Maybe the best part of the day is I get to share my my hobby I guess my passion with of a fairly full room full of others who I hope have at least share part of my passion for music music instruments and also especially the piano so what I want to do today is are sort of two things first of all I want to talk to you about some of the some of the history of the piano some of the ins and outs of how pianos function how we listen to them and so on I also want to to criticize physicists which I don't and most never do but but I was as you'll see if I go through this talk some of our some of the paradigms we use in physics about simplifying things and treating things you know there's this famous quote of Einstein make every problem as simple as possible but no simpler Well sometimes physicists go a little overboard in that especially with regards as we'll see to a musical instrument like the piano so that would be kind of an underlying theme which I think which you'll see come back several times through this talk so what I'm going to do is begin by giving you a very general interruption to the piano. It's overall structure all remind you of how the strings in the soundboard in the hammers and so on are all connected to make this this instrument the piano is in some ways unique in that we know it's exactly when it was invented and we know why it was invented unlike instruments like well the organ or the or the violin where we don't quite know who started it first and and some of it for the piano we know exactly who start who who made the first pianos and we know why you made them and that's an important part of the story on talk about then I'll move on and talk about some interesting aspects of the of piano acoustics and here's where I'll show you it's not just the simple physics that we learn in physics one simple physics and get us a long way but it can't get us all the way to a real musical instrument the one that we're used to listening to that a piano that was only based on the simple physics wouldn't be a satisfying piano but we'll see how to make a satisfying piano as we go so some of those aspects include in harmony. This is a piano with tones they are quite harmonic as we're used to learning in physics one will talk about how piano tone changes with loudness and why it's extremely important for the instrument also talk about the variations of how piano tones vary with time how that how their time decay how it behaves and why it's interesting and a little bit about bass notes and a psychoacoustic issue known as the missing fundamental something it was first noticed by a physicist called home health so you may have heard of and also a talk in general about the importance of psychoacoustics the way we listen to music in general affects the way the piano was structured and is is put together and they're not I've run out of time I want to talk too much about my own work in modeling of various parts of the piano and then I'll wrap up by reminding you the main message is what we really learned from this kind of work. So this is a picture of a typical grand piano this one sits in my living room it's this it's a it's a modern well modern It's about one hundred year old Steinway but it's and for the purposes of this talk I'll show lots of photos of this of this instrument these photos are in my my book I think this one was lifted to the for the poster and I should give my daughter credit she's a photographer and she took took all the most all the photos here she did a wonderful job so there's a typical modern grand piano. You know the player sits stands here at the front and puts on the key person on the keys the strings run from front. Front to back and we'll talk more about that later and the hammers if you look on inside the hammers you push on these key levers in the front and the hammers come up at the strings from below and then fall back so the strings get hit from below and they vibrate in a horizontal plane that runs away from the player and that's the general actually probably all knew that some three of all taken apart your piano at one time or another. Now that's a modern grand piano it has about eighty eight keys it notes this is the is the oldest extent piano. Made around seven hundred twenty it was made in Florence Italy it's interesting that this that crystal this is Bartolo male Christopher he's the inventor of this instrument of this piano of this instrument it's interesting that is his he lives he overlapped with a violin maker named Stradivari who you may have heard of in Florence isn't too far from Stradivarius hometown and also in northern Italy and so this was a must have been a really fascinating and very very exciting place still to live at that time especially culturally Now this is the only known image of Christa for in fact this image now this this painting no longer exists it was destroyed in World War two and bombing of of Germany but the image persists and the reason I showed it here is that Kristof Henri is holding in his hands if you blow this up you can see this piece of paper here he's holding a this is a diagram of a key part of the piano now this is his first piano I mentioned the piano has eighty eight keys this one has forty nine OK on almost half the number and it's really has a lot in common with the harpsichord But but mainly it but it's a it's a very different road different from in the today but the main the key part of this instrument is is a. Mechanical piece that links the hammer the key levers to the hammers as well see an important part of this is that when you push on the piano key you start the hammer in motion and just before the hammer hits the strings the key lever escape for the hammer escape from the key lever is called the skate Mint So the hammer is moving freely as his the string is able to bounce back freely that's a kick and absolutely crucial part of the piano This is the diagram of the mechanical lever system that Christopher invented the fact that it's in this portrait shows that he knew exactly how important that was for the instrument. And I'll come back to that several times so let's talk let's talk about why why did Krista for you and vent this piano Why do you get all the goal is a bigger speak Italian. OK So you know what piano means piano piano and in Italian mean soft and in fact the early pianos were actually call piano forte That was one of the terms Krista for use for his instrument so piano means soft and Forte means loud you know from your musical training so it's it's a soft power but it's an instrument that can play loud and soft and now as you may not know that at the time of the invention of the piano the main a keyboard instruments or the organ pipe organ existed and the and the harpsichord and both of those instruments have a property and that if you press the key a little bit or a lot you get the same amount of volume the tone is the same whether you press it hard or soft you you you can't have you can't get expressive possibilities of playing loud and saw from note to note and that's what people want to that's what Kristof or you want that's what people like later a few years later Mozart wanted they wanted they want the expressive capabilities that come from being able to change your volume from literally from Note the note I mean with organs you play a little bit and you can reach in and you can pull lever maybe on some organs and you can then get louder or change and it's change the sound of things but you can't change it from note to note with the piano you can because of this escapement mechanism that Christopher invented for me he added the idea of a hammer to strike the strings and that came about I think at the time from people playing with dulcimer so people knew about hammers hitting strings from dulcimers but they didn't know about the idea of using the key as a lever through some kind of link is called the action which I didn't really draw here that to get the hammer into motion but then the key falls away from the hammer the action falls with a hammer just before it's a string so the hammer is moving freely if you push the key lever hard it's going fast because a little bit is going soft that's that's the that's the crucial part of the of Christopher or his invention and as I said he knew how important this was. So the rest of this drawing shows you. Very simplified physics a physicists conception of a piano is only a single string of course there are many more than one in the piano but the idea is that you have a string stretched between two and two held rigidly at two ends you hit it with a hammer excite waves on the string we know about standing waves from our physics introductory physics courses and those standing waves cause a force