My understanding is that throat singing and overtone singing are different. Overtone singing does indeed just use the overtones, using techniques to amplify them. But throat singing will vibrate the so-called "false vocal cords" at the same time as the regular ones. This creates a growling sound below the normal voice, which throat singing maintains as a pitch. It isn't being produced from overtones. They sound rather different. In overtone singing, the fundamental is the lowest pitch, with very high notes on top. In throat singing, there's this very low tone, with a normal voice or falsetto on top of it. In theory you could combine the two to get three notes, but I've not heard much of this.
You're right; I conflated them as one thing and should have made a distinction; I probably made the mistake because most "throat singing" traditions I am familiar with also make ample use of overtones. Wikipedia says: "Throat singing and overtone singing are certainly not synonyms, contrary to what is inaccurately indicated by many dictionaries (e.g. , in the definition by Britannica) but, in some cases, both aspects may be clearly present, such as in the khargyraa technique from Tuva, with a very deep, tense voice, and rich overtone enhancements and embellishments."
That sawtooth-white noise effect is like the audio equivalent of the negative afterimage optical illusion that colourises a greyscale image. Very cool.
Pretty sure the Lucier effect also has to do with the speaker quality and characteristics. To do it perfectly you'll need something that can record and play back sounds in such a way that in a room with no characteristics, the live audio and the recorded audio will sound exactly the same
The Lucier Effect is the audio equivalent of when a teacher keeps photocopying the copies of photocopied worksheets. It eventually looks like an indistinguishable mess.
@@Boukouvalas1979 a similar event occurs when running a dialogue thru Google translate a bunch of times and then back to the original language. This was done for Paper Mario by Fatguy703 and it's quite hilarious. I recommend familiarizing yourself with the original before looking at the messy translation.
11:15 I not only heard the white noise sounding a little off until the frequencies all came back into play, but when I listened the first time, I heard what I would describe as a strange “warble” effect for the first second or so, as soon as the white noise was switched in.
Same! I thought it was a byproduct of the TH-cam compression of the audio... sometimes like mp3 low bitrates sound like that... but nope. that's all in my brain. weird!
Yes! The result of AM F1(F2) is F1, F2, F1-2 and F1+F2. I've had this problem on a tube amplifier. The thing was spitting some notes with "a low octave effect", as the customer said. Took me some time to figure that the filtering caps were just a little leaky and the mix of 60Hz mains along with the lower register of the guitar was generating intermodulation. But this is the first time I see an acoustic analogue
The white noise sounded like rain falling after the sawtooth sound. I got the compression effect but I started to hear distinct drops of rain hitting the ground. Nice..
at first for me it was like a very intense static, it sounded like an electric shock almost, then it changed some. I would go back and do it again to check but it was not a pleasant experience so I would rather not lol.
I personally don't hear it, but yeah, totally possible that you (or others) do... the spectrum shown on screen is an honest representation of the tones being played, though, so anything you hear is the psychoacoustic effect as opposed to the intermodulation of the reeds.
@@CaseyConnor I hate it so much that I itched a lot I also heard baby flies when the white noise cane on after the first effect and I have a fear of flies
The triangle wave and white noise illusion is the exact same thing that happens when you’ve had an electric razor sound in your ear for some time and it suddenly stops causing everything to sound “warbly” like that. I’ve always loved hearing that effect and it’s cool to see it talked about here
Thank you for sharing all your information. I am a musician and I have tinnitus. The 6K tone that you produced was about the same frequency as the ringing in my right ear. I have other frequencies also and I have analyze them in the past with Sony Sound Forge in my studio I don't remember the exact frequencies but after hearing your 6K tone that is definitely one of them
That mono mic moving between a stereo source shows off the opposite effect you get when you try to mic up a single audio source with two mics that are setup in one of the stereo pair configurations. That’s why if you’re going to do it, you need to space it out so that the minimum distance that’s at least 3 times the distance between the mics as the distance between the first mic and the source. The phasing in stereo causes comb filtering but if it’s summed to mono it sounds even worse. The way they recorded Bowie singing on his song Heroes, they wanted to come up with a way to make the full reverb of this massive room they recorded in, while having it seem more intimate for the verses. So, they had to put several mics spaced out following that rule, and for each much they used a separate noise gate setting so that it wouldn’t trigger the effect until it got much louder-so as he sang louder it would reverberate more and it just sounded badass. I think they ended up with 2-3 mics in that chain and it sounded. Listen to the way the reverb on that track swells up and it was before automation was even close to being a think.
I believe the extra harmonics from the melodica have two different physical causes: 1st cause: Reeds of adjacent tones (f1, f2) are also mounted next to each other on the frame. They therefore act as coupled oscillators. This boils down to a frequency-modulation (FM) of each oscillator. The modulating frequency is the difference fd = f2 - f1. A sine at freq. fs frequency-modulated by fm produces harmonics fs+n*fm and fs-n*fm (n = 1, 2, 3, ... ). These make up the extra harmonics near or above f1, f2. n=1 is the strongest, so I'll focus on that one: f1-fd and f2+fd are easy to spot on your spectrogram right next to f1 and f2. f1+fd and f2-fd, have no extra peaks because they fall exactly on f2 and f1 respectively. Since the melodica has strong overtones, 2*f1 and 2*f2 are very dominant. These are also modulated by fd, thus producing the four harmonics 2*f1-fd, 2*f1+fd, 2*f2-fd, 2*f2+fd. But these form only 3 extra distinct peaks (well visible) of which the middle is from the two components 2*f1+fd, 2*f2-fd. Both represent actually the same frequency f1+f2 (2295 Hz), which is therefore also extra strong. 2nd cause: Since the two reeds oscillate at similar frequencies, they swing alternately in-phase and anti-phase. Since they share the same air-tube (which includes the mouth region), these two swing states produce different kinds of standing waves in the air-tube, each producing a slightly different amount of composite air resistance. This produces a pressure wave in the tube (mouth for harmonica) at fd = f2 - f1 which is the low frequency peak one can see in your spectrogram (200 Hz). I hope that helps. Thank you for your excellently made and very interesting series on psychoacoustics. 🙂
That helps, thanks! So are you saying that the deep tone is most likely from the shared air manifold effect, where the other products are from the coupled-oscillator modulation through the frame? Or that both modes could cause both results? Generally, I can see that if modulation is happening, these extra tones would be produced... what I'm having a harder time seeing is _why_ modulation is occurring. Take for example the difference tone, which for some reason seems like it should be easier for me to grasp: I understand beats and how they happen, but it's hard for me to see how a beat happening at a certain frequency becomes an "actual frequency", that is, what the physics of this distortion in the system are. As you know, if you combine two tones in the computer, there are beats but no difference tone, because the difference tone isn't a "tone" per se but the measured frequency of regular amplitude variations. The best I can do is imagine that in the system of air around the reeds, the overall increase in amplitude of the variations in pressure during the "loud part" of a beat vs. the "quiet part" of a beat would cause some kind of overall increase of pressure (i.e. a "real tone") due to nonlinear vagaries of fluid dynamics, but it's those nonlinear vagaries that I don't have a grasp on. (I suppose I can also use the same method to imagine nonlinear vagaries leading to the FM due to coupled oscillation through the instrument.) I can imagine that a more distant part of the air manifold, perhaps before the reeds (or in the mouth, as you said) would react gradually to the overall changes in pressure from the beats and once oscillating due to those changes would thus induce a proper tone... maybe that's the best I'll do on this until I get a degree in fluid dynamics. :-)
@@CaseyConnor Yes, I mean the first kind: The cause 1 produces the upper modes but not the single low one. Cause 2 produces the low frequency mode. I should add that I'm only pretty sure about cause 1, less so about 2. The flow dynamics is hard to model (too hard for me, at least) and I can only say with confidence that I'd be very surprised when the two states (in-phase, anti-phase) which produce significantly different flow and pressure characteristics would result in exactly the same cumulative air resistance. I agree that the difference tone is easier to grasp when assuming intermodulation caused by non-linear behavior of the material involved. This was also my first thought and it might indeed be happening to some degree. I only feel the effect would than be much smaller than observed.
I brought the last part of this video into FL Studio and played the white noise through the master with no effects. All I heard was white noise until I turned the stereo knob on the master channel to 100% mono; then I heard the melody clearly. But fun fact: the frequency response that showed on the spectrogram showed the melody playing the entire time. Today was when I discovered that the spectrogram on FL Studio's interface only shows the mono signal. Another fun fact is that their built-in Parametric EQ (Fruity Parametric EQ 2) also only shows the mono signal, but affects both mono and stereo signals. It was only when I opened up a patcher with a Mid/Side EQ that the interface showed the stereo signal as white noise. Armed with this knowledge, I may be able to better treat a muddy signal, since I now know that a Mid/Side EQ is needed to see the stereo signal coming through, rather than just a single EQ. _Edit: I meant spectrogram, not oscilloscope._
Hi. Lately, I like to 'reset' my ears by listen to 'pink noise' for a minute, with IEMs. I found tone of the sound is getting better. I also use it to adjust or selecting IEM's tips to get better sound. Thank you for very informative videos.
