The "ringing" is actually an artifact of creating lower pitch square waves by adding progressively higher and higher multiples of sine waves together, which is probably where the harmonics come from as well. It's something Fourier discovered while trying to explain the phenomena of how heat dissipates through different temperature rods before finally coming up with the now-famous Fourier transform.
indeed, it's called the Gibbs phenomenon and it is present in all waveforms that has a finite discontinuity (a jump in amplitude) like the square wave, but also the saw wave too.
he’s not wrong in sayings it’s because of the low pass filter. As you continue to add sine waves, it gets closer to a perfect square wave, so when a low pass filter is applied, it becomes less like a square as it loses the higher frequency sine waves, creating that “ringing” phenomenon. The square wave isn’t being created by layering sine waves, but every sound can be expressed as many sine waves put together, and using filters, sine waves of different frequencies can be taken out
a speaker cannot produce square waves, because it would require the diaphragm to move infinitely fast or at least extremely fast compared to a sine wave.
I noticed that even at the highest frequencies it was still possible to detect a difference, the sine wave actually feels bigger to me when I hear it but the difference becomes is distinct the higher you go. the key to hearing it is, listen to the notes switch back and forth, one transition will feel like it's dropping or shrinking and be less pronounced (sine -> square) and the other will feel like it's the same note but expanded out a bit and with a more pronounced distinction as it transitions (square -> sine)
It might be because the switching wasn't exactly at the zero-crossing point because the zero-crossing point would most often be between samples. I might at some point repeat this test and possibly investigate further. DM
Very educational video! The explanation helped me understand why a sound I was producing sounded like it was ringing. On a totally unrelated side note, one suggestion I would make would be to try blinking a few times a minute while recording. Not blinking can be a bit off-putting for a viewer, though they might not consciously process it.
@@AudioMasterclass Well I think you did great overall! That's just a suggestion for a slight course correction. I'm definitely not trying to trip you up. 😂
I feel very smart cause I actually picked up on this happening on my own while making chiptune! The explanation however makes so much sense!! I just never considered it
Amazed by the details given in a video of 9 mints. Appreciate the clear explanation on the topic. Thank you. I am getting into Vehicle Acoustics field and this would help me to look things differently. One question related to content, I am looking for such detailed explanations on harmonics in Freq spectrum. It’d be great if I could get some information on that.
This video explains very well what I've noticed when messing around with my signal generators and audio amplifiers, I thought this was from me not having a musical ear or from me not constructing my signal generators well enough. 😄
I think that my two maths A levels and one university module might be insufficient for my comment to be reliable, but it seems like the Gibbs phenomenon is due to summing the Fourier series only partially rather than letting it go to infinity, where the phenomenon would disappear. This is very similar to the ringing displayed here - If the frequency response of the signal generator (or the filter in the D to A convertor) were infinite, then there would be no ringing. The ringing occurs due to the limited frequency response because not all of the harmonics of the square wave are included. That's my guess anyway. If there are any maths or electronics experts out there who would like to add to this, or show me where I'm going wrong, please feel free to comment further. DM
@@AudioMasterclass I think that is more or less correct, any signal with some discountinuity will have this type of ringing due to the fact that you cannot take into account all the terms in the Fourier expansion. However interestingly enough the "ringing length" goes to zero as the number of terms is increased but the overshoot does not it will always be a little bit of. The Fourier series converges point wise but not uniformly in math lingo
I studied chemical engineering and NOT an expert at this realm, so I'd just point out my observation. From what I see on Wiki, Gibbs phenomenon looks symmetrical (i.e. post-ringing + pre-ringing). While the square wave shown here 3:21 displays an asymmetric gragh (only post-ringing). Although I do agree the principles between the two seem to be very similar. My guess is that perhaps they are related, but not completely the same?
@@owlmega-101 That's a good point. I'm not going to burn my brain thinking too much but I'm guessing that maths doesn't need to bother about time only going in one direction, whereas in analogue audio time definitely does only move to the right on the x-axis. That's just a thought so we await a real maths guru to illuminate the issue. I say 'analogue audio' because in digital audio, linear phase filters can definitely access the past, but only by delaying the present. ua-cam.com/video/UFrbKqRuKxI/v-deo.html DM
Very interesting that the 10kHz square wave showed multiple lines. That's because 10kHz doesn't evenly divide 96kHz, while 6kHz, 8kHz, and 12kHz do. So each cycle of the 10kHz square wave has 4.8 samples low and 4.8 samples high. But it can't subdivide samples, so it will be either 4 or 5 samples each time. The first integer multiple of 4.8 is 5, 4.8 x 5 = 24. So every 5 cycles of the square wave it completes a cycle and comes back to where it started. Looking closely, there are 5 individual lines. I worked this out before you said you were using 96kHz. I figured you were using 48kHz, the math works out the same. You definitely weren't using 44.1kHz or 88.2kHz or anything else like that.
Well, In the beginning I was honestly listening with intent, but I had to start the vid over when I suddenly realized I was actually wondering when this man was actually going to blink, if ever!🤔 Haha...to be continued👍😉 (and I'm sorry oh blinkless one..but if theres any consolation, I did subscribe to you!)
It's a flow transition in Final Cut Pro. When it works it's good. When it doesn't... well you have seen already. I've given up on that as you can see in ua-cam.com/video/Ht3pMAkzums/v-deo.html DM
Can you SEE the difference? I.e. in a flickering light? I find it easy to produce a square wave stimulus at a given frequency, but i find producing a sine wave at a given frequency hard.
It's interesting to notice that the square waves appear to begin and end exactly at the inflection points of the sin waves; that is, the individual square waves begin and end where the sin waves change from concave up to concave down. Although, I suppose this only makes sense... As an interesting side note, these points can be mathematically calculated by taking the second derivative of the sin function, setting it equal to 0, and solving for (in this case) 't', though a sin function will have infinitely many. Only the first couple points are necessary as the inflection points of sin, and thus, the structure of the square wave will repeat on a given period. Cool beans, thanks for the visualization.