on this bridge which is has to a sound board that sets the sound board into motion that acts like a large speaker and that produces a sound that we hear a vibrating string by itself produces very little sound because it moves very little air the sound board is much larger it's it's it's a metre or two or three and across that moves a lot of air and you know big speakers can make a lot of sound. OK so this seems to involve only very simple physics but let's let's go on and see what see what's more here. OK So let's compare Christopher piano to the modern piano Steinway but any modern piano would would do. There's a difference in technology of the strings we now have steel steel strings where they weren't invented until about eight hundred fifty so they weren't available to Christa for in seven hundred twenty he used mainly brass strings some pianos a time use iron but they were the they're much different in composition and much different in strings so the pianos in his day were much lower tensions almost an order of magnitude lower tension than the modern piano. Or. The modern hammers are covered with felt I thought more about that later his were different focus on one thing here as I were mentioned he is painted forty nine notes and modern ones have eighty eight so the lowest note nowadays on a piano goes down to it's called it's a zero it's the lowest key on the piano the far left and that's got a low frequency of the fundamental of twenty seven point five hertz the highest note the far right is a C. and that has a frequency of a little over four killers now want to quit so obviously the piano involved from four octaves to almost. Eight Why did it happen well it happened because people like Mozart whose pianos had five octaves and then people like Beethoven who had six and then Brahms who had seven and they wanted more and more know if they wanted to be able to do more so the question you might ask is Why do we stop that's at seven in a few octave bit octaves when I have more we can put I mean I can reach farther why not well the answer in this case has to do with the way we listen to musical tones it turns out that if I were to I could easily make a tone with a much lower frequency than this I could make a tone that's ten hearse but when you listen to that ten hertz tone you would not hear a musical tone but you hear a series of clicks So basically it's somewhere around twenty five heard sort of so is the lower limit of what we hear is a tone as opposed to a click right if you if with a ten hertz click it would be you know clicking on off it to attend to ten or ten clicks per second it would sound like a musical tone so that wouldn't be very useful to Mozart or Brahms or anyone or even John Cage probably OK. So that explains why the piano range stopped expense extend extending to below this now somebody in the audience will be a smart alec and say well what about a bosun door for if they have extra keys at the lower end well those keys really aren't used very much and I'll come back to what why they why they might why they I think they are they're a good thing a little later so whole lot of that question I'll get that I'll get back to that now why about the highest No Why not go to a higher frequency this as you may know a young person like most of the people in the audience can hear about twenty kill up about twenty killer it's OK I probably can I don't know what my range is now is I'm sure it's much less than that if you get older you know your ears can't get your range up or a part of range goes away or goes down well the reason for that is even for a young person who can hear twenty killer it's the upper limit of what tones you can hear as a musical tone ends around five kilohertz or so that is if you probably know or if we're talking more. One of the nice things about musical tones if you put two together to play an interval or maybe you play a chord and they sound good together because of the way chords and intervals are are constructed frequency wise relative to each other. But if I if I were to so if I were to play say middle season is two hundred sixty two Hertz and I were to play say a G. above that or see above that I put them together it sounds good but if I were to play this C. and then play the next G. above that you wouldn't be able to hear them in a musical or you would be able to hear their musical relationship that is you could tell that one note is higher than the other it's wise but you wouldn't be able to tell if they were nice interval or not a nice interval your ability to perceive hitch in a musical way ends a little above this frequency so again notes above this I could easily make I could put them on a piano but they wouldn't be musically useful so the way we listen to sound already has influenced how a piano has been constructed in design so this and this has a we'll see another example later of how the human to the way humans listen to sounds affects the way we perceive the sounds from a piano and other instruments. OK So what about let's get back to the physics. That this picture is a. Figure but the the strings on the far right the high notes the strings here are very short as you probably know the low end strings the bass things are much longer and now. If I suppose I want to take take this I have a string and I want to then make it design another string to produce a note an octave an octave lower so I'm going to go from say the C. right right here to a C. an octave lower Well one way to do that would be to keep the string composition the same keep the tension the same and simply make the string twice as long as you may know again from your physics that if I double the length of the string I double the wave will I double the wavelength. Which cuts the frequency in half so I made an octave after two difference in the pitch so if I if I keep doing that I think I can go from one C. to octave lower to an octave lower octal I keep reducing the frequency by a factor of two and this curve is just a just a one over frequency curve and that's called with Egger Rian scaling you may or may not know at the Thag or is credited with having designed musical scales long ago as a sympathetic who had in his spare time talked about right triangles but he also as is sort of referred to as the father of musical scales too. But if I follow that kind of scaling for a piano but I get down to the bottom end I would be something like six meters long that would be an awfully long piano it wouldn't it wouldn't I wouldn't fit in my living room at least not with my wife's consent. So what do we do well it this this dash curve shows the actual string the links for this particular piano so we follow this this but faggotry a curve pretty well just a little bit below middle C. S. C. four and then it be sort of saturate So what's what's going on here why what how power you able to saturate How can these string which isn't which is far shorter than you might have expected produce such a low a low note. And the answer is that another way to make a string vibrate at a lower frequency is to make it heavier to increase its mass per unit length and the way and so one way to do that would be just make it thicker I could take my piano string around middle C. the piano strings are about a millimeter in diameter so there are some with thin piece of steel but if I make it thicker I can make it vibrated the lower frequency which you think would be a way to make a low note the trouble is if I get down here I have to make the steel string so thick that it wouldn't be a string anymore it be like a vibrator bar and that would not make a good piano so and in physics terms or in mechanical engineering terms there be too much stiffness. In that string that bar so what piano makers do in fact this was invented before pianos were invented for other instruments is they take a steel core of a string and they wind copper around it so thick imagine a steel course of a say a millimeter to a diameter steel string with a slinky around around it. So that Slinky gives it extra mass so it vibrates slower by Brits at a lower frequency but as you know from sleep these are very flexible so you're at you've added mass without increasing the stiffness very much These are called wound strings and this is the way all all based rings and all pianos are made and two with and some of them actually have either two or even three windings around them to give you extra mass without without giving you too much extra stiffness and that's how you that's how we violate the thuggery and scaling. On a piano OK. Now. I mention So we need to bond strings to increase the mass per unit links but we want to do that in a way that minimizes the string stiffness does the physics talk so