The Harmonic Choir was well known for doing the multiple tonal techniques as well. I remember my parents playing some of their stuff back in the early 1980’s. It was so strange to me at the time but very interesting as well! And I very much appreciate your videos, so fascinating! 👍
The melodica subharmonic example reminds of me of how distortion behaves, namely that two tones individually distorted and then combined will yield a different result than when the two tones are combined first and then distorted (even when done so digitally). However, I doubt that its a related effect.
you are correct. The harmonia produces the beat frequency because some part of it is non-linear It is similar to how an AM radio is demodulated simply by passing the carrier frequency thru a diode - the half wave rectified carrier is nonlinear and the modulating audio is generated
The proper interpretation of the lower phantom tone is stritcly correlated with distorsion. A single tone with distorsion generate upper harmonics, but if you add a second tone and a tiny beat of amplitude modulation you will get an intermodulation that connected with real instrument natural distorsion generate the phantom lower tone. Matematically speaking the distorsion creates such a product between the two tones generating the math correct frequency distribuition. Thats the same reason why valve amplifier are so much worm and preferred by a lot of musician instead of modern "perfect Linear" undestorted amplifier.
The Sawtooth Habituation Effect reminded me of a similar experience I had years ago, listening to Charlemagne Palestine’s organ work Schlingen-Blängen. The sound of it has a flanger/chorus/vibrato like quality (probably naturally achieved). After listening to the 70 minute work, my ears were configured by its sound character so that when the work ended, when I spoke to my mother (I was lucky to do that - I could have stayed silent) I heard my voice having the same flanger/chorus quality in it. It was marvelous! And the effect lasted for at least 20 seconds. I listened to the last five or so minutes of the work again today but I did not achieve the same effect. It probably needs more time in order to be achieved.
The delayed speech trick is the technique used in speech jambers. You are speaking but then you hear yourself and your brain thinks it's someone else, bc of how the skull softens your own voice which isn't present in the playback, and as a result, you stop speaking to hear what the "other person", the interrupting voice, has to say but because it's your own voice, it becomes silent and you continue just to have this pattern repeated. This will either frustrate a person or they will make a mental note to block it out. This is why echo on a phone call is so irritating and troublesome and why speakerphone should be abolished.
With that last effect you covered, I went to the other video to hear the effect there, and I noticed that if I listened closely with stereo on I could still hear the melody, mind you that I had mono turned off. The reason why this happened is probably the fact that the melody plays at the exact same volume with stereo as when mono is on, the only difference is on stereo the melody is drowned out by all the other sounds, if you use a sound like white noise which is basically just every frequency you can hear played at once, you usually won’t run into many issues. But if you use a sound where some frequencies are absent, then if the melody hits one of those frequencies, it won’t be covered by it, which can cause the melody to “leak through”, another cause is if the melody is too loud (kind of like how the Huggins pitch doesn’t work if your volume is too high), which would mean you would want to stay away from low pitched percussion like sounds like a base drum, unfortunately, the first note in the hidden melody in the Doom Eternal theme is a base drum, making it easier to pick out, and kind of like the cocktail party effect (covered in part 3/5), if you know what the melody sounds like, chances are you will be able to pick out the rest of it too. It happens that both were the case there.
ferry interesting serie! Thank you! On the "two melodica subject" This effect is also ferry pressent when two clarinet players play a C and E in the high register. It is a physic effect also essential used in electromagnetic waves, radio waves. It is an essential methode to discriminate (Sorry, I don't know the right English word) the audio content pressent in radio waves. (mixer)
I'm wearing headphones (Sennheiser HD 598SR), and for some reason, when I heard the white noise coming from just the sides with no mid, it felt like the noise was either coming from behind me or inside my head. Incredibly freaky.
18:50 getting flashbacks of manually spinning my grandparents record player. It wasn't plugged in, so I would play this game where I tried to spin the record steadily enough to make it sound normal and enjoy the tunes.
11:57 : i absolutely love the effect! 16:49 : i love this, too! 01:30 : I love the idea but i think the cause is different and very simple. It's not the room but the methodology. Every iteration with its playout and re-recording introduces a EQ manipulation, the peaks and notches in the freqnecy spectrum cumulate and after some cycles you only have some remaining spikes that sound what they are. This is the strange "ringing".
Sawtooth Habituation Effect is so pronounced it's awesome! As for the Moving in a Stereo Field effect - two things come from memory: 1) A concert experience of a quadraphonic recordings by Phill Niblock - somewhat steady but harmonically rich drones come alive by turning head or moving ever so slightly. 2) An album "Matrix" by Ryoji Ikeda, which is built around this effect, and suggests listener to create own experience by movement whithin a sounding space. ++ Contrary to Matrix, which is intended to be listened from speakers, he also has a "+ / - " album, containing among others three "headphonics" tracks, which explore headphones-related phenomenas. Thanks for the series!
The effect you spoke of in the melodica intermodulation segment of the video is actually the foundation behind the science of binaural beats where two separate frequencies are played simultaneously (one in each ear). Even though theyve been isolated from each other our brain will perceive the difference between them as a third tone playing but if u take out one headphone out at a time u can hear that theres are actually only the 2
This is not a 100% accurate statement as the two frequencies are in fact, being "mixed" (Heterodyned) as mentioned and produce the difference and the sum just as mentioned. This is a well known phenomenon and has been used in radios the world over since the beginning. Ask any ham radio operator and they will explain the effect and how it is used to produce the Intermediate Frequency stage of any radio. The frequencies are higher in radio, but the results are the same. Radio's use oscillators to generate the 2nd frequency to mix with the radio signal's frequency to get the lower difference frequency.
In simple mathematical terms, when you multiply 2 waves of different frequencies together, the result is 2 new frequencies, one is the sum of the inputs and the 2nd is the difference between them. cos(A)*cos(B) = cos(A+B)/2 + cos(A-B)/2. There must be some acoustic multiplication effect going on, which of course doesn't appear when you simply sum the 2 frequencies in software.
Yeah -- it's exactly that process I'm curious about... how the physics of waves bouncing around generates the sum and difference tones... and whether the many other distortion products are just sum/difference tones of harmonics or can also come from the fundamentals, etc. Mysterious. :-)
@@CaseyConnor I’ve got a feeling that it’s just frequencies moving through the air. If you were somehow able to play a high C, then play a note 4Hz above or below it, and really increase the sensitivity of the microphone, you would probably hear a 4 Hz tone. I remember when I was little hearing two people scream and hearing a low bass tone that was changing, going up and down in pitch. The tone was ranging from 320 to 400 Hz, and the people screaming were screaming at a frequency between 2800 and3K. I think it’s just that, but the melodica is probably adding frequencies due to it being reeds or something, I don’t know much about melodicas, but I do play piano and drums. These are just my thoughts about this. I also have perfect pitch :-) :-) ;-) :-) ;-)
@@levi_gobin Yeah my math on this is shakey at best, but my understanding is that the sum/difference tones only show up when some kind of nonlinear distortion is present in the system. Apparently the intermodulation "Tartini" tones musicians hear come from nonlinearities in our own ears, and the melodica one comes from ... something :-) ... but I don't think they are otherwise present. At any rate, the mystery continues!
i accidentally found the sawtooth habituation effect after i had done some work on a song. i noticed when i started listening to other parts of the song after listening to a particular saw-y part of the song (i think i had the instruments soloed) everything i heard from then on a few minutes afterwards sounded the same as the example you gave after 40 or so seconds of the constant sawtooth wave being played. neat.
you describing what happens in the affect reminds me of an interesting phenomena than ive experienced phasing in and out of sleep in several different places. in some circumstances, it involved my screensaver, which is cotton candy pink text wobbling around, but if im half asleep, like actually half asleep, it appears to be a similarly shaded blue instead.
ok that sawtooth/white noise one actualy freaked me out, i genuinly thought you put in a shit quality white noise sample for no reason, but that was the illusion wtf lol
About the Doom Eternal one, Mick Gordon did it again on the track IV. Doom, this time with a recording of a woman whispering "Jesus loves you" backwards. Boards of Canada did it as well for an ARG in 2014, but the audio was spread amongst two Soundcloud tracks.
The overtones are created because the phase difference between the two notes you're playing produces a soundwave vibration that also affects the physical features of the instrument that made the sound; for example, A harmonica playing the two notes creates a vibration that also causes the reed of that note to vibrate. If it doesn't have a reed of that note, then even an integer value of the difference ex: 1/2, 1/3, etc. or 2x and 3x wavelength harmonic. If that exists in the instrument, it will set up a resonance that will produce the tone, including the cavity of a device if it is physically shaped to do it. So the sources of the lower tone depend on what kind of features exist in the instrument, which can sympathetically vibrate to one or more harmonics available based on the frequency of difference between the two notes played. A guitar is well suited for this. And it also explains why a guitar that is well played and often will through compaction and rarefaction of the notes on the guitar will cause the wood to reconfigure the vibrational nodes in a guitar making the guitar a better-sounding guitar over time; thus, the price goes up. 🙂 No silliness this time; I am a retired Metrologist. No, not the weather. Measurement expert! 🙂
6:55 I'm not an expert by any means, but I would hypothesize that the difference between the two tones will create peaks in the overall audio that is heard by the listener. The peaks are periodic and, as such, produce a tone with a frequency. The frequency will be the rate of the peaks caused by the two audios. Does this make sense? It's like the rhythm = pitch scenario, where the reoccurring peaks and troughs of two audios played together will establish a reoccurring beat that, when sped up, produces an audible pitch.
