The Pocket Organ: An Open Source Musical Instrument
12:57PM Jul 30, 2020
Hello everyone and welcome back for day six of hope 2020 Boyd's been flying by so far. A big thank you to all our attendees presenters and volunteers who've made this virtual session of our conference so successful that you we could not do this and we wouldn't be here. So, thank you. We have another incredible round of talks and workshops lined up for you today. Our first session is from Thomas Tempe, a migrant of migrant the global village. originally from France Thomas has been living in China for the past 10 years where he stays active in the hacker and maker maker communities. So we'll be talking to us today about the pocket organ and open source musical instrument. Please remember that we will have a hosted live q&a session after Thomas's talk with Thomas, please submit your questions the matrix chat window.
Get away Thomas.
Hey everybody my name is Thomas Dante, and today I will be telling you about the pocket organ, which is my plan for world domination.
There we go. A top 2020.
Let me start with a question. What is world domination. I say, Make me 1 million of these and credit for the work
that is my wish.
My next question is, What is our musical instrument. The answer is twofold. On the one hand, it's a device for generating fun ways. It makes music, and the other definition is. It's a user interface for making music. So, if I like if I look at the first definition. I take okay here's a kazoo
can blow air in it
Being ease of learning
wrench. So wrench means a common the octaves you can play. For example, channel can be very wide can have 5678 updates. And then if you take a look at a harmonica it only has, like, three octaves but not all of them can. The lower octava has holes in it so excited it's mainly more like two and a half octaves.
accompaniment. So that tells you whether you can play chords with your instrument. Multiple notes at a time, and it can also tell you whether you can use the instrument, while singing which unfortunately for the harmonica game is nuts. It's not possible. So on the other criteria we have one of them is movability. So, this one fits in your pocket or Pelican cangzhou It fits in your pockets. If you take a guitar it fits in a special design back. And you can take it on your back. The yellow you can take it along when you move, and then the church organ you cannot take it along, even when you move. And then there's one more criteria that I feel is phenomenal. And that is expression. So expression decides determines whether you can modulate the sound they're making. So if I take a candle, I cannot on the date much all I can do is press stronger or lighter to make more or less to give more or less loudness to myself. And the other thing I can do is control the duration of the song. I press it and then I release it. And that's all it started when it when you let off the sound. You can no longer influence it. But if I take a lot of wind instruments, they are actually incredibly expressive. So, take the kazoo. And let's make, give it a try.
Technically, expression is about relating the pitch, the loudness, and the timbre, actually, for the musician is a way of conveying emotions, that's much more powerful than what you can achieve with just a melody. When it comes to a tenant, if I look at modern electronic music instruments. They're often not very expressive. So I leave out the guitar the electric guitar because it really is making sound it's just like you pick it up with a special, special purpose microphone. If I look at a lot of the MIDI controllers that are out there. And most of them, like, look at the MIDI matrix controllers, so you have a bunch of buttons. And they're usually programmable, which is another way of make of saying, Let's delegate the instruments designed part of the device to the user. Like the user has to figure out how to turn a set of buttons into an instrument. And they're usually using on off buttons. So you turn it, you press it down it closes the circuit and then you have a note and then you release it and then it opens the circuit and then the node stops. And that is, you cannot. That is kind of weak in terms of expression. My ambition for a pocket organ is to give it a lot of expression, and it's quite a technical challenge. So it's not working yet, but let me. Nonetheless, give you a demo.
set the keys have to have some form of analog detection of where your finger sits and be able to detect if I slide it right, a little bit to the side, or if I get away from from the, from the surface. And this is not working yet. My premiere situation was with pressure sensitive kids. And if you look at, like Kickstarter projects for example you'll find a few very interesting instruments, which are detecting finger pressure. So I moved away from the technology not that it's bad, or unfit for purpose, but it's not very well matched to the very small size of my target size of my instrument. Remember, pocket organ means it should fit in your pockets. And so basically not be no larger than a smartphone. And if you want to have pressure sensitive back a lot of those pressure sensitive sensors you actually need to give it. Give them quite a lot of pressure, which is fine for a bigger instrument like to do well. but it's a little high for
forces something that is so small.
There's one more thing I'd like to address regarding a good instrument. And that is, whether it is an instrument at all or whether it is a tie. And I think there are a lot of things out there that we would love to be instruments, but actually they are test ties. So let me take one example, though, well, maybe you can think of the speakerphone. It's really fun that you can only play one note at a time, and you need the visual control loop, to be able to play, then have the tactile control loop. So that means that you can never become too old on it, you cannot train your muscle memory to actually play faster or better. So, another more obvious example of a tie is the roll up, piano, so maybe you've seen those and they're just maybe in an airport or on internet. And the idea is this a fabulous. And the sad thing is that the realization is just to facility. So, one thing is the, the delay is hundreds of milliseconds. So, between the time you think you've hit the key on your finger thinks it's the key, the time you hear, hear something has way too much time for you to build up the neural pathway for moments automation and muscle memory. And the second thing that is so poorly realized about the role of keyboard. They run a channel is that you can never tell from the finger sensation whether you actually hit the key right or not dependent on where actually you press it falls left or right. It's hit or miss. So what makes a real instrument versus a tie. A lot of people just read the question and go for the most timeless instruments that you cannot go wrong with a piano or
I'd like to explore this a little bit. What's, where do you put the breakpoint between the tie in an instrument. So if you take the harmonica. It really looks like a toy it's constructed like a tire, it costs the price point is that over time. 1020 bucks. And it's got three octaves a very small wrench like a toy, and it's missing, there are holes in the, in the wrench, like, so tight. And yet, if you look from a musician's perspective. First of all he does. It has incredible expressivity. And a lot of people have dedicated their whole career to playing the harmonica and gain tremendous recognition from your peers. So from a musician's perspective, it really is
a new instrument.
So in a way, it's a very limited instruments, but didn't those limits
stimulate the creativity of the players, and make for a greater instrument.
So if you had had a full chromatic harmonica, would it have made the instrument better or worse, and the thing is we actually have chromatic harmonica. And that so many people are playing it. And so, my take on it is that it's fine to embrace constraints. When designing an instrument. Either way, the harmonica, and the pocket organ are very much like it from the, like, construction and price points. The it's from the continuous expression type of playing being from the packability. And so much so actually that's the default sound for that pocket organ so far is read instruments, like the harmonica. So I don't think I can make an instrument that is as great as the harmonica. But I can certainly make one that is missing some of the holes that the harmonica design is living, and one of them is the ability to played with headphones. And another one is the ability to play as your sing. I was certainly will not equal its receptivity or expression, but I can certainly do a little better in its ease of learning and the way that it encourages the players, understanding of music theory.
To get to that point with the guitar. Nothing higher. So, what about we take this time from one month to compress it to one hour.
Think it's possible. Let's give it a try.
take it better.
The minute you take into you.
The new Casta.
One question I hear quite often is, why don't you just develop it as an app for your smartphone. And the answer is, actually. You could, but there are a few challenges to this. First of all that, don't you really get a good tactile feedback. You know, you hit the screen, but you don't know if you're hitting the middle of the key or the side of the. So one solution could be to have some kind of shell that you click on, on the top of the screen, that gives you this finger proper finger location information. And the next challenge would be to have a good latency, and maybe that would be possible. I haven't experimented with it. I'd be really curious to hear but I attempt in this direction. So for now, it will be a technical device. If I take a list of all the music that sounds good on the channel is practically limitless. But they all take too many fingers to play like five or 10 fingers at the same time. If I take music that you can play on a piano with the first finger, with one finger. It's also a spit. They don't sound good. So, the ID is to pull from the space of music that sounds good to some of them on to the first finger.
That is something that courts can help us to because courts are this kind of musical item that just aren't so numerous aren't so demanding and sound good and they can play with a single finger. So once I've had the first finger, and maybe I can have a second finger. And that kind of puts a little more difficult things on it, that will expand my scope as a musician. So, and the in the case of the pocket argument. The answer is, I want to put various chord shapes like the minor key. The
sevens and so on.
And then when I bring it a little higher, in terms of player maturity. And I can add some things like sharps accidentals and expression.
Let's say I start with a major chord.
And then I turn it into a minor chord.
Maybe I add the seventh
And then some more exotic chord shapes like if I want it,
augmented or diminished.
You see increasing chord shapes complexity mean more fingers. Of course you can still play a melody, just press the middle button here.
And you can have a court and then below the on top,
not happy with the sound, you can press the instrument key, and choose a different next a chromatic percussion.
For quick chrome line just press the chrome button. And
now's the time to mention the looper. The looper lets you record the sequence of notes from one instrument and play them over and over again. Now it gets interesting when you start stacking multiple loops from different instruments on top of each other. The pocket organ lets you record up to eight different loops, one for each of the buttons here. And then you can turn them on and off at your convenience. And it's built in such a way that they always stay in sync. So even if you turn one loop off and then back on, it will keep playing together with the other loops.