After about 6kHz both of the signals were identical when hooked on my own scope :) Funny to see we both use HAMEG scopes. I've the HM203-6 scope, however the line doesn't hit the edge of the display without using the x-mag button, stops about one div off each side.
haven't finished the video, but i'm going to make a prediction that the reason that they begin to sound the same is because the 1kHz square wave is just made up of a 1kHz sine wave with a bunch of higher frequencies added on. as the base wave gets higher, the higher frequencies eventually shift out of the range of human hearing.
My ex-husband did audio tech and used to put on this CD for checking your hearing at different frequencies but wouldn't tell me, just to see if i could hear it i guess, and i'd walk around the house trying to find what tv or computer monitor had been left on mute... the bright ping of those high frequencies can be inescapable!!! :O :)
Very interesting video, thank you! I am just curious about how does your digital audio workstation generate the square sound wave? What is the input electrical signal of it?
The ringing isn't caused by filtering. It's caused by a fundamental law of nature. A "true" square wave requires an infinite number of harmonics. Even if you were to create the square wave algorithmically through the equation sin(2* pi * freq * t/nyq) > 0 ? 1 : 0, to actually play it back the speaker driver would have to move at infinite speed at the square wave's edges and stop infinitely fast, which is of course impossible. Thus a square wave not a band-limited signal and can't be recreated by sampling. It's called "Gibb's Phenomenon," and it's the kind of thing you'd expect a "Masterclass" to know about... It's also the reason things like the Heisenberg uncertainty principle and the Casimir effect exist...
Just have to put on my physicist hat here for a moment: The Ringing in the signal is what is known as Gibbs phenomenon. It is an intrinsic artifact that shows up when you do a finite fourier series decomposition of a discontinuous (ei it has a sudden "jump" from one amplitude to another) signal. No piece of electronic equipment can generate a "perfect" square wave, because that would require an infinite number of sine waves of increasing frequency and decreasing amplitude added together. We have to put in a frequency cut-off somewhere when creating an actual physical signal, and as you pointed out, the available frequency range of the playback device is usually the limiting factor. But as soon as we cut off our fourier series of the square wave, mr Gibbs pops up.
Thank you for your input, which I appreciate. There was some discussion of this earlier, if you look down the comments for SpyProductions. As for generating a perfect square wave, adding up sine waves would be a long-winded way of doing it and it's simpler to flip from low to high to low etc. That of course is also imperfect due to the rise time, but a decent function generator should be able to create a square wave that is good enough for any audio purpose. I doubt whether that would prevent Mr. Gibbs from popping up though. DM
It is easy to explain, the smaller width and distortion at the top of the square wave make them more similar because the angle to go up and fall down of the sine wave is sharper. The sine wave starts to look more like a square wave at higher frequencies because the width is smaller and the top angle is sharper. Actually both start to look more like a sawtooth.
A square wave formula is A = sin a + 1/3 sin 3a + 1/5 sin 5a + 1/7 sin 7a and so on. The higher the tone, the harmonics start going out of your hearing spectrum. 4kHz x 3 = 12kHz so for mee that is the end at 52.
If you add a constant to an audio signal you get a DC offset. That will cause switching clicks and in the (probably) worst case could bias a loudspeaker cone causing distortion on positive or negative peaks, depending on the sign of the constant. That's my first thought off the top of my head but I'm sure there's more to it. DM
You would need a mathematician to explain where they come from - try en.wikipedia.org/wiki/Square_waveen.wikipedia.org/wiki/Square_wave - All I know is that they are there. You can't see them on an oscilloscope but a spectrograph will show them clearly (in a square wave). DM
If you look into the way Fourier came up with his Fourier transform, it suddenly becomes clear. Square waves can be made by adding together progressively higher and higher multiples of sine waves. 3 blue 1 brown has a fantastic video on it. The ringing he pointed to near the end is actually an artifact of that process.
Your filter which removed high frequencies is removing some of the terms of the Fourier series which comprises the square wave. This effect is more noticeable for higher frequency waves, as more of the terms are removed. This is visible in the CRO. So part of the reason for them sounding more alike is that the resultant wave forms are more alike. (and it seems my sound card stops at 12 kHz! Or at least I hope it's my sound card & not my ears!)
I'm going to politely disagree with you. This informal test works just as well in the analogue domain where the square wave can be clean up to and beyond 20 kHz, so it's what you can and cannot hear that makes the square wave sound like a sine wave at higher frequencies. The filtering visible in the oscilloscope is that due to the low-pass filter in the digital-to-analogue converter through which the digitally-generated sine and square waves are output. When an analogue version of the internet is available I'll be pleased to repeat the demonstration. Thank you for taking the time to comment, and P.S. It might be your ears! DM
@Audio Masterclass I have a question. Why does low hertz sound low. What I mean is - in my phone i have a frequency generator app, keeping the volume of the phone constant. When i increase hz from 20hz to 2k hz why does the sound increase?
Do you think our body, or electrical body, (aura?) acts as a filter to all the frequencies surrounding us? That frequency seems to be like tinnitus. Because if you listen, you can hear all that static in your ears, and perhaps illnesses get in by a dissonant tune that wasn't filtered properly. What do you think?
Hi, Thanks for the knowledge you share with us. I have an off-topic question: I'm interested about the famous $5 preamp, is there a final schematic as well as a list of all the materials? Also, I see that ebay has many types of INA217 and OPA 2137, what is suitable for this project? Sorry for the off topic and thanks for your patience.
I personally love square and sawtooth waves, since they are fun and you can hear the tone clearly (due to harmonics and overtones). But, I hate sine wave, since it is plain boring + at lower frequencies, you have to turn volume up in order to hear the tone (if the frequency of the tone is 27,5 Hz, the sound pressure has to be around 70 dB for sine tone, which means that you have to turn volume up (and with that, you're risking to damage your hearing)). But with the sawtooth or square waves, you can clearly and without turning volume up hear 27,5 Hz sound. Sine wave is boring, bland and simplistic, it doesn't have any variety. Sawtooth and square waves are more interesting and fun.
The shape is due to the low-pass filter in the digital-to-analogue converter which is inevitable in digital audio. If I do this test again, I might use an analogue function generator which will show much cleaner square waves on the oscilloscope. DM
This is how I know I am ready to turn 40, when I bought fancy headphones, started worrying about what my amps are doing, and an oscilloscope and some webpages about measuring output impedance later, here I am, worrying about how much the amps are ringing.
The ringing you see here is way above the range of human hearing, and when you do turn 40 your high-end will decline anyway. So, nothing to worry about! DM
Well done. If you create a sine wave in your DAW that is as high as you can hear and save it, you can check back again in a year's time and see how things are going. Let us know how you get on :-)
@@AudioMasterclass I put it down to wearing ear protection at live concerts, even when this was considered a very strange thing to do back in the 1990s and 2000s. The last ten years I have begun to see other people wearing them though. I also do not listen to music with exaggerated bass frequencies, e.g. club music, EDM, Hip Hop / Rap etc. Recently I had my music in the car a bit too loud in my opinion, but when I got out and shut the door I could barely hear anything was playing at all. Contrast that with other people, where you can hear and feel their car coming from hundreds of yards away. So the extreme volume and decibel levels inside the car, must be doing them incredible damage. It's so loud I would possibly physically vomit; would not like to test this theory out though! I've also always monitored my sessions and mixes at low to moderate volumes, turning up only to test briefly then back down again. I often hear things in mixes and turn to people around me to ask them about it, but they always shrug their shoulders as they cannot hear it. Even when I move my hand in time with the thing I want them to hear, they still have no idea.
This is very helpful but I can not find a reference to a very obvios point. If people are listenting to this on typical laptops peakers then the opportunity to hear what is being presented is going to be severely hamperedd by the roll off at higher frequencies. People are goong to be biased towards hearing no difference between different waveforms due to this and at higher levels there will be no sound at all in some cases. I suspect that even basic smartphones have a better presentation at higher frequences than many brand name laptops.
You're exactly right of course. I used to warn against listening on laptop speakers but I came to a point a while ago where I thought that anyone watching a pro audio channel would know this. But I might go back to the warning again. DM
This could be an issue with UA-cam's encoding, or with your computer's audio output, or your audio interface. Maybe also your headphones/loudspeakers. Other than that, an audiologist might be able to advise further.
A sawtooth wave sounds more similar to a sine wave than a square wave does but at lower frequencies the harmonics will be audible. For a 5 kHz square wave, the next harmonic after the fundamental is 15 kHz and is probably inaudible to most people, so the square wave will sound like the sine wave. A sawtooth wave at 5 kHz will have a harmonic at 10 kHz, which probably is audible. So at that frequency and lower the sawtooth wave will sound different to the sine wave. At 7.5 kHz, all three will sound the same, assuming the listener can't hear 15 kHz and above. That's my long answer to your short question, but the best thing to do will be to try it for yourself. Bear in mind that any distortion in your listening system may confuse the issue. DM
Very interesting. It's the same with tempo and rhythm in some ways. (I.G.) As tempo increases for a jazz ride cymbal pattern based on the the triplet, it 'flattens out' and becomes indistinguishable from sixteenth notes. (Basically, as the gaps in time shorten, there isn't enough space to create the triplet that would be perceived by the ear.) This is same experience here except... we have frequencies which obviously get faster as you go higher and also 'flatten out'. Or sound more 'rounded out'. It's kind of like a musician's version of relativity. :-P
Agreed. It's also interesting with an analogue oscillator to set a square wave to around 1 Hz, which will sound like clicks. Then increase the frequency slowly. The clicks get faster, then at some point above 20 Hz they will start to sound like a musical tone and the individual clicks will disappear. DM
So does this mean in a PEMF mat that has sine wave only that by lowering your frequency from 3 gauss or 1-30 frequency to the lowest setting you can get a square wave?
It would be the other way around. If your mat uses a square wave and you can set the frequency high enough it will feel like a sine wave. That's my hypothesis, but it would need an expert in this kind of therapy to advise reliably. DM
If the difference at 0:22 isn't extremely obvious then there must be a limitation of some kind in your playback system, probably a poor high frequency response. Also, since the high frequency response of the ear diminishes with age or other aural issues, then perhaps your hearing is the issue. We are not medical professionals so we cannot give advice on this, but it might be worth consulting an audiologist.
I can hear the difference up to 6 kHz. But this experiment is flawed, since your "square wave" is not really a square, which can even be seen clearly on your oscilloscope: due to filtering, most of the higher harmonics get filtered out and what's left is just a couple of regular sine waves mixed together. That's why they sound so similar.
The video covers this issue but I'll explain briefly. The waveforms are digitally generated and are output through an audio interface. This filters out frequencies above 20 kHz or so. This is why the square wave isn't exactly square. I could have used a function generator and the square wave would look square, but *your* audio output filters the high frequencies too, so what you hear wouldn't correspond with what you see. The way I did it here, what you see corresponds closely to what you hear. The similarity between the sine wave and square wave above 6 kHz or so is due to the harmonics being higher than most humans can hear, and above 7 kHz they are filtered out in the D-to-A convertor, and even if they were not filtered, you wouldn't be able to hear them. DM
While no-one at Audio Masterclass is an audiologist, our informal response is that the natural loss of high-frequency perception is part of the ageing process and does not seem to affect people's enjoyment of sound. However it may be the case that there is high-frequency content in your recording that you can't hear that younger people might find unbalanced, or even unpleasant. It would be wise therefore to get the opinion of someone who can hear frequencies at least up to 16 kHz or so before releasing your work to the public.