Yeah I'm right there with you in an intuitive sense -- but the thing is that when you play two tones simultaneously, the beat that happens between them is not itself a tone -- e.g. if you mix 100 and 150 Hz you have a 50Hz beat frequency, but that beating is not the same as a 50Hz tone, because there is no true 50Hz frequency component present. Consider the example I show in this video where I play the two tones separately and then combine them in the computer (as described with text on screen) -- the difference tone, and all the other distortion tones, are not present (it's not in the spectrum shown, and it is also not audible to your ears as it was before.) Said another way, if you have two melodicas playing the two different tones near each other, this effect doesn't happen -- it requires the tones to be produced in the same instrument. The rhythm -> pitch idea you describe is different, because playing a rhythmic sound quickly actually introduces frequency content at that frequency. Said another way, if you have a 100 and 110 Hz tone combination, you hear a 10Hz beat. But if you gradually increase these two frequencies at the same proportion (eventually getting to 200 and 220, 400 and 440, 1000 and 1100, etc) the beating effect does not become a new tone once it climbs high enough, instead it becomes a sense of "roughness" we hear and which contributes to the sense of consonance or dissonance. So AFAICT there needs to be a physical explanation for why the difference tone would manifest in the melodica, and I think intermodulation distortion is that explanation -- my best guess (as described in the vid) is that the beating has an effect on the dynamics of the instrument causing a real tone at the difference frequency to manifest... sorta how we would intuit a beating frequency to work in the first place, ironically.
The delayed speech trick is something encounter frequently using radios to communicate at work. When your voice comes out of someone else's radio Close by so you can hear what you're saying well, half a second later, it becomes almost impossible to talk coherently! 15 years doing it, and I still struggle with it!
14:20 That is an interesting video effect. Your hands fade in, but you somehow completely matched your exact facial expression, head tilt, and head location. I cannot see the transition.
10:15 Plausible explanation for the unexplained tone. Is the deep tone in question different when the demonstration is preformed by a drastically different sized person? Eg you appear avg size hight build( average ≈ 5'8" 160#) are getting the shown tones were some one 6'8" should get a slightly higher result, 4'8" slightly lower. Build Characteristic would also play in As bigger longer bones would Have a higher sympathetic resonance frequency and heavier set people would have more sound insulation on their bones
for the white noise part, i listened with my eyes closed. it sounded like the sawtooth was getting incrementally louder and then as the white noice kicked in the was a weird screaming/scraping electronic (im not 100% sure how to describe it) sound for a second before the white noise played as normal. I'm not sure if that was the intended effect but it scared the hell out of me 😂
Melodica Intermodulation is exactly as you describe: heterodyning. This same principle is exploited in radio communications to shift high frequencies down into whats known as a "base band range" by mixing with an intermediate frequency in order to make demodulation of RF signals easier. The two tones here are mixed together (multiplied in a DSP context) to provide your negative frequencies and positive frequencies.
Thanks! So are the other distortion products (meaning, besides the sum/difference tone of the two fundamentals) just heterodyne products of the various overtones? And do you have any sort of intuitive grasp on how the physical phenomenon itself happens? Meaning, what is it about the physics that effectively causes the two signals to be "multiplied" in the melodica/harmonica?
HEY I like your videos.... on the experiment of melodica, I think is an acoustic thing. maybe is resonance because 2 holes are resonating, so the area where the wind is passing is added. sorry about my English.
a lot of feedback on the lucier effect, i wonder if its due to how drivers in themselves work as well as some other engineering voodoo. possibly a relation between how microphones typically pick up sound and how speakers produce sound
For sure: the entire chain colors the response of the overall system: the loudspeakers, the room itself, the placement of the mic, the mic itself, and (to a much, much lesser degree) the audio interface.
@@CaseyConnor im gonna have to figure out a way to do this, since i dont have a speaker system and im thinking of trying it for a noise album. ive experienced similar effects from the good ol days of recording from crappy desktop speakers to my 3ds as a kid, so im wondering if theres a similar kind of interference. maybe especially tinny speakers/microphone mixed with a muddy counterpart could produce an interesting effect?
@@sunburnedshirts3724 Worth a shot! You'll probably want to adjust the gain of the mic as you go along (otherwise it can "run away" and start clipping the input, or fade away into nothing -- the trick is to keep the level near the balance point.)
I noticed another one recently, normally sawtooth waves sound very piercing and such, but I discovered that if you listen to a square wave first, then switch to a sawtooth wave of the same frequency (this has to be an instantaneous jump, there can’t be a gab between the two) the sawtooth wave sounds much less piercing than normal, I discovered this when watching a TH-camr recreating different sounds with sine waves, they did a square wave, then a sawtooth wave, but I noticed that the sawtooth wave didn’t sound like a sawtooth wave, so I rewound, skipped past the square wave, and it sounded more like a sawtooth wave
Yeah, nice observation! That sounds like a very similar effect -- I bet what's happening is this: the square wave doesn't have as many harmonics as the sawtooth wave; so while you listen to the square wave, the harmonics sort of "burn in" to your ears (just like in the demo in this video); then you switch to the sawtooth wave, which has twice as many harmonics as the square wave but which overlaps every other harmonic with the square wave, so for a few seconds the sawtooth wave doesn't sound as shrill because many of those harmonics are "burned in" from the square wave. I just tried it out and heard the same thing you describe. Thanks for pointing it out!
The sawtooth habituation effect jumpscared me somehow. When the white noise started, I heard what I can only describe as a "boing" effect and it was so sudden and unexpected that even watching the 2 arrows indicating it coming, I wasn't ready. Update: I tried it again and it made me physically feel sick. It's a cool effect but it's so unexpected the way it made me feel!
I had something similar. I was re-watching this video and when I saw that part coming I had to skip it. Last couple of times I’ve made that effect for me or other friends and I listen to it, I get a weird ass feeling! I can only describe it as that feeling of when you’re in the room all by yourself, no one in the house at all ... ... ... Silence, ... ... ... BOOOOOOOOOOOOOOOOOOOOOOOOOOOOM! The loudest noise possible happens. I really love affects like this. It’s interesting how just two sounds that when played by themselves are seemingly harmless until they are played one after the other. I wonder if there’s something similar with the reverse effect, 40 seconds of white noise or other colored noise, then switch to a sawtooth or square waves. I’ll have to try that out. It’s really interesting and I wonder what similar effects there are. I’ve seen so many “audio illusions” videos on TH-cam, but none of them are stuff like this. they just include stuff like the Shepard Tone, the Mc’Girk effect, and that’s pretty much it. I really wish there were more effects like this and others that you’ve covered in the series. Keep up the good work and can’t wait for part eight. :-)
I wonder how much the Sitting In A Room thing is more to do with resonances caused by gradual deterioration of the audio. In fact, I did this same thing as a kid, loads of times. I used to multitrack from one tape to another. I'd make audio sketches, doing one bit of dialogue and leaving a space for the next me to fill in, if you see what I mean, or sometimes I'd just try to create the sound of a crowd. Now obviously, the tape recorders I was using were hugely inferior to the kind of mics and speakers which would have been used in this recording, but in a way that emphasizes the point. You could hear the artifacts creeping in with each repetition. The audio would degrade, and on top of that, you'd get the original echo of the room, with the echo of the echo, and the echo of the echo of the echo, ETC ETC, until all the different me's sounded like they were in different places. The version of myself I'd just recorded sounded fine, but each one that went before sounded worse, and the original me sounded like I was in a different room, talking in to a tin can or something. So yes, I guess the room is adding to it, as the mic is picking up the resonances of the room all the time, but it's as much to do with the steady degradation of the audio, because no matter how good your equipment is, you're always going to get that over so many repetitions. Obviously this is why copies of music were made from the same master, and we don't copy the copy the copy the copy ETC. That's why those old tape delays sound so cool, because the frequencies are altering with each copy. If you make it long enough, the end result becomes unrecognizable from what you started off with. And in that instance, there is no room acoustics to affect the sound. So in this Sitting In A Room example, it's hard to tell how much of it is the room sound and how much is audio degradation.
If you connect a modern quality interface to itself via a cable (i.e. output connected directly to the input) you can go maybe 100 or more generations before you can notice anything, and even then the effect would be quite subtle. Even the coloration of the microphone in the demo in the video isn't noticeable. So what you are hearing in this effect is really just the effect of the room, with a small percentage coming from the response of the speakers. You could say it is "deterioration" of the sound, though... That's basically the effect of recording the room acoustics: hearing the echo of the echo and so forth, as you put it. The effect of the room is huge, and drowns out all other effects (unless you're in an anechoic chamber or something.) Even a first generation recording will see dramatic impacts on the sound from the room. I also used to do tape recorder overdubs as you describe, and in that case I would agree that the effects of the tape (hiss, loss of high frequencies, drift and warbling of pitch, etc.) and of the cheap mic I was using, we're huge. Good times. :-)
The propagation of soundwaves through the air is nonlinear. Soundwaves can be explained as variations in air pressure, but we also know that the speed of sound depends on air pressure. The nonlinearity becomes stronger as the sounds become stronger. When a full orchestra plays a very loud strike, the reverb will be slightly lower in pitch. Inside the melodica, frequencies may be multiplied by this nonlinearity, like in an electronic ring-modulator. But I don't think this effect is strong enough to explain the ghost notes in the melodica. The airflow around the vibrating tongues of the melodica must be far more nonlinear. :)
The melodica intermodulation thing is 100% intermodulation distortion. So if F1=1047 and F2= 1247, the difference tone and combination are what would be referred to in electrical engineering (particularly radio frequency engineering where signal mixing like this is very common) as "second order" intermodulation products and the other tones you see 200Hz above and below at 847 and 1447 would be "third order" intermodulation products, so 2*F1-F2=847 and 2*F2-F1=1447 (3341 and 3541 are also 3rd order products) and you can also see just a little bit of 1647 as well, which would be a fifth order intermodulation product of 3*F2-2*F1=1647, and so on, but usually the intermodulation products drop off in amplitude and get more numerous the higher in order you go.
