Dan Tepfer's Acoustic Informatics

watch the video

We caught up with composer and Culture Crush contributor Dan Tepfer at (le) poisson rouge as he was preparing his new work, Acoustic Informatics, a perfect blend of the acoustic and the electronic, where the mathematical and emotional sides of the artist play equal roles. Dan takes us through the technical explanation of his methodology and how he arrived at the balance of this unique musical experiment.

Debra Scherer: So tell us about your unique background and how that has given you an unusal point of view.

Dan Tepfer: I was born in Paris, but my folks are American, so I kind of grew up with a double culture.  I went to a French school in France, but also spoke English at home and spent summers in the United States.  I think that kind of informs who I am as a musician. It’s not so much like Europe vs. America, but I think the idea of multiculturalism is really important to me as a musician.  So I grew up playing a lot of jazz because my grandfather was a jazz pianist on the west coast, and I just got super deep into that music.  But I was also studying classical at the conservatory the whole time, so I was very self-taught in jazz until my early 20’s.  I think that’s what led to my not getting too attached to one specific style and actually being much more attached to what I feel is good music, you know?  

One of the things I’ve really gotten into recently is mixing -- going a little bit further afield, and not only crossing from jazz into classical, or into electronic music, which is something I’ve done quite a bit of, but I’ve been starting to use computers, because this is the age of computers.  I mean, we’re in the thick of it now. Artificial intelligence is already a world-changing technology and I think it’s going to be much bigger in the next 20 years, and that can act on music in a very direct way.  So I’ve started exploring that side of things, too.  And I think that connects into who I am as a musician because, at the end of the day, I’m just interested in anything that kind of surprises me, and that leads to music that I wouldn’t have expected, that I wouldn’t have gotten to in any other way.

DS: But also, you at one point almost turned away from music, correct?

DT: Right.

DS: So what pulled you back in?  That’s gotta be something about your passion for music!

DT: Yeah!  That’s a great point.  So I really grew up playing music. I started studying piano when I was six, and it’s one of the things I spent the most time on, but I was also always super passionate about science.  And I actually studied astrophysics in my undergrad, and there was definitely a time when I was considering going into research.  And the thing is, when I was starting, I worked in an astrophysics lab for a summer, and when I was starting to do that work, I realized that music just adds another layer to the day-to-day existence. I mean, certainly if you’re composing, if you’re doing things like what I’m looking into now with algorithms, you’re actually doing research, but there’s also this whole huge kind of visceral, earthy, intuitive component, an emotional component to music that I think is pretty amazing.  And I just love having the opportunity to try to bring those two worlds together.  I would say that’s what ultimately really keeps me in music. It feels like it has the potential to activate both sides of the mind.

DS: Right.  So you became interested in artificial intelligence, what have you done about it?  What’s this new project?  Tell us about this new project that you’re doing.

DT: This is a project that I’m calling Acoustic Informatics because I’m using informatics, like information technology, computers, to generate a lot of the music that you’re gonna hear.  But everything, all the sounds you’re hearing, are acoustic.  In fact, I’m also going to be joined by four members of ICE, the International Contemporary Ensemble, one of the greatest new music groups in the world.

DS: How did you choose them?

DT: Well, it’s a long story, I mean, the group itself is very well known on the scene.  But we started working together through Ben Wendel.

DS: Love Ben Wendel!

DT: Yeah, he’s a great saxophonist and composer, and we’ve done a lot of playing together.  We did one gig at the Jazz Gallery a few years back where we were joined by a couple of the members of ICE and I’m excited to have them perform with me here. A lot of the music that they’re going to be playing is also going to have to do with computers, but all the sounds you’re going to hear are acoustic.  That’s something that I’m really interested in, because to me, acoustic sound is very powerful.  It just speaks to the heart in a very direct way.  But I’m also really interested in algorithms and how those connect with music.  So what if you could hear an algorithm, but played by acoustic instruments? I find that really compelling.  

One aspect of Acoustic Informatics is that this instrument, the Yamaha Disklavier, is a player piano, so it can play on its own.  I’ve written algorithms that respond to what I’m improvising in real time.  So in other words, if I play a note, that note is going to go into my computer.  And my computer’s running programs that I’ve written that will then input the output data to the piano, which is gonna respond to what I’ve played.  And the thing is, since I’m improvising, I’m also going to respond to it, because I see what’s going on. So it’s very much a symbiotic relationship.  It feels very much like resonating with the piano, and with the computer, because there are two waves like that.  It’s responding to me; I’m responding to it.

debra scherer

debra scherer

DS: Is it one algorithm?  Or are there many?