I get to show some equations here and what I'll show you is that. It for a perfectly flexible string this physicist ideal the fundamental frequency is some what is some value set by the tension other things in slings but it will also vibrate at other higher frequencies called harmonics and they they the frequency is a harmonics for follow a harmonic sequence where just the frequency of the ends harmonic is N. times the frequency of the fundamental OK So that's that's so that's what perfect strings would do now if I have a stiff string you are you could not well physicist can look there's a new term that comes on here that's a fourth derivative if you're not a physicist worry about it the idea is this if I have for a for a very very flexible string the only thing that when I say I pluck it or I get it into motion the thing that brings it back is the tension in the ring the tension. If you the restoring force that causes the wave motion now if I have any real string has a little bit of stiffness right if I bend a piece of steel it springs back by itself so that extra springing back forces a stiffness force and that will act to increase the frequency of vibration now it turns out that it will increase the frequency of my brace in more for the higher harmonics and for the lower harmonics basically because the higher harmonics have a shorter wavelength so they're bent more OK So what happens is the uke instead of getting this pure harmonic relationship you get a little D.V. a deviation from it that depends upon the harmonic number so the second harmonic is two times F one plus a little bit and the third harmonic is three F one plus a little bit more and so on. And these are and so because they are harmonic anymore they're actually you're actually called partials so if you have a conversation with your piano technician she will refer to them as as partials if you say can my course. So this is a this is a measurement of the IF of how big that effect is for my piano my my my grand piano this is for the note A three is the A below a four forty and and remember for a perfectly harmonic set of frequencies it would be F two F three A four if so what I've done is I've taken the measured frequencies and divide them by N. so if it was perfectly harmonic they would all fall into twenty so the first harmonic is to twenty divided by one the second harmonic is is about four forty divided by two brings me back and so on so if this was a harmonics a truly harmonic sequence every every dot here would fall on the dotted line a dash line but you see they don't they go up higher up more and more and more that's the effect of string stiffness. OK And so this big this is typically a few hertz for the the fifth or the sixth or whatever partial of this of this note now another way to measure that or display it is. In terms of a unit called called since and this is not S. E N S E as in your physics course but C. and TS It's a logarithmic unit the important part about it is that one hundred cents is the distance from one note to the next note above it Lemon are no matter where you are in the musical scale so so just so this what this says is that that say the six harmonic is about ten cents too high or about one tenth of the way to the next note it gives you a sort of a scale so it's not huge but it's but it's quite noticeable in your perception if you can you can construct either pianos or you can start artificial tones that have no stiffness and compare them to one with this and you can tell the difference and judging from the results of listening tests humans prefer a little bit of stiffness it gives it makes of the note A little more interesting in the in the way that a piano tone is different in I think merges in say the tone from the pipe organ a little different tone call up in Tampa temper. Now this also besides affecting the Tambour of a single note this also affects the way pianos are tuned this is an effect called octave stretching and here's what happens. When when your piano tuner gets to your piano she tunes to say a for a say a two to a to B. a certain frequency and then she wants the next to the note the next octave up the sound good when those two notes are played together so what she will do mine was a she is a she. Is she will to adjust the note the a an octave higher so that its fundamental coincides with the second harmonic of the lower frequency If so I have a it say to twenty it has a second harmonic it's a little above for forty so she will take a four forty and make it a little bit higher than for forty otherwise it wouldn't sound good together so what she's actually done is made an octave which should be. Factor of two in Picher frequency a little more than a factor of two. This is a octave stretching and every piano is tuned in this way whether you knew it or not OK an octave is not an octave OK. And this is a octave stretching it was first discussed a lot by a guy named rails back seventy eighty years ago OK and what he did is he measured the tuning of lots of pianos and then he comes and he found they follow us on this on the smooth curve so all I've done is draw in the smooth curve so this shows you that then what I've done here is I plotted the tuning of a typical piano across the keyboard range from the low end up to the high end and so over here the scale is in cents so this means that the lowest note on the piano is about forty cents lower than it would have been for a perfectly harmonically tuned piano and the high end is about twenty or twenty five cents higher than it would have been OK so in the middle things are pretty good not too much stretching but there's a lot of stretching noticeable at the high end a lot at the low end this is now there's no. I don't know of anybody who actually calculated this curve. But what this curve is is for typical piano following this tuning curve makes the best sound best is a subjective feeling OK but if you so but this is the way your piano at home is probably tune and in fact if you if you've ever seen your piano tuner or piano she comes with a little tuning box very very smart now and she'll punch in or you have a Steinway M. She punches that in and that tells you what curve to follow for the Steinway M. If you have a bosun door for She'll pick to put the Bosenova button on your pretty smart these engineers right. OK So this is this is a this is an effect off of stretching it means that the octaves really are octaves and this tells you. That this is a this if there's one more thing I should mention. Just because your piano tuner makes. The fundamental of the upper note mass the second harmonic of the lower note there's no way she can make the higher harmonics all match OK So she's she's what she's doing is a grand compromise. That's why I said that's why I use the word best the way I meant to use it less second ago there's no ideal what there's no way to make all the octo the harmonics of one note coincide with all the harmonics or the partials of the note below it nevermind making chords and everything else OK so it's a great again a grand compromise and this means that no chords will be perfect no no of course no intervals are perfect No already no course will be perfect but this this curve is I could say the grand compromise of how you tune a piano now somebody might ask me wait a second I say what about violins or play if I'm playing my job all along with the what the piano how come my oboe can be perfectly harmonic and how can a play with something that's so terrible as a piano and I don't really know the answer to that. But I think I know I think the answer is this most musical instruments like your oboe or your singer or your clarinet or your trumpet whatever they only cover about two or two and a half octaves and once you get way into the bass over any two octave range things aren't so different than being flat and I So I think this is how you get away with it not being offended by the note of when an oboe plays with the with the piano but I don't know for sure I never I've never done the experiments myself. OK Let's talk a little bit about let's move away from strings and start talking about other things like the like the hammer mechanism I've already mentioned this is my again my physicist conception of the escapement I just wrote a little arrow here I don't want to the whole thing is too complicated picture in a second but the idea is that we push the key lever that pushes up on the hammer the hammer is moving up to the string hits this it comes. It arthritis skate from the action that's why there's a gap here and it hits the string it forms a little wavelet or a little poles and that Paul's the hammer falls away and the pulses actually two pulses travel in opposite