Embedded - 236: The Concept of Delayed Gratification

Episode Date: March 2, 2018

Roger Linn (@roger_linn) gave us new ideas about musical instruments, detailing how wonderful expressive control, 3D buttons, and keyscanning can be. Roger’s company is Roger Linn Design. We talked extensively about the LinnStrument, some about the AdrenaLinn for guitar, and only a little bit about the analog drum machine Tempest. A key matrix circuit is a popular way to handle a large number of buttons but it falls prey to n-key rollover. Roger adds force sense resistors to this (FSR example at Sparkfun). If you have an idea for an instrument, Roger has already written his response to your request for a prototype. Roger gave a keynote address at ADC '16 about the LinnStrument, including showing the sounds it can make. OHMI Trust is the one handed musical instrument society enabling music making for everyone.   Roger mentioned some other expressive instruments including: Roli Seaboard Haken Continuum Madrona Labs Soundplane Eigenharp

Welcome to Embedded. I'm Alicia White, here with Christopher White. Our guest is Roger Lin, creator of the drum machine and many other electronic instruments. Also, he won a Grammy. We're nervous. Hi, Roger. Thanks for joining us. You're very welcome. My pleasure. Could you give us a brief introduction of yourself as though you were on a technical panel in an engineering conference? Sure. Yeah, I'm a designer of electronic music instruments. I'm most known not necessarily for creating the drum machine because there have been various drum machines in different forms ever since the 1930s. But what I am known for is creating the first, what's called a digital drum machine in 1979. And it was the first instrument,

drum machine instrument to use sampled sounds. And it was used on a number of hit pop records in the 80s. And then I had a series called the MPC, which is the staple of most hip hop records and many other pop records during the late 80s and 90s. And basically, I try to make a lot of interesting paintbrushes with which artists can paint more beautiful pictures. I like that. We usually do this thing called lightning round where we ask you short questions and we want short answers. And if we are playing by the rules,

it all goes very fast. Favorite musical key. Oh, geez. I don't know. Say, um, uh,

B flat. A horn player, huh? No, actually I'm not, but I would ask for it to be a rebellious and give you a non-obvious answer. I would be ashamed to give you C or D minor. Would you, if you could only do one, would it be software or music?

If you asked me 40 years ago, I would have said music. If you asked me now, I think it would be making electronic music products. Actually, I'm not that great of a software engineer, but I just sort of put the whole package together. What's the perfect beat per minute? Oh, God. Anything below 100, I'd say.

Analog or digital? I'm more of a digital. I've never made an analog product. Would you rather play a violin or tune a harpsichord? Oh, God. That's a digital guy. I've never made an analog product. Would you rather play a violin or tune a harpsichord? Oh, God, that's a false choice. I could have put banjo in there, too. I think I'd probably like to play a violin more than tune a harpsichord, but there are infinitely more things that I would,

there are things I'd infinitely prefer to do than either. Is there a tip, a technical or a musical or any tip you think everyone should know? Before you call yourself a musician, try to learn what notes are, maybe even chords. And then if you're really ambitious, try maybe melody, harmony, or some music theory. All right, well, let's get into the interview. That's for all the DJs out there. So you said you started with this sampled drum system in 1979? Yes.

How has music changed since then? Well, I think everyone's sort of trying to find out what the best combination of human craftsmanship in music and electronics and technology is. And so what's very popular right now, which I alluded to in my earlier comment, is being the DJ, where instead of actually playing those pesky things called notes, you can just say, I'll just take some little recordings of people playing those notes, and that way I won't have to learn how to do them, but I can just combine them together in my novel way. And the truth is, there are a lot of DJs that create very compelling music, and it allows one person to create a very compelling performance. And so at the front of the room, this person looks to be very talented, and it works. It's just that all my DJ friends, the biggest complaint they have is that no one gives them any respect.

And I think that ultimately, we want our heroes to be somewhat superhuman. And if people look at what someone's doing and say, gee, anybody can do that, then it tends to make them less superhuman and more just human. But technology has changed a lot too, and it's made some things easier and more toward the DJ sort of world. But at the same time, it's given tools to everybody. Is that a good or a bad thing? I think it's both good and bad uh you know it's it's good because it allows someone uh to create compelling music without a whole lot of uh work um but the bad news is it tends to create music that the technology wants you to play as opposed to what

you want to play and so that's why we get a whole lot of music that's 140 beats per minute and without any chord changes or any melody, things like that. That's driven by the technology then. Yeah, technology wants you to make it very simple, have very few changes. If you don't know, for example, anything about music theory or chords or melodies, you're not going to be able to play them in. So what you can do is you can find a loop from a recording at some point in history and say, well, there's my music theory. That person did it for me. And it's great. It's just that there's only so much you can do to modify an existing recording.

And if you really look at it, I think you'll find that most people have the greatest respect for people that not only have a musical talent, but have developed some craft, perhaps not all the way to the 10,000 hour rule, but they've done something that makes them appear to be, to the naked eye, somewhat better than the average human. I can't quite decide if we're talking about open source software or music. Well, something like that. The work you're doing now does involve open source software, right? Well, probably a better way of stating it is my current main product, which I call Linstrument, because my last name is Lin, and it's a musical instrument. I did release the software, the operating system for it, as open source.