I know, right... I was hoping he would have more trouble so the video would be more impressive. :-) He's a musician and has experience hearing himself in headphones while recording so ... Maybe not the smartest choice on my part.
I haven't, and yeah, it's an interesting topic... I have thus far assumed that there are lots of videos on the subject already, but maybe some day. :-)
The Sawtooth Habituation Effect didn't work at all like it seemed like you were describing... Instead, for a few seconds, the white noise was all wARbLy before fixing itself.
I tried out the 2nd Audio trick & it was hilarious 😂 I think it makes people sound like they've been drinking. It's a problem public speakers have to deal with when they have to speak in front of a large audience & they speak into a microphone that sends their voice thru loud speakers so everyone can hear. There is a slight delay between when you speak and when your voice thru the loud speakers sometimes cause an echo from the room. If you're used to public speaking in situations like this(like my father, who I tested along with several others who were not public speakers) While the effect worked on the other subjects, my father was able to speak normally. I thought that was neat. For some reason this effect helps people who have a stutter speak normally & is a tool used by speech therapist. 🤓
The sawtooth effect is very cool. Not only did I get the flange thing at the end, but as the sawtooth played, it was like it got louder and louder and louder, till it was almost unbearable. I had to remind myself that it wasn't getting louder. This is a cool effect because unlike the Yani Laurel thing, it works every single time. With the Yani Laurel thing, I canot hear Yani any more, unless I bring the pitch right down. It's like, once my brain knows the trick, the trick doesn't work any more. But that sawtooth thing, in addition to the cool white noise bit at the end, is quite scary. I've heard on science fiction shows where they try to make sounds which are supposed to have a really bad affect on the people in the play or whatever it is. But that simple sawtooth was more affective.
Totally agree on the "sawtooth getting louder" effect! I debated whether to include that as a separate effect in the video but in the end i decided not to. But yeah it's crazy how it can seem to increase so much. Also agreed on Yanny/Laurel... I heard Yanny like twice and it blew my mind. Every other of the 5000 times I've listened my brain just refuses to switch off of Laurel no matter how i try to fool it.
When I was briefly listening to the sawtooth effect earlier, not only did it seem like it was getting louder, it seemed like speakers were playing it from all angles all around you, and they were all getting closer and closer and closer to you, until,... 🤯
@@erpmo3326 I can't find any videos they have posted that are the same thing as the Mick Gordon effect, but if you have a specific video to check out I can try again.
The sawtooth and white noise effect has an optical twin where you stare on a dot in a false color picture. Then the picture is suddenly switched to black and white and for some moments you see a full color picture. th-cam.com/video/3P8q_dCU3RI/w-d-xo.html
Your melodica intemodulation effect is well known to the radio world, sideband radio is dependent on the phenomenon. Any modulation between two frequencies will produce a sum and difference component. In radio, these are called sidebands. In blues music, the effect is often utilized by celebrated electric guitarists, most noticeably when they play minor thirds in association with clipping distortion. Speaking of which, the harmonics produced by vacuum tube clipping and tape saturation are widely recognized as have pleasing effects, although discussing them inevitably leads down a rabbit hole. And dithering is a known psychoacoustic tool I do not fully understand, I don't remember you covering it.
Yes, sum and difference signals are produced by nonlinearity (mathematically multiplication of the signals or "modulation" rather than addition). It's interesting that modulation happens in the melodica, an earlier commenter explains the mechanics. Nonlinearity can occur in electronics accidentally or deliberately in a mixer. It seems that your ear or your brain can do some mixing as well. The air can become nonlinear at very high volume levels (dragster or jet engine levels of sound).
The Melodica Intermodulation is not specific to the melodica. Most organ sounds (straight, witouth vibrato) produce the same effect. I remember years ago a friend demonstrated this to me in his synthesizer with and organ sound. I just recreated the experiment using a software organ. So, it is either indeed an acoustic phenomenon, non specific to the melodica anatomy, or the synthesizer manufacturers go to extreme lenghts to recreate such phenomena (which I doubt it). Playing and holding G4-C5 produces and audible C4. G4-B4 produces a D4. G4-Bb4 produces a non-tempered note between D#4 and E4!
Thanks! It's also worth remembering that there is a psychoacoustic as well as physical component (though the psychoacoustic side has always been weak for me). Try recording the G4 by itself onto an audio track, and the C5 by itself onto a track, and then play those together in the DAW as audio tracks, and look at the result in a spectrum analyzer. There won't be any C4 (assuming there wasn't any C4 in the individual clips) but if C4 is audible, it will be due to the psychoacoustic effect. Then play G4/C5 together in the synth and look on the analyzer... I wouldn't be surprised if they do simulate the intermodulation in an organ synth (i's a pretty significant part of the sound, I think) but it'd be neat to find out for sure.
@@CaseyConnor When you play the notes on the melodica the subtone is clearly audible (I’m not sure “subtone” is the word to use, but you get what I mean). When you play them separately and then combine the audios it is scarcely audible. So we suspect a physical aspect of the instrument. But in the case of the organ samples there is no such aspect - only the perceived sound by our ear. I will do the experiment you suggest and let you know if I still hear the subtone. I suspect I will. When the friend first demonstrated the effect to me, the subtones were almost as loud as the real notes! In my recent recreation I needed to pump up the volume. But they were there without a shred of doubt. Now, the subtone I’m hearing is not the difference in Hz of the played notes. I play Bb4 (466.16 Hz) and G4 (392.00 Hz) and I get something in the vicinity of 320 Hz (between D#4 and E4). Where does this come from? I will also try other intervals and see what I hear.
@@Boukouvalas1979 Oh, interesting that it's not the usual difference tone. Yeah, give it a shot. My money is on the synth simulating it, but if it's just playing straight samples then maybe you just have a really effective demonstration of a psychoacoustic effect going.
@@CaseyConnor I did the experiment today, in the way you suggested. I still hear the lower "phantom" notes. Sure enough, there is nothing there in the spectrum analysis. So this is purely psychoacoustic. It is interesting though that the notes need certain sound characteristics to come out. No matter how fast/loud or for how long I strum piano keys on the same notes, I never listen to lower notes there. Sound has to be continuous for sure. Organ sound works well. Accordion samples also did the trick. The fact that I could hear the exact same by-product notes 20 years ago, in my early 20s, means that this is not an age "condition". Also, the friend who presented the effect to me was listening to the same phantom notes. Finaly, as I said, in that example, they were shockingly loud, almost as loud as the real notes!
@@Boukouvalas1979 Nice! What's the organ synth or samples that you were using? I've had the same experience of sensitivity to the instrument chosen - I think you mainly need a sustained tone with relatively rich harmonics, but that's just a guess. Perhaps that's why they are reported in violin as well.
At first I thought the lucier effect was going to be dumb but it sounded really cool by about iteration 16 and 22. by 47 its pretty much homogenized and boring. Cool trick.
My understanding is that throat singing and overtone singing are different. Overtone singing does indeed just use the overtones, using techniques to amplify them.
But throat singing will vibrate the so-called "false vocal cords" at the same time as the regular ones. This creates a growling sound below the normal voice, which throat singing maintains as a pitch. It isn't being produced from overtones.
They sound rather different. In overtone singing, the fundamental is the lowest pitch, with very high notes on top. In throat singing, there's this very low tone, with a normal voice or falsetto on top of it. In theory you could combine the two to get three notes, but I've not heard much of this.
You're right; I conflated them as one thing and should have made a distinction; I probably made the mistake because most "throat singing" traditions I am familiar with also make ample use of overtones. Wikipedia says: "Throat singing and overtone singing are certainly not synonyms, contrary to what is inaccurately indicated by many dictionaries (e.g. , in the definition by Britannica) but, in some cases, both aspects may be clearly present, such as in the khargyraa technique from Tuva, with a very deep, tense voice, and rich overtone enhancements and embellishments."
Bowdie biestefdeld
This is the one
Those are actually called subharmonic undertones, or "anomalous low frequencies" (ALF)
That sawtooth-white noise effect is like the audio equivalent of the negative afterimage optical illusion that colourises a greyscale image. Very cool.
Totally. I meant to draw that analogy in the video but forgot. :-)
Pretty sure the Lucier effect also has to do with the speaker quality and characteristics. To do it perfectly you'll need something that can record and play back sounds in such a way that in a room with no characteristics, the live audio and the recorded audio will sound exactly the same
The Lucier Effect is the audio equivalent of when a teacher keeps photocopying the copies of photocopied worksheets. It eventually looks like an indistinguishable mess.
I recall reading about an artist who actually did that
@@Boukouvalas1979 a similar event occurs when running a dialogue thru Google translate a bunch of times and then back to the original language. This was done for Paper Mario by Fatguy703 and it's quite hilarious. I recommend familiarizing yourself with the original before looking at the messy translation.
Or a wax hand
@@rustyshackleford5166zxmany?
The deep fried meme effect
11:15 I not only heard the white noise sounding a little off until the frequencies all came back into play, but when I listened the first time, I heard what I would describe as a strange “warble” effect for the first second or so, as soon as the white noise was switched in.
That trick also works with triangles, sines, squares, and trapezoids
Same! I thought it was a byproduct of the TH-cam compression of the audio... sometimes like mp3 low bitrates sound like that... but nope. that's all in my brain. weird!