directions OK and then they go back and forth along the string OK Now what's interesting is that. Well this picture actually is not quite correct because this Paul's here this was the hammer hits the string quite pretty close to one end and so the this pulse the one that goes to the the play at the close of the player will actually return first and it will return before the hammer is actually falling away so it's not just a really quick pulse but it's a longer post appeared that time for pulses to return Here's a picture of the action this is not a very this is hard to see from this photograph all I want to impress upon you is that it's very complicated it's the most complicated lever I think any physicist has ever seen OK or every And if it was ever wants to see OK but it and it's I don't know how to describe it other than if I show you a video but I don't have it here but I would mention that the piano action was that we have today was designed to one hundred fifty years ago or so and I think there are more patents for how to make a lever like this than any other patent with the piano OK or just there are just hundreds of pads like everything they can make a better lever. Now piano hammers. This is the picture of two hammers this one's attached to the to the hammer shank this one's at the she has removed these are there's a scale it's about two inches there and this these are basically a. Wooden. Felt cover wouldn't mallets OK The core is made of I think it's this looks like and usually the shank is made of maple and this is felt as it's wrapped very tightly around here you see the this is the trouble hammered smaller and weighs a lot. A few grams and this is a bass string hammer weighs about ten grams it is much bigger and what the way these are made is you you get a piece of wood made in the shape of the hammer head it's like it's like a long piece and you wrap the felt around it and you could like a jelly roll OK so you get one it's sort of a match set now Christof or it didn't have the technology to make felt this way we felt was knowing that you could make it in a reproducible way his hammers are made very differently soon after Krista fori people started making hammers with with leather coverings and so early ones were made with leather until the felt technology caught up and I have a lot of pianos with leather hammers and it's what's remarkable is a leather hammer sound awful lot like felt hammers OK and I could talk more about that I can talk about all these things for much longer if you want to hear it and I can talk more about that later if you want to when I know. So that's what I am you know hammer looks like and you've probably seen that if you look inside your piano. Now let's talk about a little bit about piano hammers and piano strings and this this title should really be how should a piano hammer be modeled not string. Now it's physics we know about well the piano the piano hammers this is a bury bad sketch of a piano hammer the idea is this is the hammer this is the head of the hammer hitting a string and the hammer hammer it comes out hits the string and the hammer the fell compresses a little bit hits the string then bounces back now as physicists were all taught it's ingrained in our in our psyche that everything that compresses is a is a is a spring is a linear Spring I mean that's what I mean by every course we take it's linear Springs OK So our first reaction is to model this or mathematically as a linear spring means that the force from the hammer is proport directly proportional to the compression Z. That's what I mean by linear spring in those you take a physics course as you probably have nightmares about this to. Now if I were a PA If I were to try to model a real piano. As a linear spring it would fail miserably it doesn't it's not linear at all in fact if it was a linear spring it would be a disaster for the piano it be absolute disaster so let me first tell you what they are they behave and I'll explain why it would be a disaster here is a plot isn't something we measured in my lab measure the hammer force as a function of the compression and a typical For typical hammer blow the compression is maybe half a millimeter maybe a little bit more and what you see is not linear at all it's got a lot of curvature and that curvature is described approximately by about a cubic law in some hammers that is into the two point five summers into the force is no there's no fundamental number there but it's a lot bigger than one and what that means is that at the higher compressions the hammer is effectively stiffer than at the low compression towards effectively soft OK so if I play a note with a of a very gentle key press and only press the key to get the hammer going to little bit I'll only use the soft part of the hammer characteristic If I press hard play a loud note I'll get the loud the the harder part of the characteristic and if you think about it that's exactly what we want for musical instrument. I said before that this is the pianoforte it means soft loud I really meant more than that. When I play a loud note I want the tone color to change size and not just the volume I mean think about it you could you could be playing music of music in your on your in your car off your M P three player whatever and you can turn you can built into piano music and you can turn the volume way down you can still tell when the piano is being played loudly as opposed to being played soft if you hear the change in tone color or the way I think about it if if I'm playing the piano in one room my wife was saying in the basement she can still tell when I'm playing loud or soft not from the volume but from the tone color same thing is true for other instruments think about SAIC. Clarinet or saxophone or if I play a saxophone loudly the tone color changes or trumpet right all those things the to me the key part about the expressiveness of an instrument is that the tone color changes when you change volume it's not just a change in vine that makes it expressive it's a change in tone color. And that's what exactly what these non-linear hammers do for the piano effect people with the Kristof Henri's ambers they are non-linear with just about the same exponent as modern am and this is why a leather hammer sounds us about the same as a felt hammer they have about the same exponent there and this non-linear area central so that here's my here's my knock I have how I forgot the precise business of going on this is my knock on physicists we are so obsessed with linear things that here the non-linearity is an absolutely essential part of the instrument it wouldn't be the same without the non-linearity OK so since Apple where making the physics too simple and assuming it's linear for simplicity we've missed everything important support. OK So again that's the note to the message there now let's keep going with this OK this is the use of measurement again on the arse my Steinway piano of a change in the tone color with volume I promised you this I'm going to show you know the real measurements so what I did hear was I played the same note A for forty three different volume levels very loud sort of medium and very soft and then I did an analysis spectral analysis of the of the signal and I and I plotted here that the strength of the different components so this is the partials of the fundamental partial number one second partial third partial Now again this is not going to again be to criticize physicists when we draw our our spectra in our physics classes we think we always think we tend to think that the fundamentals are strongest and then they get they get weaker from their right it's within that right Paul right and so here the soft no is indeed like that the the the. The fundamental the first partial is the strongest and then it gets weaker by about a factor of ten in power and another factor maybe one hundred power so it does get weaker as you go up and up in frequency but if you look at the loud note the second partial is actually the strongest one. OK now it's kind of amazing because we still hear this as the pitch of the for of the sound. But again this this wouldn't happen this change in the curve wouldn't happen if we didn't have this nonlinearity change in tone color so that the change in tone color is not a small effect it's a large effect OK And this is not this is not just special for this particular note I'll show you some more results in a second this is a is a streaming important effect for many many no misses is a four four is in the middle of the range and we're going to see it gets worse while it gets better as you go down to that in the base. OK So let's talk more about about this. Went on to a fun time so let me just