But that's not the main point of the instrument. It's just, it's a way to give people the equivalent of what used to be in acoustic instruments, the ability to take your instrument to a luthier or a craftsman and have him or her modify it to meet your needs. With electronic instruments, there's a brick wall. And so the open source software allows you to do that to a degree. As a non-instrument, as a non-musician. You're also not an instrument. I'm also not an instrument. Well, maybe take an analogy. Let's say, for example, you're a guitarist and you buy a guitar from the Fender or the Gibson company, and you like it very much, but you'd like to have what they call lowering of the action. You'd like the strings to be closer to the frets,

and you'd like perhaps the frets to be what they call honed, so they work better when you bend strings. So you would take it to a guitar craftsman and say, I can do that, I can take this mechanical guitar and I can hone the frets, or I can remove the frets and put in bass frets, which very strenuous rock players used to like to do, and they can lower the action or raise the action, and things like that. And so you can't do that in an electronic music instrument. And so

the nearest thing to that, I think, is to release the software as open source, so someone can make modifications if they have such skills. That's really unusual, at least in my experience with electronic instruments. I mean, either it's an analog device completely, in which case, yeah, you can open it up. And if you're an electrical engineer, you can make modifications. But for people with digital instruments, I don't think I've ever seen that before. There are some instruments that are open source. And it usually is, like in my case, there's smaller some instruments that are open source uh and it usually is is uh like in my case there's smaller time instruments smaller companies where people have uh they not

only want to make a living selling instruments but they really have a an interest in in um providing something uh an empty camp a canvas that people can use to upon which to paint. Having seen a little bit of your instrument and watched you play it on a YouTube video, it seemed like it could do anything. I mean, you had it sounding like a violin at one time, and it had drums, and it had a whole symphony, and a lot of lights, which I will ask you about. Is it hard on the musicians to have no constraints?

Well, there are constraints. And first of all, maybe for the listening audience, let me give you a quick description of what it is. If you look at the instrument, it looks like a big grid of little square rubber pads, kind of like you may have seen in drum machine drum pads, except smaller. They're just three quarter inches size placed next to each other, and there's 200 of them, and they're arranged in a big rectangle, and the idea of it is they're lit up, too, and so the idea is, the metaphor is the stringed instruments. You'll have eight rows, each with 25 of these small rubber pads.

And each row will be a series of chromatic notes. La, la, la, la, la, la, la, la, la. As though you're playing the increasing frets on one string of the guitar. And then each of the eight rows of these 25 pads is like another string. And so each one is tuned a little bit higher than the last one. And what Lindstrom does that's unique is instead of these just being on-off switch buttons, like a MIDI grid controller, there are instruments, for example, called Ableton Push or Novation Launchpad, and they have these buttons with lights.

What this does, each of these little buttons, they sense the musical note C, and then slide your finger up to the musical note D, and you'll hear the pitch of the resulting sound slowly slide from C to D, or then up to F or up to G. And so it gives you the ability to do what you do on a violin, for example, to change the pitch like that. It's also sensitive to your pressure, so the harder you press, the louder the note is, the softer you press, it becomes quieter again. And then further in the third dimension, if you move your finger forward and backward, it changes the timbre of the sound, the tone. For example, if you're playing a violin sound and you move your finger towards you, it would be the equivalent of playing the bow near the bridge, a very thin sound. But if you move your finger towards you, it would be the equivalent of playing the bow near the bridge,

a very thin sound. But if you move your finger away from you, it would be the equivalent of playing the bow of the violin closer to the neck, which is a more round and mellow sound. So by having these three dimensions control the parameters of note loudness, pitch, and timbre, you're able to get a level of virtuosic or skillful control of music as you would be on a real violin or a guitar or a saxophone. And the idea of it is, it's what I think the ideal new interface for performance of music of electronic sound. And it's my best guess at what people will be playing in 20, 30, 50 years from now. That's probably too much.

No, no, that was fantastic. But now I need to know about these, I want to say 3D joysticks, but a joystick is really only two directions, X and Y. What does this sensor look like? Well, actually, it's a pity this isn't a visual show, but if you take an instrument apart, there's a metal panel at the top,

and in the center of this metal panel is a big rectangular cutout where these 200 notepads are that are three-quarter inches square. So if you take that top panel off, those rubber notepads are actually one big molded sheet of silicon rubber. Underneath it is a thin plastic sheet, and that's the sensor I designed. And that sensor is, I mean, if you want, I can got a smart audience there. So there's a thing,

your trackpad on your laptop, for example, it senses X and Y. And actually, if it's an apple, it senses pressure also. So that would be 3D sensing of one touch. And then if it's an apple, once again, it'll be multi-touch. Well, in a sense, what Linstrument is, it's like 200 little three-quarter inch square pressure-sensitive track pads, in the sense that it's sensing left-right movement or forward-backward movement or pressure. And it uses a technology called force sensing resistors, often called FSRs. And to make it lower cost and not have to wire together 200 of these individual things, which you couldn't buy anyway, I use a printing process on mylar sheets where a resistive, a fixed resistive ink is printed in a very precise process in rows on the bottom sheet and columns in the top sheet. And then over printed on top of that is this

force resistive, force sensing resistive ink. And so on both the lower sheet of the sensor and the upper sheet of the sensor, and then there's some spacers in between the two, so they don't touch unless you are pressing on them. And so to try and not get too technical about it, the best way to say it is that at the edge of each of these eight rows and 25 columns, there are a bunch of analog switches. And so what the software does is it addresses just one row column intersection at a time, and it addresses it in such a way that it's measuring pressure using these force sensing resistors. And then it does this very quickly across all 200. And then once it finds a touch, which means you're playing it, then it says, oops, I better get the X and the Y as well. And so that's probably the best way to do it.

And if you want, I'll get into more detail, but I don't want to snow your listeners and get them to turn off and listen to some sort of another show. We don't actually care about our listeners. Sorry, listeners. I hope that's not a shock to you. It's mostly how excited I am. And really, this is cool. So the 4-synth theme, you are treating it like a matrix. You have so many buttons here.

Yeah. When we do digital buttons, we often treat them in a matrix fashion with the rows and the columns so that our GPIOs aren't exploded into a billion. You're doing the same thing, but with analog switches. Yes, because I'm actually measuring analog voltages. So when I, for example, scan very quickly while you're not touching it for pressure on any of the row column intersections, which equate to the actual three quarter inch buttons on the surface, I'm looking for pressure. So maybe the best I can describe this, and if you tell me if I'm getting in too deep, let's say I want to measure the pressure of the lower left cell. So I would address the leftmost column and the bottom row. And so the intersection

of those would be the lower left cell or row column intersection. And what I do is I connect both ends of the lower sheet, which is the row, to ground. And I connect both ends of the upper sheet, which is the column, to the analog to digital converter, pulled up to a voltage. So what that does is, if I'm not touching it, the analog to digital converter pulled up to a voltage. So what that does is if I'm not touching it, the analog to digital converter will see the maximum voltage, which is about 3.3 volts. As I start to press on it, the little microscopic bumps in the force sensing resistor ink at the top layer start to touch against the little microscopic bumps in the force sensing resistor of the lower layer and they touch just a little bit first so you get a little bit of current between the two and it

starts to lower the voltage because it's starting to bring it to ground and you effectively have a voltage divider between the pull-up resistor and the ground potential of the lower layer. So the more you connect those two layers of force-sensoring resistor ink together, you start to connect them together, and that pulled-up voltage starts to get pulled down towards ground. And so the pressure, the measurement of pressure, is inversely proportional to the voltage at the analog-to-digital converter, if that makes sense. It does. This is amazing.

