The piano is a conventional grand, but with digital interface and camera, the composer is separated from it by air, playing without touching. It’s a Theremin interface for a keyboard instrument.

Piano post-modern? Gestural post-digital?

Whatever it is, in a work composer Benjamin Martinson composed for player piano, computer, and Kinect camera, the piano work holds up as musical content – compositional gesture, not just gimmicky digital hand-waving. Martinson himself looks oddly isolated and awkward, a man making rough mime gestures in unseen water, molasses, and wind. I can’t tell whether this is more about our expectations of human movement, or of the piano, or perhaps a combination, that makes these seem incongruous. But the modal rushes of notes from the piano, brash and bright, are confident. I found myself re-listening to the music with my eyes shut. Somehow, then, this strange camera interface allows Martinson to conduct his work, to more completely connect his brain to the musical output.

The music is somehow effortless, and the effort of the composer then becomes infinitely more satisfying than a player piano alone.

To put it another way – Benjamin Martinson’s music is beautiful, earnestly clear without being overly sentimental. (I enjoy his acoustic works.) The interface here becomes extension rather than impediment, as he continues to work across media.

Player pianos and cameras – old technology keeps colliding with new, in endless recursive loops. Here’s what Benjamin says about his work:

A Microsoft Kinect tracks the motion of my arms and legs, and sends this information to a Python program on my laptop. This program uses this information to compose musical instructions, which are sent to the MIDI-controlled piano.

Torrent attempts to find a new sonic world in the familiar piano, by playing it in a different way. Effects like echos or tremolos are applied to the outgoing notes to create a thick, sometimes almost electronic-sounding texture in a completely analog way.

The piece is composed of three main sections. In the first, the performer controls the music by playing an imaginary piano, and in the second he controls it by conducting. In the third section, the intensity of the music is calculated by the distance between the arms and the legs, and the speed of movement.

To check out my other work, please navigate through my YouTube channel, or visit my website at benjaminmartinson.com

Thanks, Benjamin – and Seth Boyer, for the tip!

  • Korhan Erel

    I don’t know whether I like this or not. I can’t really see the point, I guess. Also, he’s got the keyboard layout wrong. I’d put the lower keys on the left hand. And, I think the performance aspect would have made more sense if he also disconnected himself from the audience instead of looking them in the eye.

    • markLouis

      He DID get the keyboard layout wrong! WTF?!

      Is this guy (and everyone around him who didn’t know or didn’t tell him) an idiot? Is he not a musician?

      Or did he lay it out this way to mimic the position of the piano, where the high keys are to camera right and his left hand is to camera right?

      Either way the video is supposed to be about MUSIC and any musician with keyboard experience watching this, I bet, will be stuck trying to figure out why he’s hitting high notes with his left hand. Either he’s an idiot really big time or he made a calculated visual decision that I wouldn’t have made.

      I stopped the video, I couldn’t watch it after realizing you were right about the layout business. I know it’s shallow to judge a technology demonstration by a incidental aspect like this, but I can’t get passed it. There’s just too great a possibility the guy is an idiot.

      (And Yamaha diskclaviers have been doing remarkable remote demonstrations for years, of course without the Kinect aspect.)

      • Korhan Erel

        I think it’s way too harsh to call him an idiot. He probably did make that decision of matching the piano visually.

        I have seen much better stuff done with the Kinect. This performance looks interesting at first because it actually has a piano in it. Pianos are always impressive and attractive, whether they are played by this guy or a cat.

        Leaving the technological aspect of this video aside, I think the music is utterly uninteresting in the first half, but gets more interesting (or less uninteresting) as he also does by stopping air piano playing and actually makes meaningful gestures as if he is conducting an invisible pianist.

        Finally, anything that involves movement of this sort must be well choreographed and practiced. He’s doing a better job than I would (I’m hopeless), but still, I think he needs to approach this again with a dancer, choreographer or a performance artist.

        • markLouis

          I probably over-reacted, because I just typed up my reaction right away, but I don’t know. A lot of programming these days is what real programmers call “scripting”–just routing commands to appropriate destinations. It doesn’t take a lot of skill. I don’t know the details of this guy or what he did so I apologize if I offended anyone by over-reacting. (And I apologize for typing “passed” when I meant “past”.) But I agree with you, I’ve seen better stuff with the Kinect. And I agree with you, too, that any kind of movement based demo should involve some choreographer, and there are student choreographers who’d love to get involved in this kind of stuff if pros aren’t available. So I feel stupid that I tossed around “idiot” so much,and I’m sorry. But I stand by my reaction, however, just more politely.