@@AudioMasterclass thanks, I never tought about it before, I used to make older people ( parents) listen to pieces of work of mine before the releasing to make sure the sound was balanced from 2K to 10K, it's like using a natural bypass, thanks for your answer, I really like your videos, cheers from Algiers!
In a speaker with a woofer and a tweeter, the woofer handles most of the power. That's normal for speech and music. If you drive the speaker hard with a sine wave above the crossover frequency, or a square wave which is rich in harmonics, you could send the tweeter more power than it is designed for. The voice coil will partially melt, breaking the circuit, and there will be no output.
We don't have experience of this, but running audio equipment from a modified sine wave inverter doesn't seem like a good idea. To be sure, you might consider contacting the manufacturers of your equipment and see what they say. There's an interesting thread at www.reddit.com/r/audioengineering/comments/1m5c6t/pure_sine_wave_vs_modified_sine_wave/
![[UA] NAVI проти G2 Esports | Blast Premier Fall Final](http://i.ytimg.com/vi/QWlIm4FGESQ/mqdefault.jpg)
Get your 30-day FREE TRIAL of the Audio Masterclass Music Production and Sound Engineering Course www.audiomasterclass.com/free-trial?aid=12
wat
The Square wave is basically just full of sine waves in a Frequency harmonics construction
well anyway i'm going to play with my analog moog synths i got two hardwares
Man those edits between your dialog are trippy. When you morf into yourself.
It's time travel. There's a flux capacitor just out of shot.
@@AudioMasterclass SuperCoool
This is fantastic! David's eye contact is on-point - amazed he didn't blink once the entire video.
I did all my blinking during the oscilloscope shots. DM
@@AudioMasterclass Blink twice if you were forced to listen to sine waves!
𝘾𝙤𝙣𝙜𝙧𝙖𝙩𝙯 oﾟ*｡o
／⌒ヽﾟ
∧∧ ／ヽ )｡*o
(･ω･)丿ﾞ￣￣゜
ノ/ / ｯﾊﾟ
ノ￣ゝ
@@chaoticsystem2211 🎉
The "ringing" is actually an artifact of creating lower pitch square waves by adding progressively higher and higher multiples of sine waves together, which is probably where the harmonics come from as well. It's something Fourier discovered while trying to explain the phenomena of how heat dissipates through different temperature rods before finally coming up with the now-famous Fourier transform.
indeed, it's called the Gibbs phenomenon and it is present in all waveforms that has a finite discontinuity (a jump in amplitude) like the square wave, but also the saw wave too.
he’s not wrong in sayings it’s because of the low pass filter. As you continue to add sine waves, it gets closer to a perfect square wave, so when a low pass filter is applied, it becomes less like a square as it loses the higher frequency sine waves, creating that “ringing” phenomenon.
The square wave isn’t being created by layering sine waves, but every sound can be expressed as many sine waves put together, and using filters, sine waves of different frequencies can be taken out
a speaker cannot produce square waves, because it would require the diaphragm to move infinitely fast or at least extremely fast compared to a sine wave.
but it isn't. because these waves aren't created using additive synthesis. ya dunce.
best part is seeing your voice in oscilloscope and watching bass overpower rest of the noise sibilants
got that podcast voice 🥰
Video: Can you hear the difference between sine and square waves?
Me who's spent months using Famitracker: *The Expert*
Eggspert
But Dat 4 bit quantized TRI channel is where the real FLAYVOR lives
but, there's no sines in 2A03... *huuuuummmm*
(yeah i know about the ex-chips, but i couldn't resist! Also /\ bass rules).
Excellent demonstration! Thank you.
You're welcome. DM
These videos would have been a lifesaver when I was first getting into synthesis and learning the essentials
Thank you for your comment. It would be interesting to make a video on basic synth waveforms, VCA, VCF etc so I might put that on my list. DM
Audio Masterclass that’s a great idea and would be very helpful for many people
This was really interesting. Thank you for the excellent and clear explanation
You're welcome
This is one of the best videos I have watched in the past year or so
Thank you. We have some new material in preparation. We hope you enjoy it as much.
I noticed that even at the highest frequencies it was still possible to detect a difference, the sine wave actually feels bigger to me when I hear it but the difference becomes is distinct the higher you go. the key to hearing it is, listen to the notes switch back and forth, one transition will feel like it's dropping or shrinking and be less pronounced (sine -> square) and the other will feel like it's the same note but expanded out a bit and with a more pronounced distinction as it transitions (square -> sine)
It might be because the switching wasn't exactly at the zero-crossing point because the zero-crossing point would most often be between samples. I might at some point repeat this test and possibly investigate further. DM
Very educational video! The explanation helped me understand why a sound I was producing sounded like it was ringing.
On a totally unrelated side note, one suggestion I would make would be to try blinking a few times a minute while recording. Not blinking can be a bit off-putting for a viewer, though they might not consciously process it.
As you can see, I'm not a natural presenter. And now I have to think about blinking as well... DM
@@AudioMasterclass Well I think you did great overall! That's just a suggestion for a slight course correction. I'm definitely not trying to trip you up. 😂
Never have I ever wanted to hear a function generator from youtube in my entire life.