DT: No, I’ve written maybe twelve so far?

DS: And are they all working at the same time?

DT: No.

DS: Or are you calling up one specific algorithm on your computer?

DT: So, the way I’m thinking about it right now, and this is something I’m just starting, right?  The way I’m thinking about it right now is that an algorithm is a piece.  So, an algorithm is a way of functioning, it’s a set of rules.  And it doesn’t in any way dictate what the music is going to be.  It doesn’t dictate what the data are going to be.  It dictates how things are going to be processed.  So it’s a way of doing things as opposed to a thing.  And that’s what I love about it -- we could call it process music.

DS: Okay...

DT: You know?  So in other words, I’ve written these algorithms and each one of them, to me, is a piece.  Each one of them guides me towards a certain kind of sound.

DS: But what I’m trying to understand is, how do you know which one of the twelve is being used?  Or do you use one at a time?

DT: Oh!  I just turn them on and off.  I’ll use one at a time.  I could use multiple ones at a time, but to me, at this point, I’m trying to keep it really simple because the more direct and simple it is, the more I can really engage with it as an improviser.  If it’s doing things that I really don’t understand, then that makes the relationship a little bit more complicated.  And I think I’ll get there!  But for now, I’m really digging using very simple things.

DS: So it’s the same as when you play improv jazz with somebody for the first time.  You guys are communicating in a certain way, with looks and sounds, and little tics, and I would assume that the more you play with somebody, the more you sort of know them, just like in any kind of relationship.  So it’s like you’re just building that same kind of relationship, but with the specific algorithm that you’re playing with at that moment.  Is that true?

DT: I think that’s a really good way of putting it!  There’s maybe one difference though. The reason that I’ve gotten into this project in the first place is that music has always been a mix of the algorithmic and the spiritual.  If you look at Bach-- somebody that I really, really love, and I play a lot of Bach’s music -- many of his pieces are guided by a set of rules...well, all of his pieces are guided by the rules of counterpoint, which has this long tradition of how to make one melody work with another melody.  

That’s a tradition that goes back centuries before Bach.  So those rules, already, are functioning underneath all of his music.  And those are totally algorithmic!  I mean, those are rational rules that he is using to support the spiritual or emotional message that he’s putting forward, right?  And then on top of that, he often has actual process rules in his pieces.  So an example would be a canon, where one voice starts playing and then that same voice will be repeated a little bit later, like in the third Goldberg Variation; the first voice gets repeated verbatim while the very first voice starts doing something else.  I mean, that’s an example of an algorithm right there.  And it’s followed really strictly through the piece.  So that’s true for Bach, but it’s also true of a lot of my favorite composers like György Ligeti, one of the greatest composers of the second half of the 20th century.  He used a lot of rule-based processes.  So, what I’m getting at here is that the computer is strictly algorithmic, there’s nothing emotional about it.  And there are no real choices.

debra scherer

debra scherer

DS: Are you sure?

DT: I’m pretty sure!  Because it’s really applying my rules to the letter.  It has no choices, it has no liberties, you know?

DS: It didn’t get disappointed by it’s tangerine breakfast?

DT: No.

DS: It’s not bringing any emotional baggage into the performance?

DT: No, it’s cold and calculating. Ha!

DS: So far!

DT: So far!  We’ll see what happens. Artificial intelligence can get pretty deep.  This is very simple artificial intelligence for now.  I hesitate to even call it artificial intelligence, I mean, it’s just some rules that are being used.  But the point is that a lot of the greatest music has been made with some really hard and fast rules -- cold, calculating rules, plus ineffable spirituality.  And in what I’m doing, I have the computer really taking care of these very hard and fast rules, and on top of that, I’m improvising totally whimsically, doing whatever I want.  In that interaction, I think there’s something really interesting.

DS: Right, that goes back to the two sides that you were talking about to begin with -- between science and music.

DT: Exactly.

DS: So you’re the music, and the computer is the science?

DT: Yeah!  That’s a really good way of putting it!  It means that it frees me. I don’t have to be the science! I can just be the music!