      • usesomeimagination

        Dude, it’s a virtual piano. He didn’t get the layout “wrong.” He can scale it anyway he wants, vertical, left to right, right to left, diagonal. That’s kind of the point, to escape the physical limitations

        • Korhan Erel

          No it’s not a virtual piano. It is an actual piano. But of course, he can do whatever he wants with it. Getting the keyboard layout “wrong” was just an initial observation of mine.

  • Korhan Erel

    I don’t know whether I like this or not. I can’t really see the point, I guess. Also, he’s got the keyboard layout wrong. I’d put the lower keys on the left hand. And, I think the performance aspect would have made more sense if he also disconnected himself from the audience instead of looking them in the eye.

    • markLouis

      He DID get the keyboard layout wrong! WTF?!

      Is this guy (and everyone around him who didn’t know or didn’t tell him) an idiot? Is he not a musician?

      Or did he lay it out this way to mimic the position of the piano, where the high keys are to camera right and his left hand is to camera right?

      Either way the video is supposed to be about MUSIC and any musician with keyboard experience watching this, I bet, will be stuck trying to figure out why he’s hitting high notes with his left hand. Either he’s an idiot really big time or he made a calculated visual decision that I wouldn’t have made.

      I stopped the video, I couldn’t watch it after realizing you were right about the layout business. I know it’s shallow to judge a technology demonstration by a incidental aspect like this, but I can’t get passed it. There’s just too great a possibility the guy is an idiot.

      (And Yamaha diskclaviers have been doing remarkable remote demonstrations for years, of course without the Kinect aspect.)

      • Korhan Erel

        I think it’s way too harsh to call him an idiot. He probably did make that decision of matching the piano visually.

        I have seen much better stuff done with the Kinect. This performance looks interesting at first because it actually has a piano in it. Pianos are always impressive and attractive, whether they are played by this guy or a cat.

        Leaving the technological aspect of this video aside, I think the music is utterly uninteresting in the first half, but gets more interesting (or less uninteresting) as he also does by stopping air piano playing and actually makes meaningful gestures as if he is conducting an invisible pianist.

        Finally, anything that involves movement of this sort must be well choreographed and practiced. He’s doing a better job than I would (I’m hopeless), but still, I think he needs to approach this again with a dancer, choreographer or a performance artist.

        • markLouis

          I probably over-reacted, because I just typed up my reaction right away, but I don’t know. A lot of programming these days is what real programmers call “scripting”–just routing commands to appropriate destinations. It doesn’t take a lot of skill. I don’t know the details of this guy or what he did so I apologize if I offended anyone by over-reacting. (And I apologize for typing “passed” when I meant “past”.) But I agree with you, I’ve seen better stuff with the Kinect. And I agree with you, too, that any kind of movement based demo should involve some choreographer, and there are student choreographers who’d love to get involved in this kind of stuff if pros aren’t available. So I feel stupid that I tossed around “idiot” so much,and I’m sorry. But I stand by my reaction, however, just more politely.

      • usesomeimagination

        Dude, it’s a virtual piano. He didn’t get the layout “wrong.” He can scale it anyway he wants, vertical, left to right, right to left, diagonal. That’s kind of the point, to escape the physical limitations

        • Korhan Erel

          No it’s not a virtual piano. It is an actual piano. But of course, he can do whatever he wants with it. Getting the keyboard layout “wrong” was just an initial observation of mine.

  • frup

    This seems like a good leaping off point for exploring greater levels of complexity. What happens for example when syncing Kinect up with 2 piano’s and 2 players? Add to that a generative component via Noatikl’s “listening voices” or a 3rd person processing the playing live and things could start to become uniquely interesting.

  • frup

    This seems like a good leaping off point for exploring greater levels of complexity. What happens for example when syncing Kinect up with 2 piano’s and 2 players? Add to that a generative component via Noatikl’s “listening voices” or a 3rd person processing the playing live and things could start to become uniquely interesting.

  • Lebron

    I’m trying not to dismiss this outright but I really dont get the point either. It seems like he did this just because it was possible without thinking about why you’d be doing something like this. The music being played could be played with more expression and more relevant theatrics if he sat down at the piano next to him and played it.