I mean, this would have required a lot of design work and thought. How long did this take to develop? It was about a four-year process to develop the product, and much of that was the sensing system. But now I've been able to patent it and i got the patent granted so uh at least it gives me some um uh patent protection if in fact someone tries to rip it off when i do this with a button system with a digital uh matrix key matrix system there are sometimes some keys I can't press at the same time and be able to differentiate which is which. Do you have any of those problems or have you laid it out so that you can tell when multiple keys are pressed? Are you referring to a problem in any matrix

keyboard called the N key rollover problem? Yes. Okay. That is true. The problem with this sensor is that if any of the listeners don't know what the N key rollover problem is, is that if you have a matrix of keys and you press three corners of a rectangle, that allows the fourth corner of that rectangle, which is not pressed, to be pressed because you're pressing all of the parts of that rectangle. So normally in, for example, computer keyboards, they have a diode in each of the switches of the computer keyboard in order to stop that from happening. And you can look up the principle of that. The problem with doing this in this, my sensing method, is that you can't put a diode there because you're connecting actual force-resisting ink to force-resistive ink.

So what I thought about is, is this a bad thing? Well, it turns out that when you're playing music, you very rarely play three corners of a rectangle. And the nature of laying out the musical notes on this matrix, like a stringed instrument, is that you have the same note appearing on multiple strings at different locations. So in the very rare cases where you would find yourself with this end key rollover problem, then you would simply play that fourth note on it. First of all, the software blocks the fourth note. If you're pressing three corners of a rectangle, it says whether the fourth one is pressed or not, which I can't tell, me being software, I just won't play that fourth one. And so you're forced then to play it somewhere else. And at first I was somewhat concerned about it, but after 1400 instruments,

people don't really care about it. First of all, in expressive performance, which is what this is good for, in other words, bringing the expression of a saxophone or guitar or violin to electronic sound, you're often playing melodies anyway, one note at a time. And so when you're playing chords, for the rare cases that those chords happen to show up, there's another way to play them. So it hasn't been a problem. And you don't have debouncing problems because you are doing this force-sensitive resistance. We have debouncing problems, but we debounce it in software. And one of the problems with debounce is there's a constraint in musical instruments. You don't want any latency. When you press that button,

you want it to sound immediately. So what we do is we just play it on the first touch and then debounce it for the release. Oh, okay. Yeah. And so in other words, you can't, the software will not allow you to release it a few milliseconds later. It requires that you hold it for a certain amount of time, and that ends up being the debouncing. So low latency is incredibly critical to musical instruments, even more than most other buttons. Sure. But how fast does fast have to be? Well, it actually, it's interesting.

There are two issues. One is latency and one is jitter. Now, let's say you're at a stadium concert and there's a rock band, someone, you know, somewhere in their 90s, like the Rolling Stones. Well, the guitarist, Keith Richards, is often about 30 feet away from his guitar amplifier. And sometimes he'll move out onto a platform into the audience. He'll be 50 or 60 feet. Now, for every one foot away from his guitar amplifier, that's one millisecond of delay. So if he's 30 feet away, he's hearing his guitar 30 feet

later, yet he has very good rhythm. He's able to compensate for that. So the brain is actually very good at compensating for latency, but if one time he plays a string it's 30 milliseconds, the next time it's 60 and the next time it's zero, it's going to be very confusing to his brain. He's not going to be able to compensate that. So you want to keep the latency low but you also want to keep the jitter low in the instrument it takes four milliseconds to scan all 200 cells and so the the typical latency from playing a note to outputting it is two milliseconds plus or minus two milliseconds and people have found no problem with this when playing very fast chordal parts, like you might imagine Stevie Wonder playing a clavinet sound

on the song Superstition. You would imagine that if you're more than 50 years old, I guess. But at any rate, there are cases where you want to be very rhythmic about that, and you want that to be very tight. And so that level of latency and jitter has not been a problem for people i would hope not well no people can be very sensitive i mean if you talk to my brother christopher's brother is a guitar player he's a guitar player he's very sensitive to latency having i think trained himself to to notice it more than others but yes a few milliseconds here there can be annoying when you're mixing and stuff. But when you're playing, it's less of a problem.

Well, no, it's actually, the only problem is when you're playing live. When mixing, it doesn't matter because there's the time to come off the recording. Right. But when you're playing live, usually when people, and I don't know, I'm sure your brother is very adept in understanding what he's talking about, but many guitar players or other musicians, when they play into a computer, they talk about latency, but often they're talking about latencies of 30 to 40 milliseconds, particularly because that's determined by the audio buffer

size. And Windows basically doesn't like music. It sort of puts up an ugly frown whenever music comes its way. And so they require you to do a lot of tuning, and it gets very, very unruly as you do this. But eventually, if you're doing nothing else, it'll work well. By comparison, the Mac, the latency is very, very low. And you can tune that in your system to be even lower. You mentioned expressive control. And you said, working with things like a saxophone. And what do you really mean by expressive control? Is this just being able to push the buttons in all of the different ways? Or how do you define it? I would say probably the best way to describe it would be if you look at a great guitarist, maybe a Jimi Hendrix or an Eric Clapton or one of those,