In case you're doing accompaniments for seniors. You probably don't want to use the looper, they're probably not on time that you really want to follow them. When you're jamming with friends. However, the looper might actually proved to be very useful because it gives you a very predictable and written pattern and chord progression that other musicians can have fun on top of you might have noticed metronome ticks, as I was recording the looks. So this is a trick to let the instrument know about the musicians rhythm is useful for one thing because it lets the instrument quantize, the loop duration, so that they always fall very sharply on time. And, otherwise it also lets you do some cool tricks like changing the tempo. After recording and might be useful in the future for things like arpeggios
the screen here allows you to go through multiple levels of manual, killing area but giving you immediate feedback. When you play a chord is going to display the name of that code in big fat capital letters and nothing else. And then when you also do things like setting the volume or other features is going to give you immediate feedback. So that means that in terms of UI philosophy, we are also stick stepping close to the same principle as we had for music discovery, you're going through, increasing levels of complexity including maybe mood time like key combinations. And this.
You get immediate feedback through the display.
Let's have a look inside
the pocket organ is built around a custom PCB with a microcontroller running micro Python micro Python is this fabulous open source project. So when you plug the USB ports, into a computer you instantly see a drive USB drive with all the PI source files in it. So that's the glutamate in packability. Right now I'm using the STM 32, f four or five microcontroller, which is a very beefy microcontroller. So the beauty of micro Python is that I can port it to a different microcontroller with very limited porting effort, the PCB also has some embedded memory and accelerometer, and all that display. And the rest is capacitive touch sensors for the kids. The keyboard is actually the most tricky part of the project because monitors cannot tie you and other bands that you can use directly on an instrument. You have to somehow make them yourself. In a previous iteration, I used pressure sensitive silicon keys with something called Villa stat which is a piezo resistive plastic meaning that the harder you press the lower the resistance. So you just put a sheet of plastic between the PCB and the silicone keyboard, and you measure an analog resistance. The problem with this technology works, some instruments use it successfully. But the pressure you have to apply to get a detectable signal is actually pretty high. So that's fine for larger instruments like the duelo, but it's a new high not ideally suited for such a small factor instrument, as the pocket organ. So now I'm trying to develop a capacitive sensing keyboard. I've been taking inspiration from the CPU, which cause it's possible to have those very responsive
keys. Let me give you a demo.
In terms of pool chain, I have three different tools I'm using for the development. First is the code itself. The software is hosted on GitHub, along with a page, explaining the project. The 3d design, the enclosure and so on are on fusion 360, which allow public sharing including an offer and
free hacker license.
And the PCB design is hosted on the Tron eta, which is the Chinese SR version of Eazy E da, which also offers free hosting and public publication of open source projects, so you can see the PCB you can clone the project and you can make your own. Now I'd like to talk about the open source nature of the project. So this is a hardware project right. And what when you complete an open source project, it means you've written all the code, fixed all the bugs, published it somewhere on GitHub and people can just download and use it for hardware. That doesn't quite work out, right. You cannot let users download a copy of the instrument and plate. You have to somehow manufacturer it before putting it in the hands of individual parents. So, this is where the million units, challenge comes in. 400 to come to life, you have to have a factory. You have to have suppliers, you have to have distribution network marketing. Those are all things that are usually done for profits and not well suited to the scaling down to the level of an individual hack also considering that I have a day job, which I'm not planning to quit. So, this puts me in this strange and interesting position of trying to make. Open Source hardware and then developing the project and then handing it over to someone who will take care of all those issues. The typical business case for another instrument is high price, low volume. And that's exactly why it's probably not going to make it into the hands of many beginners. In addition, those instruments are monopolies of their respective owners, being protected by patents and copyrights and trademarks. And that's why I believe there is a niche for an open source open source, low costs, musical instrument, we just have to somehow crack the problem of manufacturing and distribution. There is a crowd that's making very low price high volume products, though. And these are the toy manufacturers. Actually there's a different crowd which are Chinese copycats. So these are not very regardant on intellectual property. They're into making whatever they making cheaper and higher volume, so that they can make more profit. Those people are the ones making millions of roll up pianos. So, why is it that nobody made a good roll up channel. Well, turns out the market is already saturated with very cheap solar panels, and they don't care about music, they're just here for the volume. So what if I was someone who was able to serve them with a design that's for something that is a valuable instrument like the pocket organ. And then, their job would be to manufactured it, and to sell it at the lowest possible price, with the highest possible volume. The 10 minutes with working with copycats, is that you cannot really control them. They might be tempted to say remove expression, just to save on the manufacturing costs. So one solution could be to register the pocket or can brand, and then granted us only to those who respect a specific set of requirements. While I do believe in open source. And I do think it makes the most sense in the design phase, it could actually become a hindrance. In, looking for manufacturing partner. So, as such, I'm willing to, at one point close by to the source to encourage third parties to invest in manufacturing. If I cannot find a manufacturing partner then I might just run limited volume crowdfunding campaign, and then maybe I just do the manufacturing engineering myself. And then I sell assembly tricks for small scale manufacturers to maybe sell on Tindy the product itself is actually quite simple to assemble. It's just a one sided SMT PCB two plastic shells. A battery and a few screws.
