braces and semicolons; }
This is where you want to be using venv and a package manager that supports semantic versioning, but yeah it's not trivial. NodeJS is a nice example of baking this in by default with NPM.
One of the physics labs I did in college was optics, and one of the cooler demos we set up was a laser interferometer. Similar in principle to what you're describing here. We were able to tune it to visualize thermal layers and flow around e.g. a small airfoil (finger sized).
transcribethis AI fixes this (AI Transcriptions/ Audio to Text). "Make A Legit Sound Camera"
Man I like you dude. You're straight to the point and just real. And I like the knowledge and just way cool. Keep going man
I can imagine some very Orwellian application for this kind of tech
This video was so fucking worth the wait
You have a new sub. My wife is HoH and any technology that can bring a localized visualization to sound is extremely fascinating. Thank you.
nice practical brainstorming!
Since some time I have searched for people who managed to invent ways to locate mosquitos, and it seems your video is a very nice source of inspiration. And no, it is not really needed to shoot down mosquitos with a laser, actually to make mosquito trap it is enough to know when they pass by to either trap them mechanically by a net or one can use either vacuum , pressurized air or combination of to remove the bug.
AFAIK the software for an audio camera thing using multiple microphones is just a 2D Fourier decomposition. It's the same problem the long baseline radio telescopes face. The optimum sensor layout isn't actually a square array, it's putting them at weird locations, as far apart as possible, that are not multiple wavelengths of each other because you get more resolution. When you don't have a fully filled array, you get diffraction pattern effects but most sounds are multispectral so the bits you don't want won't usually line up in the frequency domain.
I have no use for any of this, but your videos are so interesting and well thought out, I love learning all these seemingly magical audio things. Thanks!!
Visual reflection detection of audio is super useful for treating studio and listening rooms. Being able to see the reflections and treat them where necessary is a super time saver with no real trial and error required. And it is recommended that you rent a very high end camera for this because you really only have to do it a couple of times to map out the rooms, provided you're not the kind of person who moves things around frequently. It's hard surface reflections that cause the most trouble. Of course, the "golden ear" club will try and tell you they can do it with just their ears so they can seem like they have ultimate hearing but, the eyes are better than the ears in most cases, especially this one.
I always enjoy the work you put on here. Brilliant stuff! That bird tracking section could be incredibly useful for species preservation
i love this channel because it combines my obsession with music and nerd shit at one place
Great video, you've definitely become one of my favorite channels over the past year. Just one question: how do you know the heat wasn't generated through the speaker itself rather than the audio waves coming out of it?
I already spot the UMA-16 v2 USB mic array from miniDsp. It‘s a nice one and I‘d love to have one that even has more channels. Looking forward to your implementation…
I think Steve Mould, or someone in that same sphere, made a video about acoustic imaging a few years ago. The other day I was thinking about it again and marveling about how since we typically record sound at around 44 khz, when we use acoustic imaging we are essentially capturing the data at 44,000 fps, meaning it is inherently a high speed camera.
Very cool video, i have a feeling the thermal imaging bit was due to friction between the film on the speakers and the plastic underneath
This concept is actually a really great way for a human presence sensor. I've seen other projects that map out bird calls to the specfic species, but to measure a heart beat could even be useful in a non invasive medical setting.