or Django Reinhardt, or if you look at a great saxophone, saxophonist, jazz saxophonist, or great violinist, they don't just strike the note and let it ring out, and there's no change between the start and the end. They bend it, the note, they vibrato the note. If they're a violin, they slide between notes, and then they gradually ease into a vibrato, and they change, if they're a wind player, they change their loudness of the note over time. They're changing. Sometimes it's a very soft timbre, and then they change the timbre, the tone over time too. For example, a guitar player will pick the note near the bridge for a thin sound, pick it near the neck for a more mellow sound, or change his guitar pickup for different tones. Or a violinist will use the edge of the bow or

the flat of the bow or use a variety of different bowing techniques. And it's this infinite combination of variations of the sound under skilled hands that makes it more interesting to us as listeners, makes it more emotional, and makes it more expressive. And so, now, we have this sort of a false choice existing today. If you're a musician, you can play one of these older acoustic instruments, like the guitar, even the electric guitar, or the saxophone or the violin, or any number of other acoustic instruments in the orchestra, and you can develop skills of expression, vibrato, bends, slides, slurs, embouchure on wind instruments, all these tricks

that they use to make it more emotional, expressive, and to try and capture the emotion of a human voice and all the variations inherent in that. Or your second choice is to step into the world of electronic sound. And what's your main interface for that? The standard black and white electronic piano keyboard, often just called the MIDI keyboard. Now the problem therein is that a MIDI keyboard is little more than on-off switches. They turn the sound on, albeit at different volumes. If you strike it hard, it turns the sound on at a loud volume. If you strike it softly, it turns the sound on at a soft volume. And then you have the choice of nothing. You wait until you want to make the note stop and you let it go. Now to be fair to these MIDI keyboards they have a couple of extra features. They've built a

couple of sideways knobs into the left side of the keyboard. One is called a pitch bend wheel and with that you can press a note with your right hand while with your left hand's finger, you can move the wheel forward and back, and you can get a very unnatural sounding big slide, right? Which ends up no one's used because it simply doesn't work. It's impossible to develop virtuosity on it. And what they also did is they gave you a second wheel called a modulation wheel. And what that does is gradually adds a mathematically perfect vibrato to your sound. So you sound like someone who's always playing the same on every note. Consequently, people don't really use it once again. And the third thing that these MIDI keywords do is,

if you press the key down to 95% of the travel and you continue pressing, it senses the pressure of pressing harder, which is pretty much useless because by the time you've pressed at 95%, the note is already sounding. And so if you wanted to simulate, for example, the gradual onset of a saxophone sound, you couldn't do it because pressing it up to 95% turns it on. And then from that point, you could add a little bit of volume to it. So what Lindström does, and by the way, there are four other instruments in the market of what are called expressive controllers like Lindström that attempt to do, is to bring all those subtle variations over the duration of a note

with a skilled human interface, or a very capable human interface uh bring the expressiveness of those acoustic instruments to the performance of electronic sound and that's this is a solution to that false choice of those two elements it gives you a third way which gives you the advantages of the world of technology and the expressiveness of the world of acoustic instruments. As we moved to digital instruments, synthesizers and whatnot, they became easier to play, easier for the beginner to play. Yes. And this sounds like I would need more music theory, music understanding, understanding of how things are expressed and emoted in music.

This sounds like something that would be very hard to learn. As somebody who cannot play the piano or any musical instrument and sings only in the shower to protect their dog's ears, this sounds really complicated, but I know Christopher was looking at it and drooling. How do you... Who is your target audience? Is it the trained musician? Well, I would love to have as many highly skilled musicians as possible

because they would be able to apply their skills, which they presumably developed on an acoustic instrument, and apply that to playing on an instrument. And there are some very, very talented ones. And then there are a lot of people that feel that this is a better way to be able to develop these skills than to play a violin. I think one way of looking at it is that there are two problems with acoustic instruments. Number one, the man-machine interface, or shall I say the woman-machine interface, is severely crippled by the need for the instrument to make acoustic sound. And the second thing is that if you play an acoustic instrument, you must learn

one human interface for one sound. And then instead of pressing a button, you have to learn a whole new interface to play a different sound. For example, if you play guitar, that's a very good sound. You say, well, I'd like to play saxophone. Oh, well, you have to learn an entirely new human interface. So you can't switch between sounds. And one of the great conveniences of electronic instruments is that you can just press a button and switch to a different sound, right? So let me talk a little bit about that first point, that the human interface is severely crippled by the need to make acoustic sound. Let's take the example of the violin. Now Now the violin is, for the past few centuries,

is the primary solo instrument in the orchestra. It's the most beautiful instrument. It has this wonderful high voice and skilled players of it make beautiful music. And I do very much appreciate the skill of a great violinist. But let's take a look at that human interface. In order to make those high tones, the string must be very, very short, which means the fingerboard must be very, very short. And up at the top of the neck, those notes are only a few millimeters apart. That's really, really small. Plus, it has no frets, because you want to be able to slide those notes wherever you want. It's really hard to play that thing. Oh, and let's not forget, you have to hold it between your chin and your shoulder. Now, if I, as an instrument maker, came up with such a human interface and went into a store

and said, hey guys, I've got this great instrument. It makes these wonderful high tones. It's really hard to play in tune. And since the body is so small, you have to scrape the strings with horse's hairs all the time. Oh, and you have to hold it between your chin and your shoulder. They'd laugh me out of the store. It's only because it was the best thing that was available a few centuries ago. But here we are at a time when electronics has allowed us to separate the human interface from the sound generator. So the human interface now can, God forbid, be optimized for the human and not for the generation of sound. So that means that what I did in Lindström is I said, regardless of what sound you're trying to create, what is the ideal human interface for performing music? And Lindström is my best guess at it. How is it better than a computer? I mean, the computer has so many well-understood interfaces.

How does music come out better with this? Well, computer, first of all, the input surface is a typewriter keyboard. And I think it would be hard to argue that a typewriter keyboard is a good way to play music. It's not even a good way to type. It's not even a good way to type? And so you want different human interfaces for different human activities, right? So for playing music, what you need is that ability to ideally express the sound in those three different parameters that I spoke of. One is the loudness of the note over time between the start and finish. One is the pitch of the note, vibratos, sliding between pitches, bending notes, things like that.