And then the cost of the, of the components right now is around $30 per instrument for prototypes and if I can bring it down to 10 or $12 per for small series from podcasts, then I can have a shelf price of 150 dollars. That would be my target. And now my call to action. If you're a manufacturer, please contact me if you want to be an early adopter, maybe they would the product. Maybe contributes a few lines of code, be welcome on the GitHub. And if you're really just waiting for an excuse to learn music. You're welcome to follow the project on Facebook, and then you'll be informed whenever it goes on actual sale. In any case, it was my pleasure talking to you. I know Wish you fabulous. Hope 2020. Time for the q&a.
your cloud your life.
Hey thanks Thomas, that was a fascinating presentation so tell us Have you always been musically inclined or what was it that inspired you to really build a pocket organ.
Its way. Yep. But, oh thanks to hearing myself Sorry about that.
So, yeah, I was one of those guys. I, I wanted to learn music but I always got put off by the difficulty of getting into instruments like the piano. And so, and then I always felt like I can tweak do things a little better as crushing my age.
tell us a little bit how you sourced all the components for the pocket Oregon what was that process like I mean you live in China, so I assume it's probably a little easier for you than it is for any of us here who would be trying to do that but tell us about that and then maybe you can give us some advice to people who are looking to manufacture their own devices, about how they can get started in that and find supply chain.
Oh, that's a very interesting question. Actually I learned about PCB design and so on. At the China. Hackers Fado Shanghai hackerspace seems again. Hi. And I'm using a Chinese. Open Source friendly eta, so you can throw your PCB on a web interface is very very well done. And then they have an English version which is on my side is called is Eazy E da you log in, you select either open source or close to a public or private project design your, your components, and they're actually linked to a PCB shop. So when you're done you'll click send to the PCB shop, and then you can order your PCBs and they're delivered at home, but that's just the first step. The second thing is, they also the same shop also does components, they're like a Chinese version of Moser, and they have this huge components database. And because they're the same family, you can actually take the components from their database and use it on your PCB design and the footprints and the semantics all come ready made and tested as if you were flex very huge time saver. And then you'll have your bomb, so you can click on the phone icon and you can order those components. And then as they say even more extraordinary service where you can. They offer SMT prototyping. And let me say that again. They do SMTP can place a robotic for just one or two pieces. Normally you start at 1000. And they do Excel for one or two pieces and the price is just incredibly cheap, it's the same price as normal. When you get for for high volume production, and in about slight between the time I click pay on my on my browser and the time I get the products on my doorstep is like three days. I have party soldiers populated and soldered chips, called PCBs coming into my doorstep. In this one by this one has this greenish color by me but most of the other components are made by the our SMT actually. This cost me is 300 un for two to two boards slash $50.
So do you have to live in China to be able to access that or is that something that someone from outside of China would be able to, to, to take advantage of.
For some of those services like easy Do you can access anywhere, and it's in English, but. And I think the PCB survive, is also offered outside of China, but for the SMT. It's only offered in China, as far as I know, and one of the reasons might be the IHS and CEO regulations, because I have no, the free lead free certificate or anything.
It would be incredibly difficult to.
And so any other tips or tricks for people about how to plug into the supply chain there.
If you speak Chinese.
A lot of doors open because you can like everything is d materialized and you cannot use a remainder in China to to buy things, and have them shipped wherever you live but if you speak Chinese and you already do that, most likely.
All right, we have a couple of audience questions here. So first of all, question I noticed you have the sharp symbol and one of the keys, but not the flat symbol, I assume the flat button is for making a note sharp.
Why not the equivalent for flat.
