Kristin Miltner on nurturing software and programming for video games -- plus visual art as preparation for sound-work.

The album Grains by Oakland, California-based musician Kristin Miltner is full of fidgety algorithmic chaos and patches of soft noise. The five succinct tracks that comprise the set manage to be rambunctious and sedate at the same time — for all that activity, when taken in stride, becomes a highly textured flow.

The ease inherent in Grains (released on the San Francisco label Praemedia) is due in no small part to Miltner’s voice, which informs several of the tracks. She sings tones that, for all the digital processing, maintain a loveliness that never gets too far from being recognizable as human.

Miltner studied music at Mills, but that was only the latest educational experience in lifelong studies that began with violin training and proceeded through two BFAs. Prior to recording Grains, Miltner teamed with musician Mark Bartscher; together, as Miba, they recorded an album, The Corplate Porblem, which, much like Grains, emphasized granular synthesis that located within samples tiny fractures of sound.

Miltner recently took time to discuss the differences between recording solo and as part of a duo, the ongoing effort involved in developing audio software patches, how her visual arts background informs her composing, and her day job designing sound for video games.

Marc Weidenbaum: The use of voice on the first song on Grains, “Grains Need Water and Sunlight,” is especially distinctive. In the course of recording that piece, did you alter the way you sang to fit the digital processing that you were applying to your voice?

Kristin Miltner: For all of Grains, I was using the software I built in Max/MSP. I know that I have set the buffers in my patch to record for 10 seconds, so if I’m singing into the patch, I try to time it so I sing a few notes or say a few words or whatever, and the whole improv lasts about 10 seconds.

Weidenbaum: Are the sounds on Grains ones you heard in your head and pursued, or ones that arose through experimentation and discovery?

Miltner: Both happen, and to answer that question specifically I feel like I should tell you about how I work. I use a specific piece of software that I wrote and am writing in Max/MSP that very much dictates, for better or for worse, the way I sound. I have been working on the same patch for at least five years and keep adding to it. I developed it to be very good at sounding a certain way — the stuttery, rhythmic theme that you hear on Grains. I like to think of it as something that cuts lacy patterns into samples and live input. There’s another abstract description of how it works by Jorge Boehringer — he describes it an an octopus opening and closing multiple doors in a very long hallway.

Anyway, inevitably, as a result of building it that way, as a result of choosing and eliminating, that’s the way I sound. We mutate each other as we grow symbiotically. I hear sounds out in the field in terms of what my patch will do to them. If another musician is playing an instrument or playing me a recording they made, or I am out in the world listening to a sound, I think about what it would sound like if I brought it into the patch.

Weidenbaum: Please describe the patch more in detail.

Miltner: The heart of the software patch is an instrument that allows me five buffers that I can scan live any way I choose; I set the BPM (if I want to have one; sometimes I decide to let the tempo change gradually, in a “random walk”) the playback speed, how many pulses (think of the pulses as long-ish grains of the sample in the buffer) it scans, the length of each pulse, if there are any rests where no pulse is heard, what range of the sample in the buffer it’s currently scanning, etc.

As far as the vocals go, one layer will be a result of vocal improvisation that is “caught” in the scanner and repeated by the computer in a way that I like. The vocal improvisation is usually in response to some synth sound I have made in Logic or played on piano or Wurlitzer or Rhodes, etc., and have brought into one of the buffers. I will then “hear” harmony parts that go with the first vocal, and sing and record those into other buffers. I then chose a “landscape” that goes with the voice — samples of noise, or birdsong, or water, or someone crashing two pieces of metal together — and scan those samples along with the vocals. I set their rhythm patterns, their beat divisions, and then I have all the pieces of the composition constructed.

After all my layers are built, I start weaving the layers together, taking some out, leaving others in, stacking the layers in different ways. I adjust the rhythm by means of a beat division table, which shows me whether a particular count is in, or is resting, and how long the pulse is. I also select different portions of the waveforms in the buffers, which results in a chord change and/or a textural change. I do this until I get something that sounds like a halfway coherent composition. I will play a few different “beginnings” and “endings.” All this goes into one or two long recordings that contain a few different takes and experiments.

Then I edit. For Grains I used Peak instead of ProTools to prevent myself from doing too much editing. I stayed in a two-channel environment so all I could do was cut from take to take, maybe take the beginning of one take and put it on the middle of another take; that is as serious as the editing got this time. That’s basically all Mark Bartscher and I did with the Miba CD The Corplate Porblem (2004, pax). I tried to follow the model we used then. I want all the layering and on-the-fly composing to be done in the Max patch; I think that’s what keeps it sounding spontaneous.

Weidenbaum: Your description of the patch sounds very personal. Did you have that sense of attachment to the patch when you started working in Max/MSP, or did it develop over time?

Miltner: Over time I guess I realized that I was growing a creature with personality quirks, with tendencies to be stubborn about certain things, and to make other things happen without even being asked. Little favors here and there, unexpected surprises. It became so much like working with a person, albeit a warped and twisted person. I think it’s my visual art background that makes me see it that way. It’s so much like sketching something, and then backing away from it to see if it looks right, and then re-drawing a firmer line, assessing the line, re-working, assessing, listening, making a new line, pausing, making different lines, applying various amounts of pressure, resulting in a thinner or thicker lines. Assessing the results. I had the same problem with drawings and paintings as I do with the patch — I could never be done with them either.

Weidenbaum: Please choose one song from Grains and describe how it came to be.