And one is the timbre, varying the timbre to make it more interesting during performance. And so to do that, by convenience in this particular universe, we have three dimensions. And so it's very easy to map those three dimensions to those three musical parameters. And so you want something where you can, you want a surface that can recognize the pressure of your finger, the x-axis, left-right position of your finger, and the y-axis, forward-backward position of your finger. And you want this to be polyphonic. You want to be able to independently recognize the three-dimensional position continuously of all fingers that are

touching the surface at one time. And so this is why something like an instrument or one of the other expressive controllers is a better interface for playing music than, for example, a bunch of on-off switches like the computer keyboard or even the on-off switches of an electric piano keyboard. Going the other way, can I use this as my keyboard? Because I'm pretty sure I could map that into something really fantastic with this sort of interface. You mean, for example, using the instrument to be your QWERTY keyboard? Yes, with this level of of flexibility you could

embed a lot more information into the keyboard itself although this goes back to trained trained instruments or trained musicians uh because this isn't a beginner instrument in order to get the most out of this you really have to understand it learn it as a beginner instrument. In order to get the most out of this, you really have to understand it. Learn it as a new instrument. Well, you know, you make a very good point. Certainly someone could create a query to keyboard that uses the gestures we all know. And for example, the keys are in the same position. But instead of a caps key, for example, you just press a little bit harder to get caps.

And then, for example, instead of the alt key, you would just slide your finger forward a little bit, something like that. Musicians who play acoustic instruments tend to get a degree of pride from, oh, I figured out, as you say, this archaic physical interface, and I can play it with competence, and look what I can do. Do you find that's like a

barrier to introducing a new music human interface? Yeah, not all of it is attitude or a willingness to accept new ideas. A lot of it is just the fact that humans tend to learn new physical gestures very slowly to develop that muscle memory. So, for example, if you've played a standard piano keyboard for 20 years, you've developed a certain amount of skill. And that, I think, first of all, it makes it difficult for someone to start from scratch on a new human interface. But the other thing, too, is if had I played a piano for 20 years, I developed a certain amount of skill and I enjoy having that skill. And I don't want to lose all those gestures that I've practiced by going to a new interface. So how do you convince people to come to your interface?

Well, a lot of the instrument players, or as I call them, instrumentalists, are people who tend to be open to new ideas. A lot of them are people who don't have so much musical skill, but they respect what I'm doing. And they're saying, here's an ideal new man-machine interface. And I would prefer to learn using this because it has advantages. It is not hard to play like a violin, but you can more easily get a violin sound. And just to elaborate on that, not only are the notes, the notepads in an instrument spaced the same as the average pitch of fingers on the human hand, but you're aided by the software, because no matter where you initially strike your finger within the three quarter inch pad, it starts out being in tune.

And then it does some other tricks, like if you slide from a C note up to an F note, and then you stop. If you're a little bit out of tune, once it recognizes you've stopped, it gradually slides you back on to tune. So this makes it, it's like having frets and no frets. Before you touch the string, the instrument, the roll, the frets are on. But then once you touch it, it takes the frets off. And then once you finish a slide or a vibrato, it puts sort of soft frets back on. So this is an advantage. What it does is it isolates the hard parts of learning an instrument from the easy parts and the things that you want to happen. What you really want is you want to be able to develop that expressive skill that makes people cry, that makes people laugh, that makes people feel joy. But you don't necessarily want to make it hard for them to do something

if the hard aspect is not necessary in order to create that joy. And that's the purpose of Lentzschmidt. I like this. I like the accessibility of it. Yeah, let's not make things difficult just because that's what we've always done. Yeah, yeah, I'm not trying to make it hard for people. I'm just saying that there are certain things that are required to be appreciated at large in our culture and get people to actually buy a ticket to sit down and watch you do a concert. And that is you have to develop skills that move them emotionally. And if everybody can do it, then nobody wants to pay to go see you do it by yourself.

People will go to a dance concert and they'll enjoy having you perform in the background while they're dancing and drinking their Red Bull and vodka. Or, for example, someone would love to go to a movie and see a movie score, but few people will actually pay to sit down in a concert where someone performs their movie score. The same thing in pop music. One thing that's happened over the past 20 years is the concept of an instrumental solo has virtually disappeared from popular music that is electronically generated. For example, electropop in its various forms, or hip-hop. There's no more solos, because everyone's playing music with on-off switches, either MIDI keyboards or drum machine pads. And so people have generally said, that's very interesting for background music, but it's not very interesting for foreground music. In other words, I wouldn't actually pay to sit down to see a concert of someone performing this, but I would certainly love to hear it in the background. And so what's happening is all this electronic music,

by virtue of the fact that the human interface is a bunch of switches, has been relegated to the dustbin of background music. And what we're missing are foreground artists. The equivalents, the modern equivalents in contemporary music of the Jimi Hendrix's, the Django Reinhardt's, the Miles Davis's, the Ixac Perlman's, those sorts of people. I'm probably going a little bit away from the embedded stuff. That's fine. They're used to it by now. Yeah. they're used to it by now yeah most of us don't work with artists as our customers and uh they have a reputation musicians also have a reputation how is it different to have them as your customers well you know it's um in my case what's interesting about an instrument is it tends to draw people who love ideas.

Because they buy into my reasoning behind the design of this. And they say, I agree with that. I'm going to use this and choose this as my instrument. I'm going to be an instrumentalist. The type of people that value ideas are really cool people. I've met some wonderful friends and they're all over the world. They're not necessarily in major cities. Often they're out in, if you look at the map, they're outside of cities, but they can be in

cities too. It's just there are people that love ideas. For example, also, I get a lot of people buying an instrument who are buyers of a stringed instrument called the Chapman stick. Oh, yes. Okay. I was going to ask about that. Thank you for making the connection. So that's the same sort of type of person who is excited about the idea of Emmett Chapman having created a better stringed instrument that are drawn towards what they hopefully believe is what I've created is a better electronic instrument. And they're humble people. They tend not to ask