*One video clip please*
You should probably avoid this link ua-cam.com/users/results?search_query=function+generator
I feel very smart cause I actually picked up on this happening on my own while making chiptune! The explanation however makes so much sense!! I just never considered it
You sound like paul mccartney
And look like Jeremy Clarkson
And he also has a bass lol
I was going in the comment section to comment that he reminds me of Paul Mccartney. Hahahah
To a Chinese person, maybe. But then, I can't tell the difference between someone from Hong Kong and someone else from Wuhan.
He really does.
Very professional demonstrations and video, and very mindful of the viewers' constraints! Much appreciated by a curious physics student.
Thank you. As a curious physics student you might also wonder whether the square wave is the only waveform I could have used for this test. DM
This was actually really fascinating and very well put together!
Amazed by the details given in a video of 9 mints. Appreciate the clear explanation on the topic. Thank you. I am getting into Vehicle Acoustics field and this would help me to look things differently.
One question related to content, I am looking for such detailed explanations on harmonics in Freq spectrum. It’d be great if I could get some information on that.
www.audiomasterclass.com/one/understanding-audio DM
Thank you so much for this video!
This video explains very well what I've noticed when messing around with my signal generators and audio amplifiers, I thought this was from me not having a musical ear or from me not constructing my signal generators well enough. 😄
Great video but I would avoid using morph cut, it looks a little strange.
We appreciate your comment and would add that jump cuts used to look strange until they become commonplace.
Very well explained, Sir!
Great video! I think what you call ringing on the square wave signal is Gibbs phenomenon
I think that my two maths A levels and one university module might be insufficient for my comment to be reliable, but it seems like the Gibbs phenomenon is due to summing the Fourier series only partially rather than letting it go to infinity, where the phenomenon would disappear. This is very similar to the ringing displayed here - If the frequency response of the signal generator (or the filter in the D to A convertor) were infinite, then there would be no ringing. The ringing occurs due to the limited frequency response because not all of the harmonics of the square wave are included. That's my guess anyway. If there are any maths or electronics experts out there who would like to add to this, or show me where I'm going wrong, please feel free to comment further. DM
@@AudioMasterclass I think that is more or less correct, any signal with some discountinuity will have this type of ringing due to the fact that you cannot take into account all the terms in the Fourier expansion. However interestingly enough the "ringing length" goes to zero as the number of terms is increased but the overshoot does not it will always be a little bit of. The Fourier series converges point wise but not uniformly in math lingo
I studied chemical engineering and NOT an expert at this realm, so I'd just point out my observation. From what I see on Wiki, Gibbs phenomenon looks symmetrical (i.e. post-ringing + pre-ringing). While the square wave shown here 3:21 displays an asymmetric gragh (only post-ringing). Although I do agree the principles between the two seem to be very similar. My guess is that perhaps they are related, but not completely the same?
@@owlmega-101 That's a good point. I'm not going to burn my brain thinking too much but I'm guessing that maths doesn't need to bother about time only going in one direction, whereas in analogue audio time definitely does only move to the right on the x-axis. That's just a thought so we await a real maths guru to illuminate the issue. I say 'analogue audio' because in digital audio, linear phase filters can definitely access the past, but only by delaying the present. ua-cam.com/video/UFrbKqRuKxI/v-deo.html DM
Very interesting that the 10kHz square wave showed multiple lines. That's because 10kHz doesn't evenly divide 96kHz, while 6kHz, 8kHz, and 12kHz do. So each cycle of the 10kHz square wave has 4.8 samples low and 4.8 samples high. But it can't subdivide samples, so it will be either 4 or 5 samples each time. The first integer multiple of 4.8 is 5, 4.8 x 5 = 24. So every 5 cycles of the square wave it completes a cycle and comes back to where it started. Looking closely, there are 5 individual lines.
I worked this out before you said you were using 96kHz. I figured you were using 48kHz, the math works out the same. You definitely weren't using 44.1kHz or 88.2kHz or anything else like that.
Thanks for this!
Well, In the beginning I was honestly listening with intent, but I had to start the vid over when I suddenly realized I was actually wondering when this man was actually going to blink, if ever!🤔 Haha...to be continued👍😉 (and I'm sorry oh blinkless one..but if theres any consolation, I did subscribe to you!)
Thank you. Good work.
You're welcome.
Why does the background follow your head at 1:52? It's an edit of some kind, I just can't picture what you were fixing!
It's a flow transition in Final Cut Pro. When it works it's good. When it doesn't... well you have seen already. I've given up on that as you can see in ua-cam.com/video/Ht3pMAkzums/v-deo.html DM
Great video and explanation. Thanks
Can you SEE the difference? I.e. in a flickering light? I find it easy to produce a square wave stimulus at a given frequency, but i find producing a sine wave at a given frequency hard.
It's interesting to notice that the square waves appear to begin and end exactly at the inflection points of the sin waves; that is, the individual square waves begin and end where the sin waves change from concave up to concave down. Although, I suppose this only makes sense... As an interesting side note, these points can be mathematically calculated by taking the second derivative of the sin function, setting it equal to 0, and solving for (in this case) 't', though a sin function will have infinitely many. Only the first couple points are necessary as the inflection points of sin, and thus, the structure of the square wave will repeat on a given period. Cool beans, thanks for the visualization.
That was interesting, thanks. Although, my cat wasn't pleased
Funny... Either the man does not Blink his eyes or we are blinking in sync, and can't tell when he does, lol...
After about 6kHz both of the signals were identical when hooked on my own scope :)
Funny to see we both use HAMEG scopes. I've the HM203-6 scope, however the line doesn't hit the edge of the display without using the x-mag button, stops about one div off each side.
very informative
haven't finished the video, but i'm going to make a prediction that the reason that they begin to sound the same is because the 1kHz square wave is just made up of a 1kHz sine wave with a bunch of higher frequencies added on. as the base wave gets higher, the higher frequencies eventually shift out of the range of human hearing.