come back to make a couple of quick comments about something I mentioned quickly but I pass on over to perhaps too quickly too fast this is a picture again of that of that thing of the little wave was going away from the hammer I mention the hammer says the string it generates weight pulses that go opposite directions and the first wave pulse the one hits is the close and comes back well as you know physics it hits this and it gets inverted and even if you have a physics course you know when you play when you play jump rope on you don't do that anymore but when you have a rope vibrating you know if a wave hits the end of a string it comes back inverted so this pulse comes back upside down and hits the hammer and when it does that the hammer still in contact with the string and so it causes a large Rican pression of the head of the hammer back and forth and back or so this this is a picture of the hammer force on the string as a function of time so when I play the note This is when the hammer hits the string and then you get all these vibe. Up and down up and down as these Wait polls go back and forth back and forth back and forth between the short end and back to the hammer so in terms of who hears it doubly anybody doubly OK So you know that this is basically the execution function of our system and it's going to be this frequency content which gives us the tone color later OK it's complicated right and you see it simple so it's business I would tend to draw sort of a simple smooth curve but again I would be wrong and so again I'm I'm I'm criticizing my own discipline that's OK You think so but you want to model a piano sound you've got to take this into account and we've got to do that but I want to time and talk about it today. OK Let's talk about another interesting problem. Where should the hammer strike the string. Sound simple you see you see here. That I've got this hammer is pretty close to one end of the string and if you go to if you look at my Steinway or look at most modern pianos the hammer strikes the string about one seventh or one eighth of the way along from one end. Now why is that important. It's important because of this. Those partials are different what are called vibrational mode vibrational patterns of the string and the lowest partial is is sketched here the string just goes up and down together it's a standing wave and probably all seen that in many you have the second partial looks like this part of the string go up and down and up and down in the center has what's called a note in the center there's no motion here at the middle and the third how much it looks like this you get like one of ours but it's up and down in different ways and you get to knows now the point is this. If I suppose I hit the string with a hammer right there in the middle there's no way I can excite this kind of this mode and putting it in the place where doesn't have any X. and the amplitude. So if I hit that if they have piano hammer were to hit in the center it could not excite the second harmonic second partial in fact who here is a guitar player. Have you ever tried to find the center of the guitar string with your eyes closed. OK you have OK walk out tell us how to do it if you close your eyes and you pluck the string in different places when you pluck it at exactly the center all these even harmonics go away and the tone color changes in a way you can get to know the answer you know OK OK it's an honest man appear in front so but the point is by listening to the tone color you can tell when the the second and forth and so on our minds are present or not present and likewise if I pluck the string here one third of the way or two thirds I will not generate the third in the six and so on so this is where you hit the string determines again part of the picture the tone color so for the piano if you strike the string at the one one seventh of the way out you won't generate the seventh in the fourteenth and so on harmonics so. And this is this plus going to show that let me say before I go any farther I have no idea why eliminating the seventh harmonic makes a better piano tone the somebodies decided that it's a good thing you don't want the seventh of the harmonics that's a matter of perception and how we think about the word best again but that's that's the way pianos are made nowadays OK Krista forced you know actually different but the modern pianos are all pretty uniform in that it's about the one seven point for all the strings if you look at your piano you'll see that the strike point is designed very carefully and here's a plot to show some of that. To see this carefully I what I did I actually took the spectrum of the lowest note on the piano a zero low very lowest note and this is the actual the power at different frequencies as a function of frequency so every peak here is one of the partials and you can see that there's a it goes up and there's a big dip here that's about the eighth part. There is no Dipper on the sixteenth partials is not perfectly clean we're engines are engineers here not business but it you can see a definite pattern that it goes down to around eight around sixteen around twenty four and so on so this this is again showing you this effect of the strike point. So again where the hammer hits a string has a big effect on the tone color OK there's actually no one another effect I want to show from this plot if you look down here you see that the peaks down here get pretty weak down going down toward the fundamental down the lowest partial So let's blow it up and look at that a little bit more. OK it's the same as the same plot on the left. That show just now and I've blown up this lower part to right here so this is the this is the spectrum at the low frequency end of the lowest note on the piano. Now I want to I will tell you that's where the fundamentalists. I don't see it either OK. There's a second partially a little bottle guy there then third fourth fifth and sixth and you can see that some of the highest biggest part of the other six and so parts of these are the hot these are the stronger partials and the fundamental is basically missing. Gone and now this is an effect that was actually understood and recognized by Helmholtz a very famous physicist also very famous in terms of his physiology I mean he invented the instrument that that eye doctors use to look into your eyes is still used today a very very clever guy or yes very clever guy. And he understood this is missing and then he asked OK Why is it missing or here's the real mystery the fundamentals missing OK so what I don't care but when you let's talk but the real question is this when I listen to this note I hear the pitches being this pitch this twenty seven Hertz I don't hear it being fifty five or this or this or that you. Might have thought that the IF that the apparent pitch would be the pitch of the the strongest harmonic but it's not it's the pitch of this guy who's missing. So somehow we hear the pitch that isn't even there that's the mystery and it's called and it's called the missing fundamental what it's called. And the the idea that helmet was put forward which is basically still believed today a with with some important modifications is this when you here's a here's a very Here's our back to the fishes a schematic of a part of a spectrum you can make a spectrum like this where this is the power as a function of frequency so I can no fundamental power that I get power to have three or four or five so what Hell most pointed out was he said look these these partials are separated by F. one the the value of the fundamental So maybe what your ear is doing is some kind of sophisticated spectral analysis which actually and assigns the pitch to be the difference between these frequencies and not the fundamental and the lowest value and for low notes that seems to be a pretty good explanation of most experiments are there some people are still doing experiments today with people but how we perceive tones but that seems to be explain what's happening and this explains the the mystery of the missing fundamental is actually the mystery to mysteries the first mystery is why we perceive this to be the pitch and I kind of explained how well how most is kind of explained it but the other mystery is why is it missing in the first place why do I get why is there no power in that at that frequency why why did my piano my good piano I paid so much money for why did you give me anything at this at this you know if I could short change with a low notes what happens so that's the second mistake I want to talk about. I will get to it before I get where I can explain that I thought a little about sound board member the sound board the far end