questions. For example, I have another product for guitar players called Adrenaline, which sort of takes a whole bunch of, in software, equivalents of modular synthesizers, puts them into a little guitar pedal so you can do all kinds of cool stuff. And guitar players, I've noticed, they have a certain behavior. What they tend to do is if they don't understand something, the first thing to do is call you while they're holding the phone between their ear and shoulder and opening the manual. And so, listening players, what they tend to do is they tend to have

understood the concept of delayed gratification. And they'll say, I don't understand something. I think I'll take a look on the website and look at the FAQs, look at the manual, which I think is fairly well written, and maybe post something on the forum. And they're really nice people. And when they call up, they're humble, and maybe post something on the forum. And they're really nice people. And when they call up, they're humble. And they're just really cool people. And I couldn't ask for a better set of people to work with. And as I said, I've made many friends. Your business is relatively small. Do you have any advice for small technology companies? I mean, small as in it's mostly you,

right? Yes, it's actually not only mostly me, it's only me. I have no employees. I have no office. I work out of one room in my house. However, I have a lot of outsourcing. For example, I use a manufacturer south of San Francisco to not only manufacture my products, but also to drop ship them. So for example, I'll come in in the morning and I'll get some orders by email from my automated websites. And I'll just enter the order and then send over the shipping order to the manufacturer. And even though I have about 18,000 customers, as I say, most of them, I think, are pretty good about understanding the product. And I try to make products that are not hard to understand and so that people can focus on music and not on engineering. And so there aren't that many questions that come up.

And usually, if there are questions, they're very quickly answered on the website with the information that's all there. And so I'm able to do it myself. And actually, I tried having employees for a while. I thought I only needed one person at a time. I went through three different people. And I found it was just quicker to do it myself. Now, I will say that, yeah, I'm 62 and I've been doing this for a long time. And so a lot of things I can do very quickly because I've done them before.

One thing I would say is that one thing I've learned about having a small business is that 90% of the time is spent fixing mistakes. And if you simply think through things carefully and don't make mistakes, there's very little work to do. Again, good advice for software, yes. No, just start coding. And the other thing, too, is software is a beautiful art form. And I'm only so good at it. And I can't do everything and yet get the depth of knowledge

and concentration that is needed to do good software. And so I usually start things out and then I have a brilliant engineer who lives in Belgium. His name is Gert Bevan or if you're Belgian his name would be pronounced something like Gert. And he when I started working with him I recognized that this is someone extraordinary. And so the first thing I did was cut him in on the action. Every time an instrument is sold, he gets a piece of the action forevermore because he was that valuable. So I would say that software is complex. And if you're not a software engineer and you're working with someone, when you find somebody good, never let them go.

Do you have advice for other aspiring instrument creators? Well, I would say often what people email to me is they say, I have a wonderful idea, and I'm sure it would make a million dollars. And most of the time, they are very passionate about their idea, but only from their point of view. And so, I think they usually don't realize that the reason that their instrument has not been made is either it's too expensive. For example, someone would say, you know, I've got the perfect idea for an instrument. If you take this $5,000 instrument and combine it with the features of this $7,000 instrument and make it for $50, it would sell like hotcakes. Or another thing is they'll say, you know, I want something here to do 47-tet microtonal music. And I'm sure once people understood how good it is, everyone would play this way.

That means dividing an octave into 47 equal uh pitches well not a big market for that yeah a big market for that and there are merits to it and i'm sure from one's point of view it could seem like that is the case but it's not and uh and generally i think if you just look at the world of music, you can find out what the public generally embraces. And moving too far from that is a good exercise in finding people who share your value of that particular idea. But the mistake that's too easily made is someone takes that idea and they say, could you make a prototype of my idea? And I say, well, I'm very sorry. You have a very wonderful idea, I'm sure, but I only make prototype of my idea? And I say, well, I'm very sorry. You have a very

wonderful idea, I'm sure, but I only make products of my own design. And what I've done is I've received this question so many times that I have a page on my site on the FAQ page for the instrument. And one of the questions is, I have an excellent idea and I'd like to prototype it. Can you give me any advice? And so I have a page that has this little essay, and it starts out and says, of all the ways to lose all your money, one of the most effective is to make a prototype of your product idea. And so often my fond people will show up at these trade shows called NAMM, N-A-M-M, which for some odd reason stands for International Music Products Association. And they'll have rented a booth there for $3,000. And they have made their prototype. And they're

expecting to sell millions of these things. And you see them sitting there by themselves, and no one comes by. And it's kind of a shame, because they borrowed all their parents' money to do this. And their parents are sitting there suffering in their newly rented apartment, and this person has wasted all the money. And so on that page, I suggest in electronic music products, there are various ways to prototype your idea yourself. You can take all kinds of little human interface elements that are very cheap, like Korg has a line called the nano keyboards or others. A guy named Keith McMillan has some other low-cost things. And then you can use some software called Max MSP, which is very musician-friendly,

where you connect the dots together with little lines, and you can do signal processing. Or there's a free version of it called PD, Pure Data. And if you set yourself a budget of $1,000 and just connect these things together to make something that's as close as possible to your instrument, you can do a proof of concept, and then you can show it to other people and see what they think. I would suggest doing that. And it's a great time to do that

because there are so many little elements of instruments today that you can put together and connect via the MIDI cables and software and things like that. So many things have gotten cheaper, especially over your career. I mean, it used to be but 3D buttons on a system. Do you see other sensors or technologies that are going to enable interesting instruments? Well, actually, to my knowledge, there are no 3D buttons. There are some joystick-type keys, which are 2D, but I'm not aware of any button that did what I tried to do, because I looked. And there may be something now that's on the market, but the best I could find were pressure-sensitive trackpads. But your question was, is it cool because they're so cheap or something like that, right? I'm sorry, I missed it.

