Behind the Scenes With the Stanford Laptop Orchestra
Ten days before the big concert, the members of the Stanford Laptop Orchestra are performing technology triage. Rehearsal has only just started, but already, things seemed to be falling apart. First there was trouble with the network that connects the laptops to one another. Then one of the laptops crashed; its human component, a graduate student named Juan Sierra, groans loudly. One of the hemispherical speakers emits a low, crunchy noise, like a fart.
The orchestra members have gathered at Stanford’s Center for Computer Research in Music and Acoustics to rehearse a new kind of musical composition. Together, sitting on meditation pillows in front of MacBooks, they create songs that stretch the definition of music. The orchestra plays laptops like accordions, turns video games into musical scores, and harnesses face-tracking software to turn webcams into instruments. But at this rehearsal, the Stanford Laptop Orchestra (SLOrk) looks less like the symphony of the future and more like an overworked IT department.
“Slorkians! Lend me your ears,” shouts Ge Wang, the SLOrk’s founder and director. He wears a grey T-shirt and black pants, as he does every day, his black hair down to his shoulders. Wang gives the group five more minutes to troubleshoot and then, he says, it’s time for rehearsal to begin.
Fixing a broken network isn’t as simple as a replacing a snapped string on a violin. But in a laptop orchestra, the potential for disaster is part of the delight. Since it was founded in 2008, the SLOrk has been making music that surprises audiences while it subverts the concept of orchestral performance. The compositions, part-machine and part-human, don’t always go according to plan. Technical difficulties are all but guaranteed. Now, as the orchestra prepares for its tenth anniversary show on Saturday, June 9 at Stanford’s Bing concert hall, it’s playing with those same principles—and shaping the next decade of musical experimentation.
Stanford’s Center for Computer Research in Music and Acoustics lives in a Spanish Gothic mansion on campus called The Knoll. Originally the house of Ray Lyman Wilbur, Stanford’s president in 1915, the estate sits on a high hill where two of Stanford’s main roads bisect; from a back window, rolling green hills give way to the horizon.
The program was founded in 1964 by John Chowning, a composer by training who’d come to Stanford for his doctorate in music composition. Chowning had never seen a computer before, so when a colleague showed him a paper about programming instruments with machines, Chowning was intrigued. A few years later, he would create the Center for Computer Research in Music and Acoustics—abbreviated as CCRMA, and pronounced like “karma”—as an offshoot of Stanford’s new AI laboratory. It would be a space for musicians, like him, as well as Stanford’s litany of engineers, scientists, and programmers.
Computer music proved fruitful for Chowning. A few years after founding CCRMA, he would make a major breakthrough by discovering frequency modulation synthesis, the technique used to elicit pure-tone sounds out of machines. It could make a stroke of a key sound like the reed-tone of a clarinet, or make a cell phone ring tone sound like a recognizable song. Chowning patented the technology and licensed it to Yamaha, leading to the Yamaha DX7, the first commercially viable digital synthesizer, and the rise of electronic keyboards. It became Stanford’s most lucrative patent at the time. A few years later, in 1986, the university gave CCRMA the mansion on the hill.
Since then, Stanford’s program has created mathematical models to simulate the crisp sound of a Steinway, or the sliding sweetness of a violin. Other programs sprung up across the nation, in universities like Princeton, Columbia, and Johns Hopkins. The field has devoted considerable energy to reproducing acoustic instruments in digital formats—but it’s also invented entirely new ones.
“Nothing’s better at being a cello than a cello,” says Wang. “So we’re not trying to make a cello. We’re trying to make something you don’t have a name for yet.”
“The question of the future of instruments is an interesting one,” Chowning said in a Stanford press release from 1994, introducing digital waveguide synthesis, which would pave the way for a new class of electronic synthesizers. “Some people think that totally new instruments will be developed and take over. But I don’t think so, because so much of music is tied to repertoire and tradition, which is tied to specific instruments.”