So I have to say, equal temperament so the flat a C, C, C flat would be a B sharp. That doesn't work. The see if that would be a B but the B flat would be a, a, you have to like. And I understand that musically the sharps and the flats are not the same thing, but it's kind of an advanced topic.
So, it seems a one Pittston chat said, If kids can learn or quarter in school they could probably learn the pocket organ so have you thought about promoting this inside of schools as a tool to help children learn music music theory and just enjoy playing musical instruments.
Well, I'm certain I would crush it personally makes sense. That's my first target is actually like learning through fun. So, school skips school shooter and they have to learn because it's their curriculum. And I'm guessing like there is so much inertia in in the school system that they probably don't want to exchange, unless you have like some kind of like pilot program. I guess being the instrument out there is one of the three that the topic of my last the last part of my talk is three one of those big big questions and I don't have the answer yet. So I all I know is that if I want this to be successful at some point I have to build a huge community around the instrument. So I don't have money because I'm not expecting to earn any money from from the instrument itself, it's just a money sink not for source. So, if, like enough people are interested and involved I'm sure things will happen by themselves. If not, then. No. So I understand it takes time to have the community grow. And if that's the case that's why I'm starting now. That's actually hope is my, my bet on on finding expressive people and being so happy they call it organic growth like getting into the, into the movement, and then two years down the road when we have something that's good enough. And people can actually pick it up and play with it.
So this is a really creative exercise creative project merging technology and you know more of the art side music. What advice do you have for people who are trying to do something like this on their own.
Um, hey you know I found. That's just how I'm wired but I guess that's how most of most of us in that kind of environment are wired in
right to break the break the silos break them go beyond the habits. And then I talk a lot about this to people, and I get like very different answers some reactions. Some people are just horrified that it's, it's not as good a piano as a channel would be as a real channel. Then I just ignore them and then some of the other people are curious about the idea or, I get very enthusiastic responses from people who would like to learn music. And then I try to actually that's one of the part one part of the project is actually to get input from other people to incorporate into the, the instruments, isn't there's no way you can have like a good design if you're designing code.
Oh, that leads to another question from the audience. Have you ever played a duet with anybody with two or three of the pocket organs playing together.
So what I do, do is I try to go to the local hackerspace jam sessions, and
The next door Chang has some sessions have been kind of disturbed with COVID
prime time and so on. But it's on the plan for the next session to actually bring my working instruments and use it as
a back on track. I think that's actually one of the great use cases for it.
Unfortunately the current instrument has too much lag, you don't see it on the video but it has too much lag to be very pleasant to play for using with the looper and it's perfect.
Someone recommends you tick tock a duet of to the instruments.
I sorry a tick tock,
yes, create a tick tock video of a duet with two of the instruments to, kind of, generate interest in.
Actually yes I do have other instruments, I can just like record myself and stay on top like this is definitely something I can do as a homework. Let's see, let's play like this. I do that as a homework I put a link on the, on the GitHub and whoever goes to check it out. can have the video.
Fantastic. I, people would love that. So we've got about two minutes left on this q&a session. Tell us a little bit about how you ended up in China, started in France, you ended up as a natively speaking Chinese in China.
Yeah, so I started learning Chinese like 25 years ago and the first 15 years because my wife is originally from China, that we, we met in France. We married there we started our lives here we built a straw bale house there. And one day she calls me and says, Thomas, my boss said, you go to China. And I'm sorry I already said yes. So we moved to China and it was like it was a yes or no question for me either I wasn't able to return Chinese to a point where I could find my own job and did my police work for myself, or it was a three year
expat to life.
period where I would just stay at home and do whatever and then go home. So I succeeded in learning Chinese to the right level and finding my first job.
the rest is history.
Well, that's fantastic. And on behalf of all the hope 2020 attendees and presenters and volunteers. We want to thank you for sharing your passion with us today and enlighten us a little bit about what the process is like to take that idea from something in your head to a physical device, you know, especially given how complicated that be can be in this. This changed world today. And we look forward to seeing what you come up with next so Best of luck to the audience come back at the top of the hour for our next talk 20 years of scary technology city Tex Gravesend and until then keep on hacking.
Thank you, Thomas.
Hello from Kansas City
has been like a bright light in the pandemic summer, a comment, circulating every two years, which always leaves a trail in my mind. During quarantine. I've been working from home. Yes. And for some reason I started sketching this year. I've also been missing my makerspace hammerspace community workshop. And of course, looking forward to hope, I hope to see you in two years.