Oh, my question was really, are there newer technologies? Are there different sensors, different LEDs, different things coming out that are interesting to you for future instruments? Oh, very much so. Yeah. I'm working on a new drum machine. And one of the problems with what I do is most things are fairly low volume. I don't sell millions every day. And so what I have to do is try to piggyback onto existing economies of scale. And one of the great ones is Raspberry Pi. And by the way, the instrument was created based on the arduino due circuit which is the due means two in italian and it's uh it's based on an 84 megahertz arm chip and basically i just took that same circuit and then put all my analog

switches and scanning technology uh on it uh connected with some glue to the those circuits now what i'm doing for this new drum machine, which needs to make sound as well, is I'm piggybacking it on the Raspberry Pi 3 technology. And what's a very interesting thing for people who are designing embedded systems is the Raspberry Pi, which I'm sure a number of your listeners are familiar with, comes in a different form factor for embedded use, and they call that the compute module. And what it is, it's a little card, it's maybe four inches by

an inch and a half, and it fits into a standard DIMM memory slot with 200 edge pins on it. And what this is, is the whole Raspberry Pi circuit, which you can buy for about $25 each in quantities of 100. And so I'm just going to make my motherboard of my drum machine and have a slot for this little Raspberry Pi compute module on it. Because if I had tried to do this myself, it would cost me probably $70 or $80 in parts. Even if that, it'd probably be more. This compute module has everything the Pi has. It's got the Broadcom 4-core 1.2 gigahertz ARM chip, which is a cell phone class processor. It's got a gig of RAM, and it's got four gigs of flash, all for $25.

And it works out very perfectly in an embedded product. And so what I can do is I can base my system on Linux and I can use, if you have any Linux users out there, you use a function called build root. So you can build a stripped down minimal Linux, which doesn't have all the overhead that Linux would normally have. So it's more responsive. And you can base a product, an embedded product on Linux, which is much, much easier than if you were to try and start from scratch with like the Kyle tools on ARM or other open source free tools where you have to find yourself an RTOS over here, find yourself a Flash operating system over here,

find yourself this and that and piece them together because Linux has all that stuff to start with so i'm very excited about it the nice thing about it is with four cores of 1.2 gigahertz i'm able to do audio sound generation so now i can just look for linux tools or open source software to help me with sound generation and with sequencing and some of the other things i need to do. I think for a small production system, we sometimes forget that the idea is make it cheap to create. If you're going to end up making a lot of them, then you can start cost-reducing and removing the Raspberry Pi

and putting something cheaper in. But if you're looking at making hundreds or thousands hundreds or thousands yeah especially at a good price point that that you know that 25 isn't killing your margin um yeah that makes perfect sense well my product said to be more expensive for example instrument there are two sizes one's a thousand bucks and one's 1500 bucks and the drum machine will be in that same price range that I'm making. But what I do is, my specialty is not making the cheapest things and not making the most expensive

things, but I have sort of a nice little niche that's somewhere in the middle there which appeals more to professionals or to more skilled musicians, more higher skilled musicians anyway. And they're willing to pay a bit more for that. But these products are not quite ready for the mass market. You know, most people who are playing music will look at an instrument and say, huh, I don't know what this is. You can move your finger around.

Why would you want to do that? So I don't have to appeal to everybody, but by keeping my expenses low and having a wife who works at Apple, I'm able to have a fairly good living. Yes, I'm all in favor of full-time working spouses. Yes, yes. You mentioned the silicon backing and molding it. Sorry, this was a long time ago and i i just i have this question i want to ask before we close do you um how are you doing that silicon are you injection molding

it are you do you have well i always find a vendor who does that sort of stuff because it's not the sort of thing for my contract manufacturer i just want them to be able to assemble uh commodity pieces together circuit boards parts blah blah, blah, blah, blah, blah. In fact, I try to design circuit boards, if I can, to use all parts out of the DigiKey website or out of the Mouser website so that I know that no one part going missing or going obsolete is going to stop my production, right? But for certain custom parts, like the rubber touch surface, which is molded, I have to find a vendor for that. So for example, I used a prototype shop in Silicon Valley, where I'm located, to make a prototype with a vacuum process first, etching a, I created a 3D file, and I told them just to etch that in negative so that you could pour the silicon into it,

and it would make a prototype. And once I proved the the concept then i took it to my vendor and they charged a bit more to make a full aluminum mold out of it from which they shoot the silicon and they just deliver me um a big box of these sheets every month and then do the leds get put onto the board under the force sensitive resistors or above or in somehow in this? Well, the LEDs are on the circuit board. And then on top of the circuit board is the touch sensing sheet, which uses the force, the printed force sensor, force sensing resistor ink. And that's fabricated by a company called Tangio, and they're located in British Columbia. And then on top of that is the silicon sheet, which is fabricated by a company which I want to give no more business to because they're horrible. And then on top of that is a metal panel. And this makes it very easy to repair. I'm sure if I was on iFixit, I'd get a 10.

But I'm below the radar. So yeah, I try to make it as easy to assemble as possible. I try to, basically what I try to do as a small time customer of my contract manufacturer, who's quite big, they're called Lima Electronics. And they're in south of San Francisco in Brisbane. They're excellent. But because I'm a small customer, I try to make myself as attractive to them as a customer. So I do all the work for them I can. And whenever they're about to release a production run, I go there and I do a final QA myself. So I try to quantify it into test instructions so that the people there who aren't musicians can make sure that it comes out perfect. And then when I do my QA, it's sort of the final test.

But I try to make it as easy as possible for them, no headaches as possible. So therefore, they'll like me as a customer and continue to make the products for me, which is great for me. And when they actually have far larger customers, it would be easier because they'd make more money with them. Okay, so you do QA. You're a test engineer because you really need to be. You do the hardware. You do the software. And the R&D. Most of the software. Maybe not all the software, but most of the software. The R&D, the mechanical, the concept, the business, the marketing.

Are you tired? Well, they say when you own your own business, you work half days, 12 hours. The other one I heard that I liked was if you own your own business, you get to choose whichever 18 hours you choose to work per day. I like the if you own your own business, your boss is a jerk. Yeah, internal conflict. No, I'm not. But the thing is, it does also come back to, well, two things.