My ex-husband did audio tech and used to put on this CD for checking your hearing at different frequencies but wouldn't tell me, just to see if i could hear it i guess, and i'd walk around the house trying to find what tv or computer monitor had been left on mute... the bright ping of those high frequencies can be inescapable!!! :O :)
You may have heard this too in the past www.google.com/search?q="hear"+tv+scan+line+frequency
thank you, dear.
Which is more efficient??
They didn’t sound identical at all to me. But I do hear higher frequencies than the majority. Interesting video!
@MF Nickster then I don’t know why they didn’t sound identical, but they didn’t.
Amigo acabo de encontrar tu canal y es genial, muy educativo y entretenido muchas gracias, saludos desde México 🇲🇽
This guys a natural electronic engineer but in rock n roll.
Very interesting video, thank you!
I am just curious about how does your digital audio workstation generate the square sound wave? What is the input electrical signal of it?
Both waveforms were created using the signal generator plug-in in Pro Tools.
I can’t hear the 12 KHz. Is that a significant clinical observation!
The ringing isn't caused by filtering. It's caused by a fundamental law of nature.
A "true" square wave requires an infinite number of harmonics.
Even if you were to create the square wave algorithmically through the equation sin(2* pi * freq * t/nyq) > 0 ? 1 : 0, to actually play it back the speaker driver would have to move at infinite speed at the square wave's edges and stop infinitely fast, which is of course impossible.
Thus a square wave not a band-limited signal and can't be recreated by sampling.
It's called "Gibb's Phenomenon," and it's the kind of thing you'd expect a "Masterclass" to know about...
It's also the reason things like the Heisenberg uncertainty principle and the Casimir effect exist...
Just have to put on my physicist hat here for a moment: The Ringing in the signal is what is known as Gibbs phenomenon. It is an intrinsic artifact that shows up when you do a finite fourier series decomposition of a discontinuous (ei it has a sudden "jump" from one amplitude to another) signal. No piece of electronic equipment can generate a "perfect" square wave, because that would require an infinite number of sine waves of increasing frequency and decreasing amplitude added together. We have to put in a frequency cut-off somewhere when creating an actual physical signal, and as you pointed out, the available frequency range of the playback device is usually the limiting factor. But as soon as we cut off our fourier series of the square wave, mr Gibbs pops up.
Thank you for your input, which I appreciate. There was some discussion of this earlier, if you look down the comments for SpyProductions. As for generating a perfect square wave, adding up sine waves would be a long-winded way of doing it and it's simpler to flip from low to high to low etc. That of course is also imperfect due to the rise time, but a decent function generator should be able to create a square wave that is good enough for any audio purpose. I doubt whether that would prevent Mr. Gibbs from popping up though. DM
Pls make video about Pulse wave
So is it true that pre/post ringing isn’t audible at all on any type of filter because it’s happening above what we can hear?
It is easy to explain, the smaller width and distortion at the top of the square wave make them more similar because the angle to go up and fall down of the sine wave is sharper. The sine wave starts to look more like a square wave at higher frequencies because the width is smaller and the top angle is sharper. Actually both start to look more like a sawtooth.
Nice bass
A square wave formula is A = sin a + 1/3 sin 3a + 1/5 sin 5a + 1/7 sin 7a and so on. The higher the tone, the harmonics start going out of your hearing spectrum. 4kHz x 3 = 12kHz so for mee that is the end at 52.
I wonder how would a constant sound like? Something like y=10. I guess we won't hear anything because there's no frequency.
If you add a constant to an audio signal you get a DC offset. That will cause switching clicks and in the (probably) worst case could bias a loudspeaker cone causing distortion on positive or negative peaks, depending on the sign of the constant. That's my first thought off the top of my head but I'm sure there's more to it. DM
But where do the harmonics come from? And shouldn't they be visible on the scope? Or maybe it's zoomed in too much to see them? Great video, thanks!
You would need a mathematician to explain where they come from - try en.wikipedia.org/wiki/Square_waveen.wikipedia.org/wiki/Square_wave - All I know is that they are there. You can't see them on an oscilloscope but a spectrograph will show them clearly (in a square wave). DM
@@AudioMasterclass You could also plot an FFT of the waveform. I don't miss calculating those by hand.
If you look into the way Fourier came up with his Fourier transform, it suddenly becomes clear. Square waves can be made by adding together progressively higher and higher multiples of sine waves. 3 blue 1 brown has a fantastic video on it. The ringing he pointed to near the end is actually an artifact of that process.
Your filter which removed high frequencies is removing some of the terms of the Fourier series which comprises the square wave. This effect is more noticeable for higher frequency waves, as more of the terms are removed. This is visible in the CRO. So part of the reason for them sounding more alike is that the resultant wave forms are more alike.
(and it seems my sound card stops at 12 kHz! Or at least I hope it's my sound card & not my ears!)
I'm going to politely disagree with you. This informal test works just as well in the analogue domain where the square wave can be clean up to and beyond 20 kHz, so it's what you can and cannot hear that makes the square wave sound like a sine wave at higher frequencies. The filtering visible in the oscilloscope is that due to the low-pass filter in the digital-to-analogue converter through which the digitally-generated sine and square waves are output. When an analogue version of the internet is available I'll be pleased to repeat the demonstration. Thank you for taking the time to comment, and P.S. It might be your ears! DM
I feel its aberration from filter
Excellent! Thank you very much!
You're welcome. DM
the difference is that a sine wave is sine and a square wave is square
@Audio Masterclass I have a question.
Why does low hertz sound low.
What I mean is - in my phone i have a frequency generator app, keeping the volume of the phone constant. When i increase hz from 20hz to 2k hz why does the sound increase?
1) The frequency response of your ears. 2) The ability of your phone to output low frequencies. DM
@@AudioMasterclass one last question, if hit the fork hard or if i hit the fork softly.