of the string attached to the soundboard when the string vibrates the sound board is pushed up and down it's like a. Big speaker so talk about this that let's talk about the big speaker the soundboard is a piece of spruce I can't really see it here it's hidden under the strings. But it's the same kind of wood that makes the top plate of a guitar or the top plate of a violin it's a little thicker for a piano but it's basically the same wood this spruce this well is is the is the would have choice for this because it's got a very high elastic constant called the modulus and in fact the ratio of the Young's modulus to the density is a figure of merit for these kinds of vibrating woods and that and they the figure of merit for spruce is just as good or even a little better than steel. So it's excellent for this purpose. And this is the soundboard made that is put the others way it's a piece of spruce this is the top looking at the top with all the strings ever taken out of the piano if you look carefully you'll see grain running this way and what happens is there are pieces of spruce that are that are perhaps maybe six inches wide depends on the quality the piano that are glued together edge to edge and they run from lower right to upper left that on top of that is glued to the bridge of the pants with the strings are attached member this is the front of the piano where the keys would be and the strings attached or go over the bridge like they do like for a guitar say. This is the bottom side the spruce is typically half an inch of an inch thick so it's pretty thin and so to give it extra strength in this across the grain direction there are ribs on the bottom and this with the ribs look like from below so this is this is a complicated enough of thing that it's not to say it's not trivial to model in fact. I want I want to Paris Paul or anybody here by asking this question but I'll just pose it how many how many independent Alas this concepts are there to describe to the vibrations of a single piece of wood. The answer is what he said to your off by an order. Magnitude Paul you're a physicist OK. What has twenty seven different elastic constants Young's modulus ratios all these things OK So wood is used you think physics is complicate you think electrons are complicated Wrightwood is complicated. So so it's but but it can be managed and I'm sorry for the polar time. So let's talk about about. Some were again going to get this. But OK so about how about of the House on board vibrates. This is my cartoon of a soundboard. The lowest frequency but you can ask like the string have had a fundamental frequency that was lowest frequency it was it was of give you the pitch of the no there was a fundamental frequency the soundboards or things like plates also have a have different vibrational modes in the lowest vibrational modes called a breathing mode just like the way when you breathe your chest goes out and in and out and in and that's what I'm trying to show here this sound board just goes out and in that's the lowest. Vibration vibrational of the sound board the frequency that mode depends upon the size of the soundboard for a large grand piano it might be say seventy five herds so it's toward the low end of the piano but not but not all the way. Then the second harmonic are second actually we're not really harmonic the second mode is a different mode where like if I can't do this but my chest goes out in and out out of phase like that that's the second mode I try to draw here and then you get other modes that are more common to the third mode and so on and these modes there are precise modal shapes depend upon where the ribs are in the bridge and it gets pretty complicated pretty quickly but the most important thing to realize is are two things first of all the sound board the lowest frequency is about seventy five hertz and that's important because. If I try to make something vibrate at a frequency below its lowest vibrational frequency it won't move very much and so you. You just can't it's very hard to make the sound board vibrate at frequencies well below seventy five Herb's So that's one reason why the fundamentals missing a twenty seven Hertz that's way below this one seventy five it's a couple boxes below it just doesn't want to vibrate so even though I have a lot of force in the string the soundboard won't move very much. So that's one reason why the fundamental is very weak or even missing another reason is that the sound board is too small to create a frequency of a tone with a frequency that low the again these Raise your hand OK Who's it if you want about antennas No you guys don't want to buy in tennis anymore do you and. Well if you learned about antennas you would know that in general a good intent of an antenna structure that's good for certain frequency has to have a size that's comparable to maybe half the wavelength of that radiation who's who's a got it or I will look I keep looking I can't find people so the the the wavelength for the frequency is twenty seven Hertz is a bit more than ten meters maybe fifteen meters so it's much much bigger than the sound so sound was a tiny little thing compared to the frequency is trying to read compared to the wavelength is trying to radiate and that's in the other reason why the the fundamental is so weak or even missing. So so I've explained to you why the fundamentals missing and that's good OK now. To get ready for the next next part. The sound or vibration again it's a sort of the breathing mode and so if I look at the breathing mode the sound board going out and in and out and in but if I look at a particular point on the bridge I don't draw a bridge here but you can imagine that a particular point that's off center a little bit won't be measuring moving just out and in it also be rocking a little bit as a sound right we were there as my chest out a place off center will rock a little bit to the side. And that will be an. Shortened because when the string is attached to that point the end of the string won't just move out and in it will also rock a little bit and in physicist language what that will do. Is let me skip this. Skipping I'm going to say they skipped a lot of K.. So one thing. This is again the equations I need to show you questions for the mathematical physicists here. In physics we're used to dealing with second order partial differential equations OK I don't know if I ever had ever seen a fourth or differential equation in physics till I started looking at pianos and we just don't see them but who are the civil engineers here raise your hand for civil engineer come on Raise your hand so engineers see all is this is thin plate theory is called spin around for hundreds of years OK So engineers do have something office once in a while but if you want to if you want to model the vibrations of sound boards you're going to have a very harmonic kind of structure and the different the elastic conscience come in a complicated ways I can talk about later if you feel like OK so let's keep going OK let's talk about. Piano tones and how they how they decay with time so these four plots show four different notes let's see this. Upper left is the lowest C. on the piano and then we go up a couple of octave C. three over here there's a C. six and this is the highest C. on the piano now what I've done is the time scale is the same for all these plots and so it's clear that the low notes last a lot longer than the high notes and this one only ended I put I stop to the no I put the damper down and so you can see right away at the high notes decay quickly the low notes to us for a long time which you already knew when you play your piano. Now the reason for that technically it has to do with the effective impedance and talk about in impedances for the string and the soundboard and how they are the same or different OK So that's it that's a detail I don't want to go into. Instead I want to look at the details of this decay so what I'm going to do next is plot this decay but in a different way I'm going to actually look at the instantaneous amplitude as a function of time so I go to this thing to up and now I crazy and to take that away and look at the amplitude vs time and it looks a lot simpler it looks like this so it's the same curve but now I'm spotting amplitude vs time and you can see that it decays this is a semi log Plaza logarithm over. Log scale on the left and linear scale for the horizontal and it decays roughly like one curve here and then a different straight line here and this is this is called a double D. K. very profound name and what it but it definitely means that the two different things are happening on something the DK is is is has one characteristic or one