Totally, felt like a slight flanger effect
at 8:00
in the RF world it also exists and its called intermodulation (the base note that appeared is intermodulation from 2nd order)
Yes!
The result of AM F1(F2) is F1, F2, F1-2 and F1+F2.
I've had this problem on a tube amplifier. The thing was spitting some notes with "a low octave effect", as the customer said.
Took me some time to figure that the filtering caps were just a little leaky and the mix of 60Hz mains along with the lower register of the guitar was generating intermodulation.
But this is the first time I see an acoustic analogue
The white noise sounded like rain falling after the sawtooth sound. I got the compression effect but I started to hear distinct drops of rain hitting the ground. Nice..
With me the sawtooth effect made something like a bouncy sound that goes up and down in pitch like a comical spring bounce sound
Wow, interesting.
Perfect description of it.
at first for me it was like a very intense static, it sounded like an electric shock almost, then it changed some. I would go back and do it again to check but it was not a pleasant experience so I would rather not lol.
This series is so amazing. It's like a funhouse for my ears.
This is amazing!! I learned so much from the entire series !! Thanks for making these videos
Is it just me or was there still an audible difference tone 8:40? And oh my goodness that white noise thingy gave me the chills.
I personally don't hear it, but yeah, totally possible that you (or others) do... the spectrum shown on screen is an honest representation of the tones being played, though, so anything you hear is the psychoacoustic effect as opposed to the intermodulation of the reeds.
@@CaseyConnor I hate it so much that I itched a lot I also heard baby flies when the white noise cane on after the first effect and I have a fear of flies
That room you're in looks so cozy! The plant and the piano (?) it sits on, looks so nice.
The triangle wave and white noise illusion is the exact same thing that happens when you’ve had an electric razor sound in your ear for some time and it suddenly stops causing everything to sound “warbly” like that. I’ve always loved hearing that effect and it’s cool to see it talked about here
Oh interesting i haven't noticed that one... I'll listen harder next time I'm shaving...
The pitch and rhythm circulatory sounds are so trippy. Gave me anxiety, lol
These are great videos! Interesting!
Thank you for sharing all your information. I am a musician and I have tinnitus. The 6K tone that you produced was about the same frequency as the ringing in my right ear. I have other frequencies also and I have analyze them in the past with Sony Sound Forge in my studio I don't remember the exact frequencies but after hearing your 6K tone that is definitely one of them
the sawtooth habituation effects sounded exactly like flanging to me
but flanging and low-bitrate mp3 are often very similar sounds
Ahhhh I’m so excited you came back to this!
My skullcandy earbuds mix to mono when one is in the charger so I could trigger the last effect by just putting one in the charger. 🤯🤯🤯
That mono mic moving between a stereo source shows off the opposite effect you get when you try to mic up a single audio source with two mics that are setup in one of the stereo pair configurations. That’s why if you’re going to do it, you need to space it out so that the minimum distance that’s at least 3 times the distance between the mics as the distance between the first mic and the source. The phasing in stereo causes comb filtering but if it’s summed to mono it sounds even worse.
The way they recorded Bowie singing on his song Heroes, they wanted to come up with a way to make the full reverb of this massive room they recorded in, while having it seem more intimate for the verses. So, they had to put several mics spaced out following that rule, and for each much they used a separate noise gate setting so that it wouldn’t trigger the effect until it got much louder-so as he sang louder it would reverberate more and it just sounded badass. I think they ended up with 2-3 mics in that chain and it sounded. Listen to the way the reverb on that track swells up and it was before automation was even close to being a think.
I believe the extra harmonics from the melodica have two different physical causes:
1st cause:
Reeds of adjacent tones (f1, f2) are also mounted next to each other on the frame. They therefore act as coupled oscillators. This boils down to a frequency-modulation (FM) of each oscillator. The modulating frequency is the difference fd = f2 - f1.
A sine at freq. fs frequency-modulated by fm produces harmonics fs+n*fm and fs-n*fm (n = 1, 2, 3, ... ). These make up the extra harmonics near or above f1, f2. n=1 is the strongest, so I'll focus on that one:
f1-fd and f2+fd are easy to spot on your spectrogram right next to f1 and f2.
f1+fd and f2-fd, have no extra peaks because they fall exactly on f2 and f1 respectively.
Since the melodica has strong overtones, 2*f1 and 2*f2 are very dominant. These are also modulated by fd, thus producing the four harmonics 2*f1-fd, 2*f1+fd, 2*f2-fd, 2*f2+fd. But these form only 3 extra distinct peaks (well visible) of which
the middle is from the two components 2*f1+fd, 2*f2-fd. Both represent actually the same frequency f1+f2 (2295 Hz), which is therefore also extra strong.
2nd cause:
Since the two reeds oscillate at similar frequencies, they swing alternately in-phase and anti-phase. Since they share the same air-tube (which includes the mouth region), these two swing states produce different kinds of standing waves in the air-tube, each producing a slightly different amount of composite air resistance. This produces a pressure wave in the tube (mouth for harmonica) at fd = f2 - f1 which is the low frequency peak one can see in your spectrogram (200 Hz).
I hope that helps.
Thank you for your excellently made and very interesting series on psychoacoustics. 🙂
That helps, thanks! So are you saying that the deep tone is most likely from the shared air manifold effect, where the other products are from the coupled-oscillator modulation through the frame? Or that both modes could cause both results?
Generally, I can see that if modulation is happening, these extra tones would be produced... what I'm having a harder time seeing is _why_ modulation is occurring. Take for example the difference tone, which for some reason seems like it should be easier for me to grasp: I understand beats and how they happen, but it's hard for me to see how a beat happening at a certain frequency becomes an "actual frequency", that is, what the physics of this distortion in the system are. As you know, if you combine two tones in the computer, there are beats but no difference tone, because the difference tone isn't a "tone" per se but the measured frequency of regular amplitude variations.
The best I can do is imagine that in the system of air around the reeds, the overall increase in amplitude of the variations in pressure during the "loud part" of a beat vs. the "quiet part" of a beat would cause some kind of overall increase of pressure (i.e. a "real tone") due to nonlinear vagaries of fluid dynamics, but it's those nonlinear vagaries that I don't have a grasp on. (I suppose I can also use the same method to imagine nonlinear vagaries leading to the FM due to coupled oscillation through the instrument.)
I can imagine that a more distant part of the air manifold, perhaps before the reeds (or in the mouth, as you said) would react gradually to the overall changes in pressure from the beats and once oscillating due to those changes would thus induce a proper tone... maybe that's the best I'll do on this until I get a degree in fluid dynamics. :-)
@@CaseyConnor Yes, I mean the first kind: The cause 1 produces the upper modes but not the single low one. Cause 2 produces the low frequency mode.
I should add that I'm only pretty sure about cause 1, less so about 2. The flow dynamics is hard to model (too hard for me, at least) and I can only say with confidence that I'd be very surprised when the two states (in-phase, anti-phase) which produce significantly different flow and pressure characteristics would result in exactly the same cumulative air resistance.
I agree that the difference tone is easier to grasp when assuming intermodulation caused by non-linear behavior of the material involved. This was also my first thought and it might indeed be happening to some degree. I only feel the effect would than be much smaller than observed.
@@johannessteffens8800 I appreciate the perspectives, thanks!
just found you and finished watching 5/5 and was looking at the new videos at the side and saw 6/5 released 10 minutes ago
probably the most interesting yt series ive seen
I brought the last part of this video into FL Studio and played the white noise through the master with no effects. All I heard was white noise until I turned the stereo knob on the master channel to 100% mono; then I heard the melody clearly. But fun fact: the frequency response that showed on the spectrogram showed the melody playing the entire time. Today was when I discovered that the spectrogram on FL Studio's interface only shows the mono signal.
Another fun fact is that their built-in Parametric EQ (Fruity Parametric EQ 2) also only shows the mono signal, but affects both mono and stereo signals.
It was only when I opened up a patcher with a Mid/Side EQ that the interface showed the stereo signal as white noise. Armed with this knowledge, I may be able to better treat a muddy signal, since I now know that a Mid/Side EQ is needed to see the stereo signal coming through, rather than just a single EQ.
_Edit: I meant spectrogram, not oscilloscope._
Hi.
Lately, I like to 'reset' my ears by listen to 'pink noise' for a minute, with IEMs. I found tone of the sound is getting better. I also use it to adjust or selecting IEM's tips to get better sound.
Thank you for very informative videos.
The Harmonic Choir was well known for doing the multiple tonal techniques as well. I remember my parents playing some of their stuff back in the early 1980’s. It was so strange to me at the time but very interesting as well! And I very much appreciate your videos, so fascinating! 👍
The melodica subharmonic example reminds of me of how distortion behaves, namely that two tones individually distorted and then combined will yield a different result than when the two tones are combined first and then distorted (even when done so digitally). However, I doubt that its a related effect.
you are correct. The harmonia produces the beat frequency because some part of it is non-linear
It is similar to how an AM radio is demodulated simply by passing the carrier frequency thru a diode - the half wave rectified carrier is nonlinear and the modulating audio is generated
The proper interpretation of the lower phantom tone is stritcly correlated with distorsion. A single tone with distorsion generate upper harmonics, but if you add a second tone and a tiny beat of amplitude modulation you will get an intermodulation that connected with real instrument natural distorsion generate the phantom lower tone. Matematically speaking the distorsion creates such a product between the two tones generating the math correct frequency distribuition.
