Hockets for Two Voices


Hockets for Two Voices is a work of effortless beauty & concision. It sounds like Bach and Blade Runner at the same time.” — David Longstreth (Dirty Projectors)

In just over ten minutes of recorded music, Los Angeles-based composer, artist, and instrument designer ​Meara O’Reilly​ can communicate a daunting breadth of creative possibility. That’s no mean feat in today’s soundbite-obsessive world, but it’s part of what makes ​Hockets for Two Voices​ such an unusual and compelling addition to the hypermodern canon of new music.

Hocketing​ refers to the practice of splitting a melody across multiple parts, often in very surprising ways. While the form dates back to the vocal music of medieval Europe, it’s also found in indigenous folk practices from all over the world. For O’Reilly, who grew up in a household filled with classical music, but who was drawn to the study of musical cognition and perception, her pursuit of hocketing is a natural extension of her interests.

“​I feel a little like an outsider in some ways,” she admits, ​“​in the sense that I didn’t come from a conservatory background. I was very much steeped in experimentalism and noise music in college. More recently I’ve found myself crossing over in the other direction, where I’m learning to write for orchestral instruments. I really love the focus and discipline and attention to listening that’s present in classical music, where you have the careful rehearsal of these perfect, beautiful little elements. I'm interested in merging both of these perspectives in my work.”

Hockets for Two Voices​ consists of seven movements, each less than two minutes long, but it took a year to record, requiring a vocal precision that pushed O’Reilly to physical limits she hadn’t yet explored, much less conceived of. At times, each voice alone can be an off-kilter sequence of leaps in pitch and rhythm. Musical clarity only comes when it is combined with the other voice's part. The result, referred to in music cognition as pseudo-polyphony, is the perception that there are more voices at play than in reality. When writing the pieces, O’Reilly drew inspiration from psychoacoustic researcher Albert Bregman, who demonstrated how the limits of our perceptual processes and attention actively shape our experiences of music.

Translating these ideas into musical reality required a new approach. “I had to train to sing in a completely new way for me,” she says. ​“​When I was starting out, I didn't even know if some of these parts were possible to sing. I had to develop the coordination to make such huge yet precise and pointillistic leaps with my voice. I also had to figure out how to capture something so technically challenging. ​I didn’t have someone else who knew the music, so I ended up recording it one voice at a time, in sections, to achieve the accuracy that I wanted​.”

Listening to ​Hockets​ can be an ever-changing experience. Taking cues from Brian Eno and psychoacoustic composer Maryanne Amacher, O’Reilly sought to create a recording that could deliver varying sensations, depending on the listening environment. “The stereo image is a key component of the compositions themselves — they’re ideally heard on loudspeakers with proper distance between them. Depending on where you’re standing in the room, it’s possible to hear the fusion of the parts very differently.” On headphones, this stereo imaging of the two voices comes across with its own stark intensity. O’Reilly’s melodies and cadences are rooted in classical forms (her first love was Chopin, which led years later to a deep fascination with the multivalent work of Olivier Messiaen), but they also seem to morph with each listen, at times revealing nuances that suggest a synthesized voice.

All tracks performed and recorded by Meara O’Reilly in Los Angeles CA, April 2017-June 2018
Engineering by Sonny Diperri, April 21-26 2017
Mixed by Marta Salogni in London UK, March 2019
Mastered by Emily Lazar at the Lodge NYC, assisted by Chris Allgood, April 2019


Hockets for Two Voices
Vinyl / Digital
Release Date: