If most albums are postcards from a moment, Electronic Music 1994–2021 (recently released by Elli Records and reviewed here) is more like a meticulously hand-stitched quilt assembled across three decades – stitched in the dark, with occasional bursts of lightning, and an insistence on getting the seams just irregular enough to be interesting.
Roads, whose name has long been synonymous with granular synthesis and microsound, doesn’t just compose music – he architects it from the atomic level upward. The result is less a catalogue of pieces than a slow-motion evolution of ideas: early pre-microsound experiments converse fluently with later micro-montage works, not unlike old friends meeting after years apart and discovering they still finish each other’s sentences.
In conversation, Roads is as candid about obsolete mixers and $8,000 DAWs as he is about the philosophical patience it takes to work for decades on a single piece. Here, he talks about coherence without homogeneity, the pleasures of imperfection, the stubborn beauty of the Bohlen-Pierce scale, and the stubborn limitations of current AI music tools.
It’s a rare portrait of an artist for whom technology is never just a tool, but a co-conspirator – and for whom the voyage is at least as important as the destination.
Chain D.L.K.: The album Electronic Music 1994–2021 spans nearly three decades. What guided your decision to bring these specific works together into one collection?
Curtis Roads: Emanuele Battisti of Elli Records approached me several years ago about doing a new album. I completed several pre-microsound pieces from 2016 to 2021, but it was not enough material for an album.
I started looking in the archives for material. Purity (1994) had been released on the CCMIX CD on Mode Records in 2001. The rest of Clang-tint was released in a limited edition vinyl LP on the Slowscan label in 2021, but was not in streaming format. Sculptor (2001) and Touche pas (2009) are works from my microsound period. Modulude (2021) and Bubble Chamber (2021) had not been released.
This spring I gave concerts in Italy and France. I performed the album with new videos by BrianO’Reilly. This was a test, and it sounded like a coherent set. The music does not all sound the same, so there is contrast, but there are elements throughout that are consistent with my style.
Chain D.L.K.: How do you perceive the dialogue between your pre-microsound pieces and those using granular synthesis and micro-montage?
Curtis Roads: In the pre-microsound pieces (Clang-tint, Still life, Modulude, Bubble chamber) I was developing my techniques for creating phrase structure and counterpoint in electronic music. In the granular and micro-montage pieces, I was applying phrase structure and counterpoint to microsonic materials.
Chain D.L.K.: You’ve mentioned that your compositions often take years to complete. What kind of temporal or reflective space does that extended process allow for?
Curtis Roads: Some lucky pieces come together in months (for example, Tenth vortex), but most take years. I want to make something extraordinary. It is difficult work. To begin, one is trying to solve an undefined problem. I do not use a system or a master plan.
I start with an idea of the sound materials I would like to compose with. I make some sounds. Then I try to organize them. I put up temporary scaffolding and solved a local issue while also trying to plan the macro form based on structures I have already composed. The music is detailed. One of my methods is micro-montage, working with individual particles of sound to build up larger structures. It takes time.
Chain D.L.K.: How do you know when a piece is finally finished – especially after decades in development, as with Modulude?
Curtis Roads: A piece is finished when I have nothing more to add or subtract. I do not aim for perfection. For example, my piece Pictor alpha is a sequence of looping pulsar phrases. Most of the piece features a single voice. At 2 min 42 sec I added a contrapuntal voice that is asynchronous to the main voice – an improvisation in the studio. It seemed to work despite it not being perfect – a lucky accident. A part of me wanted to redo it to make it perfect, with everything aligned like clockwork. I played it for Joel Chadabe, who told me to not change it. This was good advice. My music has elements of precision, but too much precision can be tedious in music. Thus, I also seek elements of playful and loose feeling.
Not aiming for perfection helps me to finish projects. I did my best at the time. I can let them go.
Chain D.L.K.: Looking back at works like Clang-tint or Still life, how did the tools available to you in the early ’90s or ’80s shape their sound?
Curtis Roads: In my student days at CalArts and UCSD, I would take my tapes to a professional recording studio in Los Angeles to perform the master mix. In that period that there was a huge distinction in audio quality between professional audio recording gear and anything else. A professional mixer and 16-track multitrack tape recorder could easily cost $100,000. University studios could not afford this level of investment. So most had amateur gear.
Inexpensive audio mixers were especially terrible, like the Tascam Model 10 introduced in 1976 for $1900. A studio at UCSD had one so I knew how noisy and nasty they were. To my disappointment, the MIT Experimental Music Studio had a model 10 when I arrived in 1980. I had a commission to make a new piece. Rather than try to make the piece at MIT EMS, I booked a recording studio in Boston with a professional mixing console and a 24-track analog tape recorder for an eight hour session. This resulted in my piece Field (1981), most of which was assembled in one night. The professional studio had high-quality power tools.