Thats the same reason why valve amplifier are so much worm and preferred by a lot of musician instead of modern "perfect Linear" undestorted amplifier.
I love sound
The Sawtooth Habituation Effect reminded me of a similar experience I had years ago, listening to Charlemagne Palestine’s organ work Schlingen-Blängen. The sound of it has a flanger/chorus/vibrato like quality (probably naturally achieved). After listening to the 70 minute work, my ears were configured by its sound character so that when the work ended, when I spoke to my mother (I was lucky to do that - I could have stayed silent) I heard my voice having the same flanger/chorus quality in it. It was marvelous! And the effect lasted for at least 20 seconds. I listened to the last five or so minutes of the work again today but I did not achieve the same effect. It probably needs more time in order to be achieved.
Wow, neat!
The delayed speech trick is the technique used in speech jambers. You are speaking but then you hear yourself and your brain thinks it's someone else, bc of how the skull softens your own voice which isn't present in the playback, and as a result, you stop speaking to hear what the "other person", the interrupting voice, has to say but because it's your own voice, it becomes silent and you continue just to have this pattern repeated.
This will either frustrate a person or they will make a mental note to block it out. This is why echo on a phone call is so irritating and troublesome and why speakerphone should be abolished.
With that last effect you covered, I went to the other video to hear the effect there, and I noticed that if I listened closely with stereo on I could still hear the melody, mind you that I had mono turned off. The reason why this happened is probably the fact that the melody plays at the exact same volume with stereo as when mono is on, the only difference is on stereo the melody is drowned out by all the other sounds, if you use a sound like white noise which is basically just every frequency you can hear played at once, you usually won’t run into many issues. But if you use a sound where some frequencies are absent, then if the melody hits one of those frequencies, it won’t be covered by it, which can cause the melody to “leak through”, another cause is if the melody is too loud (kind of like how the Huggins pitch doesn’t work if your volume is too high), which would mean you would want to stay away from low pitched percussion like sounds like a base drum, unfortunately, the first note in the hidden melody in the Doom Eternal theme is a base drum, making it easier to pick out, and kind of like the cocktail party effect (covered in part 3/5), if you know what the melody sounds like, chances are you will be able to pick out the rest of it too. It happens that both were the case there.
5:38 kind of off topic, but do the spinning barber shop polls look like they're slowly rolling towards each other to anyone else?
ferry interesting serie! Thank you! On the "two melodica subject" This effect is also ferry pressent when two clarinet players play a C and E in the high register. It is a physic effect also essential used in electromagnetic waves, radio waves. It is an essential methode to discriminate (Sorry, I don't know the right English word) the audio content pressent in radio waves. (mixer)
I'm wearing headphones (Sennheiser HD 598SR), and for some reason, when I heard the white noise coming from just the sides with no mid, it felt like the noise was either coming from behind me or inside my head. Incredibly freaky.
Yeah, it's a strange sensation!
Probably my least favourite
18:50 getting flashbacks of manually spinning my grandparents record player. It wasn't plugged in, so I would play this game where I tried to spin the record steadily enough to make it sound normal and enjoy the tunes.
11:57 : i absolutely love the effect!
16:49 : i love this, too!
01:30 : I love the idea but i think the cause is different and very simple. It's not the room but the methodology. Every iteration with its playout and re-recording introduces a EQ manipulation, the peaks and notches in the freqnecy spectrum cumulate and after some cycles you only have some remaining spikes that sound what they are. This is the strange "ringing".
I think one can think of it as a slow motion study of the development of a microphone feedback.
Sawtooth Habituation Effect is so pronounced it's awesome!
As for the Moving in a Stereo Field effect - two things come from memory:
1) A concert experience of a quadraphonic recordings by Phill Niblock - somewhat steady but harmonically rich drones come alive by turning head or moving ever so slightly.
2) An album "Matrix" by Ryoji Ikeda, which is built around this effect, and suggests listener to create own experience by movement whithin a sounding space.
++ Contrary to Matrix, which is intended to be listened from speakers, he also has a "+ / - " album, containing among others three "headphonics" tracks, which explore headphones-related phenomenas.
Thanks for the series!
The effect you spoke of in the melodica intermodulation segment of the video is actually the foundation behind the science of binaural beats where two separate frequencies are played simultaneously (one in each ear). Even though theyve been isolated from each other our brain will perceive the difference between them as a third tone playing but if u take out one headphone out at a time u can hear that theres are actually only the 2
This is not a 100% accurate statement as the two frequencies are in fact, being "mixed" (Heterodyned) as mentioned and produce the difference and the sum just as mentioned. This is a well known phenomenon and has been used in radios the world over since the beginning. Ask any ham radio operator and they will explain the effect and how it is used to produce the Intermediate Frequency stage of any radio. The frequencies are higher in radio, but the results are the same. Radio's use oscillators to generate the 2nd frequency to mix with the radio signal's frequency to get the lower difference frequency.
The sawtooth after effect actually lasted for me into the talky after it.
You have a great way of explaining things. Awesome video series.
In simple mathematical terms, when you multiply 2 waves of different frequencies together, the result is 2 new frequencies, one is the sum of the inputs and the 2nd is the difference between them. cos(A)*cos(B) = cos(A+B)/2 + cos(A-B)/2. There must be some acoustic multiplication effect going on, which of course doesn't appear when you simply sum the 2 frequencies in software.
Yeah -- it's exactly that process I'm curious about... how the physics of waves bouncing around generates the sum and difference tones... and whether the many other distortion products are just sum/difference tones of harmonics or can also come from the fundamentals, etc. Mysterious. :-)
@@CaseyConnor I’ve got a feeling that it’s just frequencies moving through the air. If you were somehow able to play a high C, then play a note 4Hz above or below it, and really increase the sensitivity of the microphone, you would probably hear a 4 Hz tone. I remember when I was little hearing two people scream and hearing a low bass tone that was changing, going up and down in pitch. The tone was ranging from 320 to 400 Hz, and the people screaming were screaming at a frequency between 2800 and3K. I think it’s just that, but the melodica is probably adding frequencies due to it being reeds or something, I don’t know much about melodicas, but I do play piano and drums. These are just my thoughts about this.
I also have perfect pitch :-) :-) ;-) :-) ;-)
@@levi_gobin Yeah my math on this is shakey at best, but my understanding is that the sum/difference tones only show up when some kind of nonlinear distortion is present in the system. Apparently the intermodulation "Tartini" tones musicians hear come from nonlinearities in our own ears, and the melodica one comes from ... something :-) ... but I don't think they are otherwise present. At any rate, the mystery continues!
i accidentally found the sawtooth habituation effect after i had done some work on a song. i noticed when i started listening to other parts of the song after listening to a particular saw-y part of the song (i think i had the instruments soloed) everything i heard from then on a few minutes afterwards sounded the same as the example you gave after 40 or so seconds of the constant sawtooth wave being played. neat.
you describing what happens in the affect reminds me of an interesting phenomena than ive experienced phasing in and out of sleep in several different places. in some circumstances, it involved my screensaver, which is cotton candy pink text wobbling around, but if im half asleep, like actually half asleep, it appears to be a similarly shaded blue instead.
ok that sawtooth/white noise one actualy freaked me out, i genuinly thought you put in a shit quality white noise sample for no reason, but that was the illusion wtf lol
About the Doom Eternal one, Mick Gordon did it again on the track IV. Doom, this time with a recording of a woman whispering "Jesus loves you" backwards. Boards of Canada did it as well for an ARG in 2014, but the audio was spread amongst two Soundcloud tracks.
The overtones are created because the phase difference between the two notes you're playing produces a soundwave vibration that also affects the physical features of the instrument that made the sound; for example, A harmonica playing the two notes creates a vibration that also causes the reed of that note to vibrate. If it doesn't have a reed of that note, then even an integer value of the difference ex: 1/2, 1/3, etc. or 2x and 3x wavelength harmonic. If that exists in the instrument, it will set up a resonance that will produce the tone, including the cavity of a device if it is physically shaped to do it. So the sources of the lower tone depend on what kind of features exist in the instrument, which can sympathetically vibrate to one or more harmonics available based on the frequency of difference between the two notes played. A guitar is well suited for this. And it also explains why a guitar that is well played and often will through compaction and rarefaction of the notes on the guitar will cause the wood to reconfigure the vibrational nodes in a guitar making the guitar a better-sounding guitar over time; thus, the price goes up. 🙂 No silliness this time; I am a retired Metrologist. No, not the weather. Measurement expert! 🙂
6:55
I'm not an expert by any means, but I would hypothesize that the difference between the two tones will create peaks in the overall audio that is heard by the listener. The peaks are periodic and, as such, produce a tone with a frequency. The frequency will be the rate of the peaks caused by the two audios. Does this make sense? It's like the rhythm = pitch scenario, where the reoccurring peaks and troughs of two audios played together will establish a reoccurring beat that, when sped up, produces an audible pitch.