In both the scenarios will the fork vibrate at same frequency?
It is normal if i can hear the difference between the sine and the square wave at 12khz ?
Do you think our body, or electrical body, (aura?) acts as a filter to all the frequencies surrounding us? That frequency seems to be like tinnitus. Because if you listen, you can hear all that static in your ears, and perhaps illnesses get in by a dissonant tune that wasn't filtered properly. What do you think?
These are topics that I'm not qualified to comment on. But if tinnitus is an issue for you then you should consider consulting an audiologist. DM
Hi,
Thanks for the knowledge you share with us.
I have an off-topic question: I'm interested about the famous $5 preamp, is there a final schematic as well as a list of all the materials?
Also, I see that ebay has many types of INA217 and OPA 2137, what is suitable for this project?
Sorry for the off topic and thanks for your patience.
Thank you for your comment. The full schematic of the $5 preamp is available at resources.audiomasterclass.org/images/1237/ina217.pdf
Thank you very much !
I personally love square and sawtooth waves, since they are fun and you can hear the tone clearly (due to harmonics and overtones). But, I hate sine wave, since it is plain boring + at lower frequencies, you have to turn volume up in order to hear the tone (if the frequency of the tone is 27,5 Hz, the sound pressure has to be around 70 dB for sine tone, which means that you have to turn volume up (and with that, you're risking to damage your hearing)). But with the sawtooth or square waves, you can clearly and without turning volume up hear 27,5 Hz sound. Sine wave is boring, bland and simplistic, it doesn't have any variety. Sawtooth and square waves are more interesting and fun.
At higher frequency square waves started to shape like sine waves. Very nice demo.
The shape is due to the low-pass filter in the digital-to-analogue converter which is inevitable in digital audio. If I do this test again, I might use an analogue function generator which will show much cleaner square waves on the oscilloscope. DM
Very nice video.
Thank you. DM
Interesting stuff
This is how I know I am ready to turn 40, when I bought fancy headphones, started worrying about what my amps are doing, and an oscilloscope and some webpages about measuring output impedance later, here I am, worrying about how much the amps are ringing.
The ringing you see here is way above the range of human hearing, and when you do turn 40 your high-end will decline anyway. So, nothing to worry about! DM
Interesting no audio at or above 8 Khz Time to throw the test CD into the removable drive. I can still hear the 15,750 on my old kitchen TV.
2:10 Yes, Please
I think the last time I checked my hearing range last year, my limit was around 15 to 16KHz. Not bad for a 47 year old I guess?
Well done. If you create a sine wave in your DAW that is as high as you can hear and save it, you can check back again in a year's time and see how things are going. Let us know how you get on :-)
@@AudioMasterclass I put it down to wearing ear protection at live concerts, even when this was considered a very strange thing to do back in the 1990s and 2000s. The last ten years I have begun to see other people wearing them though. I also do not listen to music with exaggerated bass frequencies, e.g. club music, EDM, Hip Hop / Rap etc.
Recently I had my music in the car a bit too loud in my opinion, but when I got out and shut the door I could barely hear anything was playing at all. Contrast that with other people, where you can hear and feel their car coming from hundreds of yards away. So the extreme volume and decibel levels inside the car, must be doing them incredible damage. It's so loud I would possibly physically vomit; would not like to test this theory out though!
I've also always monitored my sessions and mixes at low to moderate volumes, turning up only to test briefly then back down again. I often hear things in mixes and turn to people around me to ask them about it, but they always shrug their shoulders as they cannot hear it. Even when I move my hand in time with the thing I want them to hear, they still have no idea.
mine is about 15hz to 19000hz
1000 Hz (censored sound)
This is very helpful but I can not find a reference to a very obvios point. If people are listenting to this on typical laptops peakers then the opportunity to hear what is being presented is going to be severely hamperedd by the roll off at higher frequencies.
People are goong to be biased towards hearing no difference between different waveforms due to this and at higher levels
there will be no sound at all in some cases.
I suspect that even basic smartphones have a better presentation at higher frequences than many brand name laptops.
You're exactly right of course. I used to warn against listening on laptop speakers but I came to a point a while ago where I thought that anyone watching a pro audio channel would know this. But I might go back to the warning again. DM
tq for sharing
2:17
I thought of "DX entrance music"
this man never blinks
That's correct. Now if only you were the first person to say that...
Cool, thx my British friend.
You're welcome, typed in a vaguely northern English accent. DM
I weirdly could not hear the 10 kHz at all, but I COULD hear the 12 kHz?
This could be an issue with UA-cam's encoding, or with your computer's audio output, or your audio interface. Maybe also your headphones/loudspeakers. Other than that, an audiologist might be able to advise further.
Likely a low frequency artifact like aliasing caused by the codec or your hardware.
could just be the acoustics of your computer speaker more than likely!
try moving your head around ..
@@adhanda2017 How would that help in headphones?
So audio sampling at rates greater than 44.1 kHz is pointless...
How dose one figure this out??
i didn't heat anything in 6khz but i did on 8khz , i think it matches my tinnitus..
I stop hearing a difference at 6khz, which matches the frequency I go deaf, 18khz.
Well spotted. Exactly the way it should be.
The same here... Though my hearing is a little less than 18kHz (I'm pretty sure I can hear a 17000Hz sound wave)
How about a sawtooth wave?
A sawtooth wave sounds more similar to a sine wave than a square wave does but at lower frequencies the harmonics will be audible. For a 5 kHz square wave, the next harmonic after the fundamental is 15 kHz and is probably inaudible to most people, so the square wave will sound like the sine wave. A sawtooth wave at 5 kHz will have a harmonic at 10 kHz, which probably is audible. So at that frequency and lower the sawtooth wave will sound different to the sine wave. At 7.5 kHz, all three will sound the same, assuming the listener can't hear 15 kHz and above. That's my long answer to your short question, but the best thing to do will be to try it for yourself. Bear in mind that any distortion in your listening system may confuse the issue. DM
The 12khz sound the same, as is not heard. At 57 I do not think I can hear much above 8Khz
Does this apply to other timbres other than square?