Miltner: “Body in Sleep” came about this way: I hunted through all my recordings for a piano sample, something low-key. I knew I wanted the last track to feel like you were drifting off slowly. I was trying to capture that time between asleep and awake. I found an autoharp sample I really liked. It was just two notes, sustained for around six seconds each. So I put that in a couple buffers, and then whispered into another one, and one had some bird sounds. I put some filters on some of the tracks, and then played them all together, overwriting the buffers with more voices, loading in other pre-recorded voice tracks that contain things being whispered.

Weidenbaum: I’ll try not to focus too much on your use of voice, but I was wondering if Laurie Anderson’s work played any role in your sense of how to invoke vocals in electronic music.

Miltner: I think in her work, each of Laurie’s songs is much more of a concept. I think of her work as more “text-based,” as in she actually has lyrics in mind, and messages to her audience in those lyrics, whereas Grains is more about texture, and the words that come to mind when I am singing are more like pieces of words, subconscious babble. Dream words. I wanted it to be an intuitive album, that invites you in without overpowering you with any specific lyrical ideas — something that would encourage the listener to invent imagery or ideas along with it.

Weidenbaum: Some musicians create a patch for each work. Others have a handful they call upon, like a small box of staple recipes. You seem to be saying you have one single patch that is your virtual equivalent of a studio. Am I getting that right?

Miltner: Not really a studio patch, but a live performance patch. The fantasy is one big performance patch that can handle anything coming at it, whether it be live input either from myself or other musicians, or any one of hundreds of combinations of my samples in a way that fits my aesthetic or fits the improv situation I am in — balance between control and spontaneity that feels right to me. Then there is no need for editing ever. An impossible dream, but something I aspire to.

Weidenbaum: Did working solo on Grains feel significantly different from working as a duo, as you have in the group Miba — and could you say, in a manner of speaking, that your Max patch is your partner?

Miltner: Creating Grains felt a lot different for me than the Miba CD, for the same reason I enjoy playing live as a solo musician: every single decision made — or not made — is one I am responsible for, so if it goes bad, I have only myself to blame. In a live situation this total responsibility makes me feel like I have more freedom, because I am not deferring to anyone else. I don’t have to feel bad about stepping on anyone’s toes or worry about feeling bad about stepping on people’s toes, or get mad at myself because I am deferring too much or not listening to what’s going on or letting my ego take over.

Mark Bartscher and I did not make a single decision separately when we edited the Corplate Porblem. We decided on everything together. It was awesome to have that experience first, because when I worked on Grains, Mark was still there; I would hear his input and observations, or what I think he would think about the tracks.

Weidenbaum: Please say a little more about the editing process. You say you want to keep the editing simple, so are you mostly looking for a good introduction and a good exit — does song form play much a role in your decision-making about how to shape a track?

Miltner: It most definitely does, but I try to create a coherent form when I am playing live — or making the takes for a CD — so I don’t have to create a form in post. I want to keep it as much like a live session as I can because I wish for this type of music to be able to be generated on the fly, and form to be just as interesting as it would be had things been added and subtracted from it. After I play the “song” in Max two or three times, and have recorded myself playing it, I bring them into Peak, but leave the takes mostly whole — like I said, I may slap a different beginning on one, like, if I remember right, the first half of “Bell Cycle” [track 4 on Grains] is take two, and the last half is take three.

Weidenbaum: I know very little about you personally — could you share some basic biographical information: where you’re from, what you studied prior to Mills, general cat/dog, coffee/tea, Pepsi/Coke type stuff?

Miltner: I’m from Omaha, originally. I started piano lessons when I was 4. I was a Suzuki kid. I went to the Kansas City Art Institute because I thought I wanted to be a painter, and there I was exposed to a lot of great video, sculpture and performance art. I ended up getting a BFA in Photography and New Media, and a second BFA in Art History. I made some Super 8 films and videos while I was there, which led me to ProTools, which led me to make multichannel audio installations. I got really into Pauline Oliveros’s work. After KCAI, I came to Mills. Now I am making sound effects and composing music for video games. I live with Cliff Caruthers, who is also a professional composer and sound designer, and helps to curate the San Francisco Tape Music Festival. We live in a house in Oakland, and we have three cats and two Ameraucana hens. I like Diet Coke, and both coffee and tea.

Weidenbaum: I’ve spoken with several musicians who work, by day, making music for games. For some it’s really simply a day job; for others the processes, processing power and tools-sets in gaming have influenced what they do in their personal music, and vice-versa. Could you describe what you’re doing at work in gaming, and how it might relate to your compositions and performance work?

Miltner: Sure. I think making game music has not only taught me a lot about the available tools and technology, but has also exposed me to the myriad ways that people use these tools. You can observe techniques, workarounds, and exploitations of these tools when you work with a team of other audio technicians, and arrive at solutions you would never think of if you were entirely on your own. Having to design small-footprint games really taught me a lot of ways to maximize a very small palette and to think about what can, if long and varied pieces cost too much in terms of space or memory, be done programmatically to enhance the design of the music and sound effects. I have built up techniques for percussion that changed the way I think about programming my own drums and textures when I make my personal work. I’ve made tons of custom instruments for particular effects I want to achieve in certain background musics. I am very “Logic matrix window”-oriented. I tend to draw and cut and paste all my notes, only actually “playing in” a measure or two at a time. People think this is time-consuming, but actually I’ve gotten quite fast at it. I love the color-coded velocities. Having your pitches on a colored grid-like display makes an incredible amount of sense to me.

Related links: Kristin Miltner's MySpace page ( Praemedia Records, which released Grains ( Pax Recordings, which released Miba's album (

Leave a Reply

Your email address will not be published. Required fields are marked *