character early in the different one later on and what's happening is this. Well sketch here here's again the hammer the string the string runs from you the player to the back of the piano the hammer hits the string and the strings vibrating in this plane OK now that that direction of string vibration gets the sound board moving a lot so you get a lot of radio a lot of sound generated and the string dies away role of the quickly but as the as a soundboard is vibrating it has this rocking motion so the end of the string is moving up and down but also rocking to the side a little bit so that means you start the string vibrating in the horizontal plane So what started out as this kind of vibration and meaning being a horizontal vibration now the whores Olla vibration doesn't move the sound board very much but somewhere in one and so it decays much more slowly so what you've done is converted in technical language physics language you converted from one polarization the vertical one to the horizontal one. And so and now. What that does is it means when you put your piano tone it dies away quickly at first but then it will linger for a long. TIME So if you're playing an awesome Chopin or whatever and you it's long passages you have that a bit of ale ability there which you wouldn't have if the thing just decayed away all the way quickly. OK. So this is this is so the piano the decay of NO is a little more complicated you might have guessed as your physicist OK as if this is now it's this this is that the decay for that lowest note on the piano that has I didn't mention it it's only a single string creating the note as you go I forgot to mention this earlier I should have I got to excited as you go up in the can a keyboard some notes have two strings as you may know and some even have three so it's still the same kind of behavior but now we're going to look at history in a note this created by two strings. And what you see is this you get you get a fantastic eight first and then you get a slow decay but you kind of get this oscillation and if is this would recognize it as well as a sort of beating effect so what had what's happening here is when you first at the string we heard when the hammer hit first it's two strings they both go up and down together and he K. a lot and then they eventually make this the K. but they start feeding S. or they get out of phase because they don't have quite the same frequency and so it's for a while they're in phase and then they're out of phase and then they're in phase so this is this is an effect called beats so and you can imagine if I have three strings it gets more complicated for a physicist this is very similar to having two modes that are coupled with some so we perturbation it's very similar physics you can write it down. But this again now this why is this important if you listen to a piano tone you can really hear this because sometimes you hear you play a note and it lingers on and you hear kind of a little bit of our bra go OK that's where it comes from. And that mic and as you probably know if you listen to your violin friends play violin they're always ready for bottle a little bit of Roboto is a good thing just ask your violinist friends and that's what that's where you get it here just from a single note OK. So the. I wouldn't I'm Paul. I Some done I thought I forgot where I was in my talking I'm sorry so let's let me just summarize and sort of recap some of this. On the face of it a piano is a very simple thing right what could be simpler than a vibrating string. But there's a lot of physics that involved here string stiffness is very important it effects tone color inharmonicity effects the way a piano is to number the octave stretching effect are crucial how piano sounds. The hammer hammer non-linearity is perhaps the most important part about the piano it's it's what makes it expressive this is the part that Christopher he really understood even three hundred some years ago. The interaction between strings leaves this very interesting time decay which isn't just a curiosity but it's really important for a broader effect which has make it really gives the tone an extra richness which which we all like to hear the bass notes you know or our puzzle are those are surprise no fundamental but we still hear the same pitch so again we see how the importance of us the listener in the whole in the whole business. And it's what I mean by I say psychoacoustics plays a major role in how we perceive tones and how the piano is made so to me the bottom line is this there's a lot of interesting physics in your piano OK thank you very much. Of. A of. Thank you very much. Or is. All. Right thank you should also. Be. It's very well if you will just say. Well if people who know me know that I love to criticize engineers and chemists but I think sometimes physicists deserve it to you like I think we have five. The question was Why don't we make a piano just out of our computers rather than out of anything else right. OK. I think it's certainly possible a company like Yamaha are doing things in that direction not not quite as much as you think you have the. No I wouldn't buy I would never dial violate a physical law I would never think of doing that. Well I'll just say one interesting thing I mean I thought you were Ed you raise a very interesting point brought to say one thing one comment about that. Yamaha makes some very interesting. Pianos that aren't acoustic piano but they're they're basically make sounds into in you know based on equal on sentence or other things and what it was interesting to me is the latest ones they work very hard to give the to give the Keil every give the action the same feel as a real piano even though it is not needed to do that but they also do is they have vibrators built into the piano so the bite so the case vibrates the way our acoustic piano would even has nothing to do with the way they create sound so that all they're doing is making it IS TRYING TO SCHOOL you into thinking it's a real piano so they decided that it's important to fool the player to make him or her comfortable OK so. It's like we have our ideal piano in our heads and we're never going to let it go even though we don't need these other things OK so I don't so maybe the piano is locked into place and it will never touch some of the parts of it when I read it I mean. Yeah I don't. Know I didn't I didn't say it was an answer your question I was just making some comments as my prerogative. Why they go from whether he covered hammer covers to tell him recovering so. I could I could you know I collect pianos and I have a lot of pianos with leather covered hammers. And it's amazing that the leather made two hundred years ago is still very good I think that I think the chemical way it was made was different two years ago that is now and it lasted longer I think the reason is that the felt was ultimately more durable when you have higher tensions and bigger and bigger strings I think the leather those wear out although I guess I have you know they have nice leather coverings and work well I think it's just a question of Dura building wants to string tension got up and it was hard hit that they were faster you know question over here. OK. Yeah I do I well I didn't write that the player but there on my website you know we did that I mean we made a whole cup I'm a whole piano based on you know pushing the hickey lever and so the calculation involves this non-linear hammer hitting the string and then take into account a complicated way form I showed you and then at the strings vibrating and gets the sound board moving and we calculate the vibrations a soundboard we did that with all the elastic conscience and things and then coupling a sound or motion to the air in creating the you know the acoustic pressure you hear so go to my website you'll hear some sounds in. That. Book in the form or do well it's funny my piano teacher used to always talk to me about how I push the key down after after him and I don't believe that I mean I just don't believe but but certainly. Expressiveness I mean a lot of it is it just is just certainly how hard you hit push the keys and things there's have been some really interesting. Studies of real per piano performance and a lot of it isn't timing and having that having the the note come a little earlier a little late in those you've probably seen those I mean not just Chopin type rubato but either things too and so that's a big part of it in fact. The way you make a harpsichord. Sound more interesting is by even though the notes are all the same volume is by zero by the timing so there's a lot a lot of that I'm not I'm not a great player so you shouldn't ask me only question and