The triangle wave also has only odd-numbered harmonics, so the test would be similar. DM
Very interesting. It's the same with tempo and rhythm in some ways. (I.G.) As tempo increases for a jazz ride cymbal pattern based on the the triplet, it 'flattens out' and becomes indistinguishable from sixteenth notes. (Basically, as the gaps in time shorten, there isn't enough space to create the triplet that would be perceived by the ear.) This is same experience here except... we have frequencies which obviously get faster as you go higher and also 'flatten out'. Or sound more 'rounded out'. It's kind of like a musician's version of relativity. :-P
Agreed. It's also interesting with an analogue oscillator to set a square wave to around 1 Hz, which will sound like clicks. Then increase the frequency slowly. The clicks get faster, then at some point above 20 Hz they will start to sound like a musical tone and the individual clicks will disappear. DM
super stuff
Thank you. DM
i realised to late that those high frequencies kind of triggered my tinnitus, be careful if you read this and you also have tinnitus
So does this mean in a PEMF mat that has sine wave only that by lowering your frequency from 3 gauss or 1-30 frequency to the lowest setting you can get a square wave?
It would be the other way around. If your mat uses a square wave and you can set the frequency high enough it will feel like a sine wave. That's my hypothesis, but it would need an expert in this kind of therapy to advise reliably. DM
additive synthesis be like "bruh?"
please i have an experiment of effect of hz in water what i need to use ?
This would be interesting but we have no expertise in sound travelling in water so it would be best to enquire elsewhere.
Video: Can you hear the sine waves?
Me: Sine From Above-Lady Gaga
Sorry..hahaha *Last Song Syndrome
Can you hear the difference between sinewave and squarewave?
YES! THEY SOUND COMPLETELY DIFFERENT! BRUH WHAT KIND OF QUESTION IS THAT!?
This is why you shouldn't comment before watching the whole video.
@@AudioMasterclass oh.
What dose it sound like on the Htz scale
This video - ua-cam.com/video/vge0GmVqUXg/v-deo.html - might answer your question.
"high-pass filter of the audio interface"???? I think it is low-pass filter (antialiasing filter of DA converter) .
Oh now I get it, to get rid of DC unit (voltage), means frequencies below 10 or 20 Hz.
Did I say that? Yes of course it is low pass. DM
At 100hz I can hear the square wave but not the sine wave. Is it my speakers or my ears?
You're hearing the harmonics of the sine wave. DM
0:22 the sound exactly the same what do you mean they sound complete different
If the difference at 0:22 isn't extremely obvious then there must be a limitation of some kind in your playback system, probably a poor high frequency response. Also, since the high frequency response of the ear diminishes with age or other aural issues, then perhaps your hearing is the issue. We are not medical professionals so we cannot give advice on this, but it might be worth consulting an audiologist.
I can hear the difference up to 6 kHz. But this experiment is flawed, since your "square wave" is not really a square, which can even be seen clearly on your oscilloscope: due to filtering, most of the higher harmonics get filtered out and what's left is just a couple of regular sine waves mixed together. That's why they sound so similar.
The video covers this issue but I'll explain briefly. The waveforms are digitally generated and are output through an audio interface. This filters out frequencies above 20 kHz or so. This is why the square wave isn't exactly square. I could have used a function generator and the square wave would look square, but *your* audio output filters the high frequencies too, so what you hear wouldn't correspond with what you see. The way I did it here, what you see corresponds closely to what you hear. The similarity between the sine wave and square wave above 6 kHz or so is due to the harmonics being higher than most humans can hear, and above 7 kHz they are filtered out in the D-to-A convertor, and even if they were not filtered, you wouldn't be able to hear them. DM
The problem is that it's not a perfect square wave, there's some sine element to it and you can even hear it more and more as the frequency increases.
Why did you skip 7khz?
In the interest of not making the video too long, some frequencies were skipped.
1:49 - the background is cursed!
i'm done at 8kh, impossible to "see" the difference for me, what about you?
While no-one at Audio Masterclass is an audiologist, our informal response is that the natural loss of high-frequency perception is part of the ageing process and does not seem to affect people's enjoyment of sound. However it may be the case that there is high-frequency content in your recording that you can't hear that younger people might find unbalanced, or even unpleasant. It would be wise therefore to get the opinion of someone who can hear frequencies at least up to 16 kHz or so before releasing your work to the public.
@@AudioMasterclass thanks, I never tought about it before, I used to make older people ( parents) listen to pieces of work of mine before the releasing to make sure the sound was balanced from 2K to 10K, it's like using a natural bypass, thanks for your answer, I really like your videos, cheers from Algiers!
My ear aw!
Is this what dog hear?
there is a rumour about using modified sine wave to audio stuff like powered speaker will damage the device is it true?
In a speaker with a woofer and a tweeter, the woofer handles most of the power. That's normal for speech and music. If you drive the speaker hard with a sine wave above the crossover frequency, or a square wave which is rich in harmonics, you could send the tweeter more power than it is designed for. The voice coil will partially melt, breaking the circuit, and there will be no output.
@@AudioMasterclass does this also apply/happen with line interactive UPS which is using modefied sine wave inverter?
We don't have experience of this, but running audio equipment from a modified sine wave inverter doesn't seem like a good idea. To be sure, you might consider contacting the manufacturers of your equipment and see what they say. There's an interesting thread at www.reddit.com/r/audioengineering/comments/1m5c6t/pure_sine_wave_vs_modified_sine_wave/