But what if, as an experiment, you took something like an orchestra—a type of musical ensemble steeped in repertoire and tradition—and subverted it with entirely new instruments? What would you learn about the nature of music, the limitations of certain instruments, or the qualities of art that transcend mediums? What would you gain from the unlikely pairing of an orchestra, “an almost archaic institution whose continued existence is something of a miracle,” as computer music researcher Dan Trueman once put it, with the technological newcomer: a laptop?
In 2005, Trueman and fellow Princeton computer researcher decided to see if it would work. The two founded the Princeton Laptop Orchestra, an ensemble of 15 “laptop-based meta-instruments.” (Wang, a Princeton graduate student at the time, was also a founding member.) They dreamed of challenging the very idea of an instrument, of an ensemble, of the relationship between human and machine. An orchestra captured the broader narratives of nations, cultures, modern institutions over time. Could a laptop orchestra provide the next chapter in that story?
The Stanford Laptop Orchestra meets to rehearse every Wednesday night in the spring from 7:30 to 10:30 pm (The late hours are a remnant of Wang’s night-owl habits as a graduate student.) It’s a for-credit course at Stanford—Music 128, cross-listed in the computer science department as CS 170—but getting in isn’t easy. The group of 15 students includes those with computer science credentials, and those with more traditional music backgrounds, but neither is enough to become a great laptop orchestra player. The most important thing is curiosity. “We’re unified by this interest to make music together with computers,” says Wang, “and to figure out what that means.”
Wang likes to call SLOrk a kitchen of sound. “We can go to a restaurant, order delicious food, and enjoy that,” he says. “But there’s a special joy in going back into the kitchen with raw ingredients and being able to concoct your own dish. The process of making—and eating—your own creation carries with it its own satisfaction.”
Every orchestra member gets a MacBook, propped up on an Ikea breakfast tray, with a meditation pillow beside it.
In the ten years that SLOrk has existed, it’s composed over 200 original works and created almost as many new instruments. Most of these works have little in common, but they all start with the same set-up: Every orchestra member gets a MacBook, propped up on an Ikea breakfast tray, with a meditation pillow beside it. The laptop connects to a homespun hemispherical speaker, made by adding car speaker drivers and high-efficiency amplifiers to Ikea salad bowls. (From far away, they look a bit like Minions.) Wang created the speakers during the first year of SLOrk, with an aim to add an acoustic element to an otherwise machine-heavy ensemble. “We want the computer instruments to seem more like acoustic instruments where the sound isn’t coming from a PA system around you but from the artifact itself,” he says. While the MacBooks and cables have been replaced a few times, the hemispherical speakers are the same ones SLOrk used ten years ago.
Every station also includes a GameTrak, a game controller with a retractable cable. GameTraks were originally used in golf simulation video games, where they could turn someone’s virtual golf swing into data points. It was a commercial flop, but computer music researchers immediately saw the appeal. “We bought no less than 100 of them at massively discounted prices,” says Wang.
The device maps movement in three-dimensional space. For a laptop orchestra, that means turning fluid movement into sound value. “It opens up the infinite space of human music, and the dancelike qualities of musical performance,” says Matt Wright, a longtime SLOrkian and one of the orchestra’s instructors. “You can put one in someone’s hands and say, ‘Here. Make an instrument out of this.'”
In past performances the ensemble has used GameTraks to operate video-games that translate into melodic compositions, or finger-plucked the cable like a traditional string instrument. One composition in SLOrk’s upcoming show introduces a new instrument, created by hanging GameTraks upside down on a beam and weighting them with various wooden blocks. Performers push them like swings on a playground to create the song. The performance is wildly playful, like watching kids on a playground discover the delightful sounds of their own laughter for the first time.
One student used a face-tracking program called FaceOSC to turn facial movements into sound.
During the SLOrk term, each student creates their own instruments, composes their own scores, and performs them with the class. There are virtually no rules, other than the limits of imagination and programmability. One student, Kunwoo Kim, used a face-tracking program called FaceOSC to turn facial movements into sound. He and fellow SLOrk member Avery Bick stared into their laptop web cams while opening their eyes wide, or raising their eyebrows, or stretching their mouth to scream, to control the pitch and tempo of the face-tracking instrument.