  • Lebron

    I’m trying not to dismiss this outright but I really dont get the point either. It seems like he did this just because it was possible without thinking about why you’d be doing something like this. The music being played could be played with more expression and more relevant theatrics if he sat down at the piano next to him and played it.

  • edisonSF

    this is an interesting conversation starter…
    taking a step away from “art is art is everything and anything” philosophy…
    (which i can gladly subscribe to..)
    i have to ask….
    how disconnected can you get?
    buttons and knobs are a step away from strings and frets…
    computers are a step back from buttons and knobs…
    touch is a further removal from there…
    now, air?
    aside from a feat of technology, what validity does this hold?
    sweeping piano arpeggios…. skittering midi scale maps… fine…
    but there has yet to be an implementation of this (that i’ve seen to date) that isn’t awkward, pained and full of weird holes and missed gestures…
    if the routine was lock step and everything was exact…. would the effect be any better??
    for me…. definitely not….
    i.e. http://www.youtube.com/watch?feature=player_detailpage&v=6btFObRRD9k#t=968s
    why stack and stack and stack on the distance?
    i tend to be dismissive of these performances and just subscribe to the “just play it and fuck the programming” school….
    of course no offense intended, my opinion is my own….
    if you’re going to stand next to instruments and wave at them, for me (the audience)… you’re showing how much of a stranger you are….

    its just bad magic…

    anyway…. sorry to rant…

  • edisonSF

    this is an interesting conversation starter…
    taking a step away from “art is art is everything and anything” philosophy…
    (which i can gladly subscribe to..)
    i have to ask….
    how disconnected can you get?
    buttons and knobs are a step away from strings and frets…
    computers are a step back from buttons and knobs…
    touch is a further removal from there…
    now, air?
    aside from a feat of technology, what validity does this hold?
    sweeping piano arpeggios…. skittering midi scale maps… fine…
    but there has yet to be an implementation of this (that i’ve seen to date) that isn’t awkward, pained and full of weird holes and missed gestures…
    if the routine was lock step and everything was exact…. would the effect be any better??
    for me…. definitely not….
    i.e. http://www.youtube.com/watch?feature=player_detailpage&v=6btFObRRD9k#t=968s
    why stack and stack and stack on the distance?
    i tend to be dismissive of these performances and just subscribe to the “just play it and fuck the programming” school….
    of course no offense intended, my opinion is my own….
    if you’re going to stand next to instruments and wave at them, for me (the audience)… you’re showing how much of a stranger you are….

    its just bad magic…

    anyway…. sorry to rant…

  • XY01

    I think a lot of people in the comments are missing what seems to be the point of these exercises in gestural performance. It IS about exploring the technology as much as it is about exploring the art.
    It isn’t necessarily about creating something amazing off the bat, it is about exploring ideas and new forms of interaction that could lead the the development of new instruments.
    I work on a similar project myself Ethno Tekh and we make loads of different instrument implementations using synths and gestural inputs using kinect. Some of them are good some are bad but they are all experimental.
    Think about the first time someone blew through a hollowed out branch. I am sure it didn’t sound as amazing as a contemporary didgeridoo but it was surely none the less interesting and exciting to see the possibilities in where it may go.
    Thoughts?

    • edisonSF

      in small part, i agree…
      experimentation is always good…

      as for didgeridoos…
      maybe the first attempt was crude…
      but what about the first time someone stood on stage, pretended to blow in a log and someone off-stage made a noise?

      it’s just too detached….
      i dunno…
      your guys sound is decent…
      nice visuals/production… .

      alas…..
      i just don’t get the kinect music connection……

      • XY01

        If the first guy playing a didgeridoo had people willingly gather around to listen while he experimented, then that’s great. I see this in a similar light.
        There is such wide and varied art and musical forms now that people are quite able to freely choose something that piques their interest.
        Obviously this is not to everyone’s liking or understanding, and perhaps my own understanding of what his intention is is wrong. Though I know myself that even though I don’t particularly enjoy the music I do enjoy the technological aspect, the way in which he has explore gesture to interact and the way in which he has built, practiced and improvised a new instrument.

        For us (Ethno Tekh) the initial motivation to build a motion driven instrument was a response to the all too ‘press play, pretend to twist knobs’ EDM scene that we grew up in. To have something that has an obvious live input, something that can be screwed up or made to sound amazing live on stage makes for a unique and connected experience. This experiment however ties the gesture to a preexisting physical instrument which is an interesting approach, and may well lead to the instrument being able to be played in new and interesting ways which you’re unable to do via traditional means, while no doubt also losing a lot of the articulation that you are able to get through traditional play styles.