Yeah I'm right there with you in an intuitive sense -- but the thing is that when you play two tones simultaneously, the beat that happens between them is not itself a tone -- e.g. if you mix 100 and 150 Hz you have a 50Hz beat frequency, but that beating is not the same as a 50Hz tone, because there is no true 50Hz frequency component present. Consider the example I show in this video where I play the two tones separately and then combine them in the computer (as described with text on screen) -- the difference tone, and all the other distortion tones, are not present (it's not in the spectrum shown, and it is also not audible to your ears as it was before.) Said another way, if you have two melodicas playing the two different tones near each other, this effect doesn't happen -- it requires the tones to be produced in the same instrument. The rhythm -> pitch idea you describe is different, because playing a rhythmic sound quickly actually introduces frequency content at that frequency. Said another way, if you have a 100 and 110 Hz tone combination, you hear a 10Hz beat. But if you gradually increase these two frequencies at the same proportion (eventually getting to 200 and 220, 400 and 440, 1000 and 1100, etc) the beating effect does not become a new tone once it climbs high enough, instead it becomes a sense of "roughness" we hear and which contributes to the sense of consonance or dissonance. So AFAICT there needs to be a physical explanation for why the difference tone would manifest in the melodica, and I think intermodulation distortion is that explanation -- my best guess (as described in the vid) is that the beating has an effect on the dynamics of the instrument causing a real tone at the difference frequency to manifest... sorta how we would intuit a beating frequency to work in the first place, ironically.
The delayed speech trick is something encounter frequently using radios to communicate at work. When your voice comes out of someone else's radio Close by so you can hear what you're saying well, half a second later, it becomes almost impossible to talk coherently! 15 years doing it, and I still struggle with it!
14:20 That is an interesting video effect. Your hands fade in, but you somehow completely matched your exact facial expression, head tilt, and head location. I cannot see the transition.
That's my compulsive overuse of Davinci Resolve's "smooth cut" transition in action. :-)
@@CaseyConnor I actually really like it, and which it had become the norm in short video content, rather than the ever present jump cuts.
10:15 Plausible explanation for the unexplained tone.
Is the deep tone in question different when the demonstration is preformed by a drastically different sized person? Eg you appear avg size hight build( average ≈ 5'8" 160#) are getting the shown tones were some one 6'8" should get a slightly higher result, 4'8" slightly lower. Build Characteristic would also play in As bigger longer bones would Have a higher sympathetic resonance frequency and heavier set people would have more sound insulation on their bones
I live I a noisy household so I sleep with white noise.
for the white noise part, i listened with my eyes closed. it sounded like the sawtooth was getting incrementally louder and then as the white noice kicked in the was a weird screaming/scraping electronic (im not 100% sure how to describe it) sound for a second before the white noise played as normal. I'm not sure if that was the intended effect but it scared the hell out of me 😂
Yeah it's a crazy effect! And the "increasing volume" illusion of the sawtooth is almost worthy of a name itself.
Melodica Intermodulation is exactly as you describe: heterodyning. This same principle is exploited in radio communications to shift high frequencies down into whats known as a "base band range" by mixing with an intermediate frequency in order to make demodulation of RF signals easier. The two tones here are mixed together (multiplied in a DSP context) to provide your negative frequencies and positive frequencies.
Thanks! So are the other distortion products (meaning, besides the sum/difference tone of the two fundamentals) just heterodyne products of the various overtones? And do you have any sort of intuitive grasp on how the physical phenomenon itself happens? Meaning, what is it about the physics that effectively causes the two signals to be "multiplied" in the melodica/harmonica?
HEY I like your videos....
on the experiment of melodica, I think is an acoustic thing.
maybe is resonance because 2 holes are resonating, so the area where the wind is passing is added.
sorry about my English.
12:30.... well, good thing i'm not editing audio today... Hello Tinnitus.
edit. i was early; 14:00 NOW my tinnitus is really pissed.
a lot of feedback on the lucier effect, i wonder if its due to how drivers in themselves work as well as some other engineering voodoo.
possibly a relation between how microphones typically pick up sound and how speakers produce sound
For sure: the entire chain colors the response of the overall system: the loudspeakers, the room itself, the placement of the mic, the mic itself, and (to a much, much lesser degree) the audio interface.
@@CaseyConnor im gonna have to figure out a way to do this, since i dont have a speaker system and im thinking of trying it for a noise album.
ive experienced similar effects from the good ol days of recording from crappy desktop speakers to my 3ds as a kid, so im wondering if theres a similar kind of interference.
maybe especially tinny speakers/microphone mixed with a muddy counterpart could produce an interesting effect?
@@sunburnedshirts3724 Worth a shot! You'll probably want to adjust the gain of the mic as you go along (otherwise it can "run away" and start clipping the input, or fade away into nothing -- the trick is to keep the level near the balance point.)
2:27 okay that started sounding like the Plasm Wraith there--
I noticed another one recently, normally sawtooth waves sound very piercing and such, but I discovered that if you listen to a square wave first, then switch to a sawtooth wave of the same frequency (this has to be an instantaneous jump, there can’t be a gab between the two) the sawtooth wave sounds much less piercing than normal, I discovered this when watching a TH-camr recreating different sounds with sine waves, they did a square wave, then a sawtooth wave, but I noticed that the sawtooth wave didn’t sound like a sawtooth wave, so I rewound, skipped past the square wave, and it sounded more like a sawtooth wave
Yeah, nice observation! That sounds like a very similar effect -- I bet what's happening is this: the square wave doesn't have as many harmonics as the sawtooth wave; so while you listen to the square wave, the harmonics sort of "burn in" to your ears (just like in the demo in this video); then you switch to the sawtooth wave, which has twice as many harmonics as the square wave but which overlaps every other harmonic with the square wave, so for a few seconds the sawtooth wave doesn't sound as shrill because many of those harmonics are "burned in" from the square wave. I just tried it out and heard the same thing you describe. Thanks for pointing it out!
11:18 I didn’t like that afect it gave me a jump scare and I nearly started crying 😢 I don’t like loud noises
The sawtooth habituation effect jumpscared me somehow. When the white noise started, I heard what I can only describe as a "boing" effect and it was so sudden and unexpected that even watching the 2 arrows indicating it coming, I wasn't ready.
Update: I tried it again and it made me physically feel sick. It's a cool effect but it's so unexpected the way it made me feel!
Wow sounds like a strong reaction :-) i agree though, it's a weird effect.
Both of you you know what I’d rather get bullied in school then hearing that white noise sound with the boings
I had something similar. I was re-watching this video and when I saw that part coming I had to skip it. Last couple of times I’ve made that effect for me or other friends and I listen to it, I get a weird ass feeling! I can only describe it as that feeling of when you’re in the room all by yourself, no one in the house at all ... ... ... Silence, ... ... ... BOOOOOOOOOOOOOOOOOOOOOOOOOOOOM! The loudest noise possible happens. I really love affects like this. It’s interesting how just two sounds that when played by themselves are seemingly harmless until they are played one after the other. I wonder if there’s something similar with the reverse effect, 40 seconds of white noise or other colored noise, then switch to a sawtooth or square waves. I’ll have to try that out. It’s really interesting and I wonder what similar effects there are.
I’ve seen so many “audio illusions” videos on TH-cam, but none of them are stuff like this. they just include stuff like the Shepard Tone, the Mc’Girk effect, and that’s pretty much it. I really wish there were more effects like this and others that you’ve covered in the series. Keep up the good work and can’t wait for part eight. :-)
Welcome back!
Absolutely amazing video series ❤
I wonder how much the Sitting In A Room thing is more to do with resonances caused by gradual deterioration of the audio. In fact, I did this same thing as a kid, loads of times. I used to multitrack from one tape to another. I'd make audio sketches, doing one bit of dialogue and leaving a space for the next me to fill in, if you see what I mean, or sometimes I'd just try to create the sound of a crowd. Now obviously, the tape recorders I was using were hugely inferior to the kind of mics and speakers which would have been used in this recording, but in a way that emphasizes the point. You could hear the artifacts creeping in with each repetition. The audio would degrade, and on top of that, you'd get the original echo of the room, with the echo of the echo, and the echo of the echo of the echo, ETC ETC, until all the different me's sounded like they were in different places. The version of myself I'd just recorded sounded fine, but each one that went before sounded worse, and the original me sounded like I was in a different room, talking in to a tin can or something. So yes, I guess the room is adding to it, as the mic is picking up the resonances of the room all the time, but it's as much to do with the steady degradation of the audio, because no matter how good your equipment is, you're always going to get that over so many repetitions. Obviously this is why copies of music were made from the same master, and we don't copy the copy the copy the copy ETC. That's why those old tape delays sound so cool, because the frequencies are altering with each copy. If you make it long enough, the end result becomes unrecognizable from what you started off with. And in that instance, there is no room acoustics to affect the sound. So in this Sitting In A Room example, it's hard to tell how much of it is the room sound and how much is audio degradation.
If you connect a modern quality interface to itself via a cable (i.e. output connected directly to the input) you can go maybe 100 or more generations before you can notice anything, and even then the effect would be quite subtle. Even the coloration of the microphone in the demo in the video isn't noticeable. So what you are hearing in this effect is really just the effect of the room, with a small percentage coming from the response of the speakers. You could say it is "deterioration" of the sound, though... That's basically the effect of recording the room acoustics: hearing the echo of the echo and so forth, as you put it. The effect of the room is huge, and drowns out all other effects (unless you're in an anechoic chamber or something.) Even a first generation recording will see dramatic impacts on the sound from the room. I also used to do tape recorder overdubs as you describe, and in that case I would agree that the effects of the tape (hiss, loss of high frequencies, drift and warbling of pitch, etc.) and of the cheap mic I was using, we're huge. Good times. :-)
@@CaseyConnor the band Neurosis / Tribes of Neurot tried the same experiment with their album "A Sun that Never Sets".
The propagation of soundwaves through the air is nonlinear. Soundwaves can be explained as variations in air pressure, but we also know that the speed of sound depends on air pressure. The nonlinearity becomes stronger as the sounds become stronger.