For Clang-tint, which was conceived in 1990, I used one of the first digital audio workstations (DAWs), the MacMix software running on Studer Dyaxis hardware. This was before Pro Tools. The Dyaxis cost me $8000 in 1988. I had to sell my beloved Studer A810 analog tape recorder to buy it. This was my transition from analog tape to digital editing. It completely changed the methodology. Suddenly, one had dozens of tracks, sounds could be freely moved around the timeline, and you could do it in your home studio. MacMix became obsolete so I switched to Pro Tools to assemble my pieces.
Chain D.L.K.: In what ways has your compositional relationship with technology changed since the release of POINT LINE CLOUD?
Curtis Roads: POINT LINE CLOUD was first released twenty years ago. My working method has not changed much in the period since. I have my custom granulators and the PulsarGenerator application. I use Pro Tools and plugins. I use my Studer A807 tape recorder for tape echo feedback effects. Artificial intelligence (AI) technology is now all the rage. In 1985, I published a paper on music and AI in a computer science journal that was widely distributed (Roads 1985). This was the era of rule-based symbolic expert systems using if-then logic. The symbolic paradigm was swept away by neural-network based AI algorithms based on training from millions of examples.
Neural network technology can handle tasks like face recognition that are too complex to encode
using manually-written if-then rules. As far as the new AI tools go, I worked with several graduate students on projects in 2020-2024 but we did not achieve great results. The musical application of AI technology is still in its
infancy.
I look forward to someone developing an application of AI to music that I can use in my workflow. I can imagine many applications. In 1981 at MIT I wrote a proposal for “An intelligent composer’s assistant.” To give an example, I would like to see an AI app that could automatically make a graphic score of my music where I could choose symbols and edit the results. It would be great to have smart sound tools that one could talk to like an assistant. For example, you give it a sound a have it make a dozen variations. Or do the same with a phrase. This could be powerful. Film composers have been using human composer assistants for decades.
Chain D.L.K.: How do you see the current landscape of granular synthesis and microsound evolving, especially with the rise of real-time tools and plugins?
Curtis Roads: Our CREATE team produced the EmissionControl2 software in 2020. It was designed as a laboratory instrument that will do anything that can be done with granular synthesis. However, it has over 80 controls on its front panel so there is a learning curve. Due to the vastness of the parameter space, numerous settings yield unremarkable results. We incorporated an extensive preset system for users, but we did not make a library of presets for users to try. So everyone using has to start from scratch.
My PhD student Raphael Radna and I are making a new granulator that he would like to make into a commercial plugin. It has fewer controls but adds some interesting new functionality like support for MIDI note performance, meaning you can play it on a keyboard. It also offers full support for microtonal scales. It will be shipped with pre-made presets.
Chain D.L.K.: “Bubble Chamber” is a new piece in this collection. What was the conceptual or sonic impulse behind it?
Curtis Roads: As I wrote in the program notes of Bubble Chamber: “I have long been inspired by the beauty of the images produced by a bubble chamber–a device for photographing subatomic particles. These images show intricate interactions as particles enter the chamber at high speed, leading to collisions in which some particles break apart or spin off in strange directions.
These interactions serve as metaphors for musical processes in Bubble Chamber. This composition focuses on the interactions of sound particles with noises and tones. In this work, a collision between sounds often terminates a phrase. The result is a cadence that resonates with the deep energy of the collision”.
Chain D.L.K.: Touche pas exists both as a standalone piece and as part of an audiovisual work. How do you think image and sound interact in that context?
Curtis Roads: My pieces can exist in multiple formats. I create them in stereo in my home studio, Lemonwood. (I have several lemon trees.) The works are whole and complete in the beautiful medium of stereo. They are designed to be heard on good loudspeakers, but of course many will hear them on headphones or even ear buds. It is not the same as hearing them on loudspeakers. When I perform live I spatialize the pieces in what I call pluriphonic configurations of multiple loudspeakers. This is the process of upmixing, from stereo to an octophonic circle of eight loudspeakers or an immersive 360-degree system like the UCSB AlloSphere, with 54 Meyer sound speakers and a subwoofer. So pluriphonic sound projection is another format. Finally, I have been working with video artist Brian O’Reilly for many years. I send him the completed audio work and he makes an interesting video to go with it. I met Brian at Les Ateliers UPIC in Paris in the 1990s, where he was working as musical assistant to composers like Luc Ferrari. He currently teaches improvised and electronic music in Singapore, but he has always been one of the top video artists in the field of visual music (sound and moving image).
So this is yet another format for my music. The visuals help some people appreciate the music. Brian made some new videos for our recent shows in Parma and Milano.