  • XY01

    I think a lot of people in the comments are missing what seems to be the point of these exercises in gestural performance. It IS about exploring the technology as much as it is about exploring the art.
    It isn’t necessarily about creating something amazing off the bat, it is about exploring ideas and new forms of interaction that could lead the the development of new instruments.
    I work on a similar project myself Ethno Tekh and we make loads of different instrument implementations using synths and gestural inputs using kinect. Some of them are good some are bad but they are all experimental.
    Think about the first time someone blew through a hollowed out branch. I am sure it didn’t sound as amazing as a contemporary didgeridoo but it was surely none the less interesting and exciting to see the possibilities in where it may go.
    Thoughts?

    • edisonSF

      in small part, i agree…
      experimentation is always good…

      as for didgeridoos…
      maybe the first attempt was crude…
      but what about the first time someone stood on stage, pretended to blow in a log and someone off-stage made a noise?

      it’s just too detached….
      i dunno…
      your guys sound is decent…
      nice visuals/production… .

      alas…..
      i just don’t get the kinect music connection……

      • XY01

        If the first guy playing a didgeridoo had people willingly gather around to listen while he experimented, then that’s great. I see this in a similar light.
        There is such wide and varied art and musical forms now that people are quite able to freely choose something that piques their interest.
        Obviously this is not to everyone’s liking or understanding, and perhaps my own understanding of what his intention is is wrong. Though I know myself that even though I don’t particularly enjoy the music I do enjoy the technological aspect, the way in which he has explore gesture to interact and the way in which he has built, practiced and improvised a new instrument.

        For us (Ethno Tekh) the initial motivation to build a motion driven instrument was a response to the all too ‘press play, pretend to twist knobs’ EDM scene that we grew up in. To have something that has an obvious live input, something that can be screwed up or made to sound amazing live on stage makes for a unique and connected experience. This experiment however ties the gesture to a preexisting physical instrument which is an interesting approach, and may well lead to the instrument being able to be played in new and interesting ways which you’re unable to do via traditional means, while no doubt also losing a lot of the articulation that you are able to get through traditional play styles.

  • Robbe K

    Ok! I love the technics behind this, but: I still think the best controller for a piano, would still be the klavier itself. I see a lot of controllerism and believe that playing an actual instrument is still one of the best ways of controlling sound…

  • Robbe K

    Ok! I love the technics behind this, but: I still think the best controller for a piano, would still be the klavier itself. I see a lot of controllerism and believe that playing an actual instrument is still one of the best ways of controlling sound…

  • djtmc

    Some of the comments remind me of the resistance to technology throughout history. We don’t often deal with new ideas very well. This work is a new composition performed by a very accomplished musician with new technology. He obviously knows what he’s doing with both the technology and music. It’s fascinating to watch and to listen. Congratulations to Martinson for having the vision to put this together.

    • Yeah; I’m mainly embarrassed that, as I was busy with Resonate Festival in Belgrade, I missed moderating some of these comments. Seriously – some of them are constructive criticisms, which I think is fine, but a couple (just deleted one) were just outright abusive. It’s clear again and again that people say things on the Internet, which are read by the intended target, as if that person *can’t see them*. People say what they think rather than what they’d say to someone’s face. That creates a hostile environment to any actual constructive criticism. And yes, I will happily, happily throw out any of those comments as worthless. Anyone unable to use basic social judgment doesn’t deserve any time spent on their opinions.

    • People also missed the point that this is a composer who regularly composes *for the piano*. That makes a lot of this discussion a complete waste of time. He experimented with an alternative interface.

  • djtmc

    Some of the comments remind me of the resistance to technology throughout history. We don’t often deal with new ideas very well. This work is a new composition performed by a very accomplished musician with new technology. He obviously knows what he’s doing with both the technology and music. It’s fascinating to watch and to listen. Congratulations to Martinson for having the vision to put this together.

    • Yeah; I’m mainly embarrassed that, as I was busy with Resonate Festival in Belgrade, I missed moderating some of these comments. Seriously – some of them are constructive criticisms, which I think is fine, but a couple (just deleted one) were just outright abusive. It’s clear again and again that people say things on the Internet, which are read by the intended target, as if that person *can’t see them*. People say what they think rather than what they’d say to someone’s face. That creates a hostile environment to any actual constructive criticism. And yes, I will happily, happily throw out any of those comments as worthless. Anyone unable to use basic social judgment doesn’t deserve any time spent on their opinions.

    • People also missed the point that this is a composer who regularly composes *for the piano*. That makes a lot of this discussion a complete waste of time. He experimented with an alternative interface.

  • I actually quite liked the resulting music in that video. I thought it was elegantly coded (well, the results are elegant, haven’t seen the actual code)

    I guess my main issue with the *tech* behind it, as I start to parse the video in my head, is that I don’t see a lot of… control. I mean, it doesn’t look like there are gestural benefits, really. Only lack of fine control.

    Of course we need work like this to push the scene into the level where one might actually develop to become virtuoic on the instrument. But for now it really feels like going through the motions. Thankfully this is where the performer-slash-coder can compensate through finetuned code and pre-planned sections with different code.

    • Benjamin Martinson

      Thanks, everybody, for sharing your thoughts. Andreas, your criticism is fair, and I completely agree. Reliable control over the instrument was a big problem, and I’m a bit surprised (but glad) that my frustration isn’t more evident during the performance.

      The first obstacle was the stage lighting–this introduced a ton of jitter into the coordinates I was getting from the Kinect, which made me have to use large gestures to get much of a reliable response. I could have done this on a dark stage, but what’s the point if the audience can’t see me?

      The second obstacle was the mechanics of the piano itself–I’d say it didn’t respond to about 60% of the notes that were sent to it. There appears to be a complex relationship between note velocity, number of notes down, amount of activity, and the likelihood that a note will actually sound. The conservatory only gave me an hour and a half with the actual piano, so I wasn’t able to account for this in the code, but I hope to model this relationship better in future versions.

      There is certainly still much to do to improve this piece, and I appreciate the honest feedback. If anybody on here has experience programming for Kinect or PianoDisc systems, I’d love to get your advice!

      • Korhan Erel

        So this was simply an experiment, a work-in-progress… Now it makes sense 🙂

  • wetterberg

    I actually quite liked the resulting music in that video. I thought it was elegantly coded (well, the results are elegant, haven’t seen the actual code)

    I guess my main issue with the *tech* behind it, as I start to parse the video in my head, is that I don’t see a lot of… control. I mean, it doesn’t look like there are gestural benefits, really. Only lack of fine control.

    Of course we need work like this to push the scene into the level where one might actually develop to become virtuoic on the instrument. But for now it really feels like going through the motions. Thankfully this is where the performer-slash-coder can compensate through finetuned code and pre-planned sections with different code.

    • Benjamin Martinson

      Thanks, everybody, for sharing your thoughts. Andreas, your criticism is fair, and I completely agree. Reliable control over the instrument was a big problem, and I’m a bit surprised (but glad) that my frustration isn’t more evident during the performance.

      The first obstacle was the stage lighting–this introduced a ton of jitter into the coordinates I was getting from the Kinect, which made me have to use large gestures to get much of a reliable response. I could have done this on a dark stage, but what’s the point if the audience can’t see me?

      The second obstacle was the mechanics of the piano itself–I’d say it didn’t respond to about 60% of the notes that were sent to it. There appears to be a complex relationship between note velocity, number of notes down, amount of activity, and the likelihood that a note will actually sound. The conservatory only gave me an hour and a half with the actual piano, so I wasn’t able to account for this in the code, but I hope to model this relationship better in future versions.

      There is certainly still much to do to improve this piece, and I appreciate the honest feedback. If anybody on here has experience programming for Kinect or PianoDisc systems, I’d love to get your advice!

      • Korhan Erel

        So this was simply an experiment, a work-in-progress… Now it makes sense 🙂

  • David

    Utter pointless crap. Can’t beleive shit like this get any exposure.
    Someone said idiot, quite accurate description of this moron.

  • Jeffrey

    After the first three minutes or so. The music got more ‘composed’ as opposed to the stupid crap he was doing at the beginning. I understand he was showing how he can control the piano. But then the piano suddenly started playing itself and he began waving his hands like a conductor. Which means he was acting like a conductor, which really means absolutely nothing. And then all he really had was a digital version of a player piano, which has already been around for more than a decade. Not really impressed.

    • You know, generally speaking I’m inclined to delete comments that describe someone’s music as “stupid crap.” I don’t think that’s ever an okay thing to say. This is generally something you say to someone, to their face, when talking about their music?

  • Jeffrey

    After the first three minutes or so. The music got more ‘composed’ as opposed to the stupid crap he was doing at the beginning. I understand he was showing how he can control the piano. But then the piano suddenly started playing itself and he began waving his hands like a conductor. Which means he was acting like a conductor, which really means absolutely nothing. And then all he really had was a digital version of a player piano, which has already been around for more than a decade. Not really impressed.

    • You know, generally speaking I’m inclined to delete comments that describe someone’s music as “stupid crap.” I don’t think that’s ever an okay thing to say. This is generally something you say to someone, to their face, when talking about their music?

  • Moco

    your comments remind me alot about what Ive be saying the Past on my blog

  • Moco

    your comments remind me alot about what Ive be saying the Past on my blog

  • PaulDavisTheFirst

    the conductor analog resurfaces again. i have nothing against conductors – they do amazing things. but here we at least see the performer actively connect what they are doing with conducting – they abandon fine scale over notes, timing, possibly even entire harmonic relationships and melody, but take control over broader aspects of the performance. just as a conductor does, leaving the actual “performing” to the orchestra members (here, mapped to some prexisting compositional notions and mechanical playback devices, that do their own thing, subject to his final, high level control).

    we recognize real genius in this type of thing, and celebrate it. which is good, very, very good.

    lets just not go confusing simon rattle for rostropovich. conducting is a skill. playing an instrument is a skill.different, therefore equal. and to the extent that this work plays around with the boundaries between them, in the end it seems to me to confirm that distinction quite clearly.

  • PaulDavisTheFirst

    the conductor analog resurfaces again. i have nothing against conductors – they do amazing things. but here we at least see the performer actively connect what they are doing with conducting – they abandon fine scale over notes, timing, possibly even entire harmonic relationships and melody, but take control over broader aspects of the performance. just as a conductor does, leaving the actual “performing” to the orchestra members (here, mapped to some prexisting compositional notions and mechanical playback devices, that do their own thing, subject to his final, high level control).

    we recognize real genius in this type of thing, and celebrate it. which is good, very, very good.

    lets just not go confusing simon rattle for rostropovich. conducting is a skill. playing an instrument is a skill.different, therefore equal. and to the extent that this work plays around with the boundaries between them, in the end it seems to me to confirm that distinction quite clearly.

  • Matt Harder

    It should be someone trained in movement – ie. a dancer – who performs the piece. Nice idea though. Lots of great possibilities.

  • Matt Harder

    It should be someone trained in movement – ie. a dancer – who performs the piece. Nice idea though. Lots of great possibilities.

  • fabz

    There isn’t much to say about comments of those who merely “see” through their eyes and “hear” with their ears. Judging without the actual deep and true knowledge of what we are experiencing is a clear manifestation of ignorance!

    Only if you really UNDERSTAND the work behind a project like this and the creativity of the mind who presents it you can really appreciate such performance.

    I wish I was able to experienced this gig live, and possibly congratulate in person with Benjamin Martinson!!

    Inspiring!!

  • fabz

    There isn’t much to say about comments of those who merely “see” through their eyes and “hear” with their ears. Judging without the actual deep and true knowledge of what we are experiencing is a clear manifestation of ignorance!

    Only if you really UNDERSTAND the work behind a project like this and the creativity of the mind who presents it you can really appreciate such performance.

    I wish I was able to experienced this gig live, and possibly congratulate in person with Benjamin Martinson!!

    Inspiring!!

  • Josh

    May be too late for the party, but this is awesome. Not sure I’d call it a composition, depending on how long he worked on the hand motions and gestures, but at the very least an amazingly cool improvisation, and in my opinion, it’s not perfect at it yet, but it’s one step closer to enabling a human to compose and control music directly with emotion. Very cool idea, and the music itself sounds very good. I thought the piece started a tad slow, but especially towards the middle and end it sounded great. On a more technical note, pretty impressive that you were able to get that much expression out of the piano without directly controlling it.

  • Josh

    May be too late for the party, but this is awesome. Not sure I’d call it a composition, depending on how long he worked on the hand motions and gestures, but at the very least an amazingly cool improvisation, and in my opinion, it’s not perfect at it yet, but it’s one step closer to enabling a human to compose and control music directly with emotion. Very cool idea, and the music itself sounds very good. I thought the piece started a tad slow, but especially towards the middle and end it sounded great. On a more technical note, pretty impressive that you were able to get that much expression out of the piano without directly controlling it.