When a full orchestra plays a very loud strike, the reverb will be slightly lower in pitch. Inside the melodica, frequencies may be multiplied by this nonlinearity, like in an electronic ring-modulator. But I don't think this effect is strong enough to explain the ghost notes in the melodica. The airflow around the vibrating tongues of the melodica must be far more nonlinear. :)
The melodica intermodulation thing is 100% intermodulation distortion. So if F1=1047 and F2= 1247, the difference tone and combination are what would be referred to in electrical engineering (particularly radio frequency engineering where signal mixing like this is very common) as "second order" intermodulation products and the other tones you see 200Hz above and below at 847 and 1447 would be "third order" intermodulation products, so 2*F1-F2=847 and 2*F2-F1=1447 (3341 and 3541 are also 3rd order products) and you can also see just a little bit of 1647 as well, which would be a fifth order intermodulation product of 3*F2-2*F1=1647, and so on, but usually the intermodulation products drop off in amplitude and get more numerous the higher in order you go.
09:23 Thoes two notes played together sounded like the EAS Attention Tone
Mick Gordon also did the same kind of thing with the track VI. Doom in DOOM (2016)
Hey your friend did really good actually
I know, right... I was hoping he would have more trouble so the video would be more impressive. :-) He's a musician and has experience hearing himself in headphones while recording so ... Maybe not the smartest choice on my part.
Have you ever gone over ASMR tingles? I've always wondered the science behind that.
I haven't, and yeah, it's an interesting topic... I have thus far assumed that there are lots of videos on the subject already, but maybe some day. :-)
The Sawtooth Habituation Effect didn't work at all like it seemed like you were describing...
Instead, for a few seconds, the white noise was all wARbLy before fixing itself.
oh that sawtooth was HARD to listen to. but that aftereffect was so worth it. that the HELL was that??
I tried out the 2nd Audio trick & it was hilarious 😂
I think it makes people sound like they've been drinking.
It's a problem public speakers have to deal with when they have to speak in front of a large audience & they speak into a microphone that sends their voice thru loud speakers so everyone can hear. There is a slight delay between when you speak and when your voice thru the loud speakers sometimes cause an echo from the room.
If you're used to public speaking in situations like this(like my father, who I tested along with several others who were not public speakers)
While the effect worked on the other subjects, my father was able to speak normally.
I thought that was neat.
For some reason this effect helps people who have a stutter speak normally & is a tool used by speech therapist. 🤓
Neat - I hadn't heard of the speech therapy angle.
The sawtooth effect is very cool. Not only did I get the flange thing at the end, but as the sawtooth played, it was like it got louder and louder and louder, till it was almost unbearable. I had to remind myself that it wasn't getting louder. This is a cool effect because unlike the Yani Laurel thing, it works every single time. With the Yani Laurel thing, I canot hear Yani any more, unless I bring the pitch right down. It's like, once my brain knows the trick, the trick doesn't work any more. But that sawtooth thing, in addition to the cool white noise bit at the end, is quite scary. I've heard on science fiction shows where they try to make sounds which are supposed to have a really bad affect on the people in the play or whatever it is. But that simple sawtooth was more affective.
Totally agree on the "sawtooth getting louder" effect! I debated whether to include that as a separate effect in the video but in the end i decided not to. But yeah it's crazy how it can seem to increase so much. Also agreed on Yanny/Laurel... I heard Yanny like twice and it blew my mind. Every other of the 5000 times I've listened my brain just refuses to switch off of Laurel no matter how i try to fool it.
When I was briefly listening to the sawtooth effect earlier, not only did it seem like it was getting louder, it seemed like speakers were playing it from all angles all around you, and they were all getting closer and closer and closer to you, until,... 🤯
4:25 the anxiety theme song
4:45 the depression theme song
5:15 ALL TOGETHER NOW
The sawtooth wave thing was crazy
The acoustic explorer is back!
Is mick Gordon effect what they've used in planet sounds?
Hmm, not sure -- what exactly are you referring to? Got something specific I can search for to check it out?
@@CaseyConnor these bright side in yt. Said planets have sounds...
@@erpmo3326 I can't find any videos they have posted that are the same thing as the Mick Gordon effect, but if you have a specific video to check out I can try again.
@@CaseyConnor th-cam.com/video/UTAPvPLb7t4/w-d-xo.html
th-cam.com/video/IQL53eQ0cNA/w-d-xo.html
@@CaseyConnor don't mind me, I'm american
The Lucier Effect could be used in a video game for theorists to solve.
Lucier effect is the audio equilevant of deepfrying pictures
The sawtooth and white noise effect has an optical twin where you stare on a dot in a false color picture. Then the picture is suddenly switched to black and white and for some moments you see a full color picture.
th-cam.com/video/3P8q_dCU3RI/w-d-xo.html
Amazing series :)
Your melodica intemodulation effect is well known to the radio world, sideband radio is dependent on the phenomenon. Any modulation between two frequencies will produce a sum and difference component.
In radio, these are called sidebands.
In blues music, the effect is often utilized by celebrated electric guitarists, most noticeably when they play minor thirds in association with clipping distortion.
Speaking of which, the harmonics produced by vacuum tube clipping and tape saturation are widely recognized as have pleasing effects, although discussing them inevitably leads down a rabbit hole.
And dithering is a known psychoacoustic tool I do not fully understand, I don't remember you covering it.
Yes, sum and difference signals are produced by nonlinearity (mathematically multiplication of the signals or "modulation" rather than addition). It's interesting that modulation happens in the melodica, an earlier commenter explains the mechanics. Nonlinearity can occur in electronics accidentally or deliberately in a mixer. It seems that your ear or your brain can do some mixing as well. The air can become nonlinear at very high volume levels (dragster or jet engine levels of sound).
infomative as always
That sawtooth wave was worse than going to the dentist. Couldn't handle it.
He's not having a stroke i promise
That made me laugh
"Something special happens at 4:20"..lol✌️💨
The Melodica Intermodulation is not specific to the melodica. Most organ sounds (straight, witouth vibrato) produce the same effect. I remember years ago a friend demonstrated this to me in his synthesizer with and organ sound. I just recreated the experiment using a software organ. So, it is either indeed an acoustic phenomenon, non specific to the melodica anatomy, or the synthesizer manufacturers go to extreme lenghts to recreate such phenomena (which I doubt it). Playing and holding G4-C5 produces and audible C4. G4-B4 produces a D4. G4-Bb4 produces a non-tempered note between D#4 and E4!
Thanks! It's also worth remembering that there is a psychoacoustic as well as physical component (though the psychoacoustic side has always been weak for me). Try recording the G4 by itself onto an audio track, and the C5 by itself onto a track, and then play those together in the DAW as audio tracks, and look at the result in a spectrum analyzer. There won't be any C4 (assuming there wasn't any C4 in the individual clips) but if C4 is audible, it will be due to the psychoacoustic effect. Then play G4/C5 together in the synth and look on the analyzer... I wouldn't be surprised if they do simulate the intermodulation in an organ synth (i's a pretty significant part of the sound, I think) but it'd be neat to find out for sure.
@@CaseyConnor When you play the notes on the melodica the subtone is clearly audible (I’m not sure “subtone” is the word to use, but you get what I mean). When you play them separately and then combine the audios it is scarcely audible. So we suspect a physical aspect of the instrument. But in the case of the organ samples there is no such aspect - only the perceived sound by our ear. I will do the experiment you suggest and let you know if I still hear the subtone. I suspect I will. When the friend first demonstrated the effect to me, the subtones were almost as loud as the real notes! In my recent recreation I needed to pump up the volume. But they were there without a shred of doubt. Now, the subtone I’m hearing is not the difference in Hz of the played notes. I play Bb4 (466.16 Hz) and G4 (392.00 Hz) and I get something in the vicinity of 320 Hz (between D#4 and E4). Where does this come from? I will also try other intervals and see what I hear.
@@Boukouvalas1979 Oh, interesting that it's not the usual difference tone. Yeah, give it a shot. My money is on the synth simulating it, but if it's just playing straight samples then maybe you just have a really effective demonstration of a psychoacoustic effect going.
@@CaseyConnor I did the experiment today, in the way you suggested. I still hear the lower "phantom" notes. Sure enough, there is nothing there in the spectrum analysis. So this is purely psychoacoustic. It is interesting though that the notes need certain sound characteristics to come out. No matter how fast/loud or for how long I strum piano keys on the same notes, I never listen to lower notes there. Sound has to be continuous for sure. Organ sound works well. Accordion samples also did the trick. The fact that I could hear the exact same by-product notes 20 years ago, in my early 20s, means that this is not an age "condition". Also, the friend who presented the effect to me was listening to the same phantom notes. Finaly, as I said, in that example, they were shockingly loud, almost as loud as the real notes!
@@Boukouvalas1979 Nice! What's the organ synth or samples that you were using? I've had the same experience of sensitivity to the instrument chosen - I think you mainly need a sustained tone with relatively rich harmonics, but that's just a guess. Perhaps that's why they are reported in violin as well.
The Lucier effect sounds like how they got the cave noises for Minecraft.
At 4:46 "...and the calliope crashed to the ground."
Lucier effect is what they play for music in horror movies lmao
listening on mono, lets rock
nice
At first I thought the lucier effect was going to be dumb but it sounded really cool by about iteration 16 and 22. by 47 its pretty much homogenized and boring. Cool trick.
4:50 to be played in elevators
👁🔺