“Using a face as a controller was a very interesting concept for us,” he says. “We wanted to deliver a human message that uses human parameters.”
Kim came to Stanford after earning a bachelor’s in mechanical engineering and a master’s in electrical engineering. He joined CCRMA because he wanted an interdisciplinary program that would let him continue engineering while also studying music; when he heard about SLOrk, he figured he’d give it a shot.
“I had no idea what was going on,” he says about his first day in the orchestra.
Soon, though, things started to click—and Kim found something in SLOrk that he’d never found before in his engineering coursework. The point of SLOrk isn’t to have a direction. It’s to find a direction.
“The engineering that I have been doing was about solving problems,” says Kim. “But in SLOrk, there’s no problem to solve. We try to cover more of the sentimental side of human beings. And I think that’s very interesting. You’re actually trying to say something about humanity through the computers.”
The nature of computer music means that SLOrk performances can sometimes be hard to grasp. The orchestra’s music often sounds like a chorus of beep-boops, or worse: Some compositions create screechy, metallic sounds, the noise a computer overlord would make when demolishing the human race. Other passages just sound weird, the result of too much randomization from the computer program used to create the song.
“We don’t always like the music we make,” says Wang. “The litmus test is: Is it interesting?”
By SLOrkian standards, “interesting” has a wide berth. A performance that makes use of technology in a novel way (like Kim’s face-tracking webcam instrument) usually qualifies, as do performances that subvert common conceptions of music (like a composition from last year’s concert, where Trijeet Mudhopadhyay invited the audience to open a web page on their phones that made musical sounds). Once, SLOrk students created a piece played by using two Oculus Rifts attached to Leap Motion controllers to track arm movement, which produced some of the sounds. The experiment brought computer music into the virtual world, even if the “music” it created wasn’t exactly nice. Wang pushes his students to think about aesthetics, even when the results are sometimes bizarre.
But SLOrk can also be beautiful. With computers, the orchestra can simulate the sound of a violin 100 feet tall. It can create sliding sounds on static instruments. And it can prototype sound through human gestures—even human expressions—to make music the likes of which we’ve never heard, or seen, before.
“Given that you have infinite options, and it’s very hard to control everything, people sometimes [create compositions] that are crazy,” says Juan Sierra, a masters student at CCRMA and a member of SLOrk. “They don’t need to be crazy all the time. It’s not impossible to create very tonal and very traditional music with computers.”
Sierra, who comes from a background in sound engineering, gravitates toward more melodic sounds. Together with Doga Cavdir, he created a piece for the last SLOrk concert reminiscent of a traditional string orchestra. Cavdir played a solo, stretching the cable of the GameTrak in melodic, emotional gestures. From the back of the audience, or by the look of concentration on her face, you would be forgiven to think she was playing a cello instead of an outdated video game controller.
In the penultimate rehearsal before the tenth anniversary show, Wang teaches the ensemble one of the oldest laptop orchestra compositions. The piece, “Non-specific Gamelan Taiko Fusion,” was composed during the first-ever laptop orchestra performance in 2005, at Princeton. Wang calls it a “classic,” in a genre that’s anything but.
The piece involves the entire orchestra, with each musician running a program on their laptops that looks like a primitive computer game. There are squares in different colors; clicking on them cues various percussive sounds. A human conductor (Wang, in this case) controls the timbre of the bells and directs different members of the orchestra to play at different times. It creates a percussive melody that starts as wind chimes, and then gives way to bossier sounds, like bongos and taiko drums. It doesn’t require much skill on the part of the players, who are effectively learning a new instrument. Just some skillful programming, and a willingness to play along.
Watching them, it’s easy to see beyond the laptop. No one here knows exactly what they’re doing, or even if they’re doing it right. But unlike nearly every other exercise in computer programming, it doesn’t matter. They’re just a group of people learning to play, as if for the very first time.