yes. It's basically the same physics question was about about square square pianos in the green and then uprights. Square piano well the earliest pianos were made in the shape of harpsichords which had this wing shape we've already seen but there was also an instrument at the time I was referred to as you already know this at that time there was instrument called the clavichord which was not square but rectangular and the the little ones were sort of about this big on the table maybe this big around and and they had the strings running sideways as opposed to front the back and they were basically the household instrument so when Mozart would go to give piano lessons to somebody they would probably play a square piano in the house and they were the ones in Mozart's time where some I could pick it up it's very light and everything else I could really carry but I could use that small. Things are basically the same for that. The bigger change was when they want to up right now the square pianos. They had the same thing happen to them happen to grand piano if you want to add more and more nos so the earliest squares were like four four octaves and then they started adding bigger and bigger I have a I have a Steinway square the bass from eight hundred fifty eight that has seven octaves and it weighs I think it weighs a ton I mean it's the it was it's huge it's got it's it's very massive and it wasn't a very good household instrument anymore at the same time in the early eighty's hundreds the technology for uprights was invented and what was really the key to technology was the action right the action I didn't talk much more about it but the action for a grand piano and for a square the Hammers come up from below. Hit the strings and fall back so you have gravity working for you for the upright piano the Hammers hit the string this way and it had to bounce back she got levers and keys and all kinds of things and it took a while for people to invent a good action and is about eight hundred twenty or so when the the actually the base of the uprights today was invented and after that the square started to go away so but the physics is so the main physics differences in the action as a post anything else back in the back there I've never looked for that so the thinking in musical acoustics is that the phases don't really make much difference the phases of War of one partial to the next so we usually ignore that. I never study that but I guess that's what people that's what people think. Or questions yes. You want to talk about that. But the question was this made me put the question Could people hear I mention this business about octave stretching and I talked about octaves being a factor too but there are a lot more ratios that go into musical scale than just a factor of two. Now. Do I start G.'s. The way we tune pianos today isn't it is that it is a method called equal temperament which means there are basically twelve notes that span the octave and so we go one twelfth of the way so that across the octave for each node but one twelfth of the way and in a log makes sense OK so the ratio of C. sharp to C. is two to the one twelfth and indeed a C. Sharp another two to the one twelfth and twelve to the one twelve's or whatever it is make two right that's what it has that's equal temperament Now some people believe that equal temperament was the worst thing ever invented. The for the following reason if I play. The notes C. and I play the note G. which is supposed to be a perfect fifth or point I suppose to be a fifth. If it's a perfect fifth then the ratio will be precisely three to two so are three to other OK and that will be a very nice pleasing ratio in my ears and I want to hear much beating other than the higher harmonics higher partials but the fundamentals won't beat but equal temperament to the five twelve divided by two is not three to two OK so. That that interval seed that she will beat when you listen to your piano and some people most of us don't like that very much and so OK Mike I can make sure I get back to your question OK. So by tuning in equal temperament we give up the possibility of perfect intervals and making really beautiful intervals beautiful walk beautiful chords. But what we gain is we gain the ability to play equally bad in any key OK which some people believe is a good thing OK And you can sort of sense how I feel about it but that's OK OK. But Equal Temperament equal temperament was what was used up till about it depends what you believe at least Mozart's time I started to change and then maybe a little even a little later than that. Now it's sort of the question of equal temperament versus versus other temperaments gets is sort of a moot question for the piano because I've already showed you that even a single note is a harmonic and the deviations of a single note from being harmonic are about the same size as the deviations from temperaments between equal temperament othersome So I think tuning a least twenty a modern equal temperament to an equal temperament wouldn't really make much difference is the end OK so I'm not sure about but but the real reason is you can marshal keys and you can play you can play in F.R. minor as well as you play in C. Major those kinds of it and that makes it easy for playing other instruments and other things like that but but I mean there's been many books written about it you probably know this by temperament right there have been there been many. Yes. Why is it that OK well I was speaking figuratively OK. I guess I'm not an expert on this there are people there's lots of these things I I think I've read about people doing experiments where they play one note on one side and one on the other they and some of the. Harmonic Motion get put together not in your single air and I think also people don't experiment is not a non-linear effect I mean it's not just non-linear mixing I mean that as a double E. You might think this is non-linear mixing two fundamental partials mixed to give us one is not that simple I think it is done at the higher order of parts I think but I I think you can look it up I swear I seem to remember. Yes. I forgot to mention that OK the question was What about the Loki's on the bows and or for why are they important OK I got to show up I have to look at my pictures OK I don't actually own a boat and or for I can afford it but. Why do we do that. OK let's let's do it here. The lowest note is one of these pay strings it's way along the edge of the soundboard OK now and it runs over the bridge close to the edge of the soundboard Now remember they're breathing mode thing if I'm trying to make the soundboard move and I'm pushing on at the edge it's not very good and so when I'm with the bows and or fried it throws you haven't seen a post over some of them have an extra one of three or four notes on the on the low end how many is it I don't know if they have about four the but for no it's leavin lower than the one on the Stein ways and every other piano and they actually give them a different color so you don't get missed disoriented when you're playing the piano and but what they do never mind if you use them or not what they do is they make a means a sound one has to be even a little wider than this one which means this lowest note is actually closer to a farther from the edge which is a good thing for it so that's why so I think adding those extra four or five most of those never makes other notes better they serve as they sort of sacrifice themselves for the good of the piano and so I think to me it's that's one of the Course important things they do so. This. OK The question was Has anybody ever tried to design a key where you could push not just down but in are now it. I don't know about that I mean there is a pedal on the on many pianos it's called the. What they want a quarter pedal or if you see that you know what that OK Out to explain so some pianists like I like most grand pianos there's a there's a lever there's a well there's one there's the one pedal you all know lift the damper So the thing vibrates more in sound sort of more resonant but there's another pedal which shifts the entire keyboard over a little bit so instead of having a hamburger three strings or two strings it's one own accord and that's where it comes from right and the idea there is the you get all you get first you'll have one string so it changes the tone color a little bit you also are playing with a different part of the hammer which is probably a little softer than the part that's been you know been been rough and up by always playing so it will change so that's a little bit of what you're saying but that's that's the only think it goes in that direction like that. A lot of.