Chain D.L.K.: Purity, which originally appeared on the CD CCMIX Paris (Mode 2001), is now recontextualized. Has its meaning or role changed for you?
Curtis Roads: No. It was the piece in which I became fascinated with the Bohlen-Pierce scale, which is both more sweet and more sour than 12-note equal temperament. I am currently working on a new granular piece based on the Bohlen-Pierce scale.
Chain D.L.K.: How would you describe your aesthetic concerns over the decades – have they remained consistent or shifted with your techniques?
Curtis Roads: I would say they are more-or-less consistent. Early in my career I was inspired by Edgard Varese’s vision for “The liberation of sound.” This remains a guiding light. I have taken a granular approach to sound since the 1970s. But I have also been interested in many other research directions. I was able to explore the Bohlen-Pierce scale, for example, but I am now at an age where I do not have the luxury to take years to explore a new research area.
For example, I am fascinated by the world of subharmonic tones and rhythms. This is beautiful sound world but it is completely undocumented. This means that anyone who wants to explore this world is starting from scratch. Oskar Sala pioneered it with his Trautonium instrument, but it could be taken much further. I began a piece based on subharmonics and then realized that it would take possibly years of experiments to understand how to organize the harmonies. Sala did not publish any theory. His magnum opus, Subharmonische Mixturen came out in 1997. He had been working with subharmonics for 65 years. I have other important projects, so I decided to leave this for others to explore.
Chain D.L.K.: How do you relate to the idea of “beauty” in electronic music? Is it a useful or limiting concept?
Curtis Roads: I prefer the term “aesthetics” as it incorporates both beauty and ugliness. My text Composing Electronic Music: A New Aesthetic (Oxford 2015) is a 480-page book of aesthetic philosophy. Aesthetic perception is a human survival mechanism. We reject things that taste or smell toxic: it might kill us. If it looks scary or ugly, we avoid it. If it looks, smells, tastes, sounds, and feels beautiful, it probably is great for us. The problems of the planet Earth can be seen as issues of aesthetics. For example, ugly invasive destruction versus beautiful peaceful coexistence.
There is also a highly subjective side of aesthetics. Each person has their own taste or preferences, which vary in time. Most of these preferences are unconscious in origin. A person prefers or likes something without knowing why.
Chain D.L.K.: How do you view your own work in the broader lineage of electroacoustic music and computer music history?
Curtis Roads: My work continues the direction of the mid-century modern composers Varese, Xenakis, and the middle-period Stockhausen among others. These composers left behind a legacy of concepts, sounds, and techniques for other people to take to the next level. I hope that I can leave behind a legacy that others will find helpful to develop further.
Chain D.L.K.: Were there artists or composers you were in dialogue with – explicitly or implicitly – during these years?
Curtis Roads: I was an editor of Computer Music Journal for 23 years and co-founded the International Computer Music Association. I worked in several music departments and music institutions internationally. I met a lot of composers. As stated, I was influenced by Varese, Xenakis, and Stockhausen.
I discussed composition with Xenakis and Horacio Vaggione, Luc Ferrari, Bernard Parmegiani, and James Dashow. I admire Natasha Barrett’s work. I remain in touch with Morton Subotnick and John Chowning. Obviously, there are many others, too numerous to name whose work I appreciated.
Chain D.L.K.: Do you think the act of composing with sound materials has become more accessible, or simply more complex?
Curtis Roads: Definitely more accessible. It does not need to be more complex. In my classes I would play the Six Preludes for Magnetic Tape (1966) by Ilhan Mimaroglu to show how evocative pieces can be made with simple materials and techniques.
Chain D.L.K.: What would you want new listeners – perhaps unfamiliar with your earlier releases – to take away from this album?
Curtis Roads: The composer is at the helm of a ship sailing listeners on a voyage. Each listener gets something different from the experience. It does not matter if the listener is unfamiliar with my previous work or whether they are familiar with electronic music in general. The music reaches out for connection, and some people are open to this aesthetic experience.
Chain D.L.K.: After nearly 30 years of composing across formats and technologies, what continues to inspire you to make new work?
Curtis Roads: I do not know where inspiration comes from. The motivation is largely unconscious. I wanted to play music since I was a child. In college, I studied music without any expectation that it would lead to a music job. It did not. My first job out of college was in the software industry. But the goal was always to find a way to make a living in music. Then the path of publishing opened up with Computer Music Journal. Later, the path of research opened up at MIT. Finally, the path of teaching opened up at the University of Naples Federico II.
Later, in my books, teaching, software, and composition, I felt I had something to say that no one else was saying. Besides the benefits to me of meaning and purpose, I believe that what I am doing is a service to others. Making art is a service to humanity.
Visit Curtis Roads on the web:
https://www.curtisroads.net/