Number one, that if you don't make any mistakes, it actually, you can be quite efficient every day. And the other thing is, and so when things are getting hectic, there's a tremendous temptation to just power through things, which often results in making more mistakes. And so if you just stop, take a breath, and you'll find that you can usually find a solution which corrects things for the future. So I tend to embrace the concept of delayed gratification. Now, that was the first thing. I said there are two things. Number two is what I said before. It helps to have a wife who is a manager of engineering project managers at Apple. She's my angel investor, effectively. You mentioned the instrument, which is the expressive control. We've talked a lot about that. You mentioned the

guitar pedal adrenaline. I knew there was a lin in it. There's one more thing you've been working on. Oh, I've been working on a new drum machine. I've been resisting my destiny for years now. Everyone says, please make another drum machine. I say, I can do drum machines

in my sleep. It bores me to tears. But by using this three-dimensional sensing of Linstrument and applying it to the idea of creating beats, you actually get some very creative ideas in there. And so that's, for example, back in 1984, I invented a feature called note repeat, which others called roll. And what that does is if you have this function on and you just vary the pressure of one of the drum pads, it takes that instantaneous pressure every 16th note, for example, and generates a new 16th note at that loudness of a note. So if you vary the pressure, the resulting sound will be sort of a and so in this new drum machine I'm working on, it's sort of like that note repeat feature on steroids. And by the way, that note repeat feature was a very important feature in the

MPC drum machines. I originally created a drum machine I made in 1984 called the Lin 9000. And then it was in the MPC drum machines. And then the Ableton company came out about three or four years ago with a product called Push, where they introduced this new innovation called pressure sensitivity. And I had to have a little bit of a chuckle because I did that in 1984. But it's still, so it's finally ready for primetime, I guess. But at any rate, so the idea is that I'm applying not only this pressure sensing, but that full 3D sensing to the creation of beats, and it'll make some very, very cool beats.

Now I've got it all playing in my head, and I have all these ideas and no way to get them out. Well, you know, which you can do to the audience, if you want to hear what this does, and this is semi-plug, but you can go to my site, which is rogerlindesign.com, R-O-G-E-R-L-I-N-N, design, and under the instrument menu, you'll see an option for videos, and there's about 60 or 70 videos there and you get an idea and among these are this ability to play drum beats cool well i think we have kept roger as long as as we should um although i now i want to go over and ask him to show me all of the things because i just want

to play with them uh christopher do you have any questions before we go? I did, and now I've lost it. Oh. Yes. So one thing that occurred to me with these new user interface designs is there's kind of an opportunity for accessibility, like making instruments that maybe appeal to people with disabilities. Have you thought about that or worked on that at all? You know, my selfishness i

never considered it when i was making it and then in 2015 i was contacted by an organization philanthropic organization in the uk called the one-handed musical instrument trust and they were their mission is to create it turns out all the orchestral instruments require two hands to play and so they have all these weird mechanical instruments that, for example, you can play a wind instrument with one hand that people have made. And they give an award every year. Well, in 2015, they awarded Lindstrom the most playable instrument with one hand of any instrument, acoustic or electronic.

And that was very nice to see. And they flew me to Bristol, England to accept the award. And in doing so, I met a number of disabled musicians. And it was very touching because these are people that would love to play instruments, but they cannot. And there are other disabilities as well. Someone has arthritis, something like that. Well, as a side product of an instrument's ability to very efficiently control expression with one finger, it allows someone to more efficiently play.

Use those things. In other words, if you're playing a violin sound, you don't have to have one hand that's developing a bow technique and the other one which is being held near your ear and your chin, for example. And so it makes it more efficient, this more efficient use of body gestures. And so as a result of that, I now offer a 20% discount to any disabled musicians if it can help them in some way. That is so awesome to have it be, I mean, it's kind of a side product of your goals of making better human instrument

interactions. And yet, that's so wonderful that someone who might have lost the ability now has it. Yeah, it feels very good. It was a wonderful visit to this celebration in Bristol, and I met quite a number of disabled musicians who were very excited about not only my instrument, but some of the other instruments that people have made that have won awards in this competition that the One-Handed Musical Instrument Trust does. There are other people with various disabilities.

For example, they don't have enough strength in their hands, so they need to be able to increase the sensitivity to pressure, for example. So it is very gratifying. One thing I should mention as we get to the end there is mine is not the only instrument that's what's called an expressive controller. There's a wonderful instrument made by an English company called ROLI, and they call it the Seaboard, S-E-A-B-O-A-R-D. There's also a phenomenal instrument, which is actually the first of these polyphonic

expressive instruments called the Continuum, made by a guy named Professor Lippold Hawken. There's another one called the Soundplane, and another one called the Eigenharp, one also out of England. And so this is sort of a revolution that's just bubbling under. And in my view, in about 10 years, people will recognize how limiting playing music with on-off switches is, and they're going to be embracing these new instruments. I think you just ended on a fantastic note, and yet I'm still going to ask you, do you have any other thoughts you'd like to leave us with oh that's about it i i um uh the only thing i would say is is that um uh i think we did mention that the the instrument uh software is open source and so i found that

i think my sales have increased by probably 20 because a lot of people who are musicians but also programmers love the idea of it as an an tablet, and they use it to create their own instruments by modifying the code or to use our user interaction mode where it becomes sort of a smart, dumb terminal, and they can write computer software for it, for example. And so if you're a programmer and you do love music and you have a creative streak, it might be something you want to check out on the website. They're going to kill us because we keep telling them about fantastic instruments. Yeah, that's okay. All our listeners will be gone soon. It'll work out. But until they've gone, I guess has been Roger Lynn,

founder and CEO of Roger Lynn design and janitor and, and janitor winner of the 2011 technical Grammy. You can check out Roger's company at Roger Lynn design.com. Link will of course be in the show notes along with links to his, uh, instruments and some of the other ones he's listed. Thank you for being with us, Roger. It's my pleasure. Thank you very much.

Thank you to Christopher for producing and co-hosting. And of course, thank you for listening. You can always contact us at show at embedded.fm or hit the contact link on the Embedded FM website. Now, a quote to leave you with. Music expresses that which cannot be put into words and that which cannot remain silent. That's from Victor Hugo. I liked it. Embedded is an independently produced radio show that focuses on the many aspects of engineering. It is a production of Logical Elegance, an embedded software consulting company in California. If there are advertisements in the show, we did not put them there and do not receive money from them. At this time, our sponsors are Logical Elegance and listeners like you.