Icons of Sound
Aesthetics and Acoustics of Hagia Sophia, Istanbul

A collaboration between Stanford University's Center for Computer Research in Music and Acoustics and Department of Art & Art History

Acoustics explores the spatial dimension of sound as it unfolds over time. A sound source produces a sequence of air pressure values over time. A listener will experience each of these pressure values, but not just directly from the source. Sound also interacts with objects, surfaces, and architectural features of the space such that each pressure value radiated from the source carries with it to the listener an imprint of the space. This imprint is called the impulse response, and the process of "imprinting" a space on a sound is called convolution.

Through the impulse response and the mechanism of convolution, listening in may be viewed as much a spatial experience as it is a temporal one. This idea becomes clear if one considers the structure of the arriving sound. The first arrivals have traveled the shortest distances on their way from source to listener, interacting with just a few room surfaces and objects. They convey to the listener a sense of the geometry of the space, source, and position. Later arrivals, having interacted many times with room objects, surfaces and architectural features, bring to the listener an overall sense of the room materials and size.

Impulse response can be collected by using a high energy (intense, loud) sound of short duration. A variety of mechanical and electrical means exist to create this: a gun-shot from a starter pistol, balloon pop, a whip crack, and a mathematically engineered test signal played by a loudspeaker. In Hagia Sophia we have used balloon pop for practical reasons: it has omni-directional propagation pattern; it is easy to use, does not require the carrying of heavy equipment; could be operated by one person; and produces a loud enough to excite the system. We did fours balloon pops, two on May 5 and 6, 2010 and two on December 9, 2010 executed in the space underneath the dome. Two people were involved: a guard of the AyaSofya Müzesi held the balloon and stood 3-5 meters away from the researcher. We recorded the sound using omni-directional microphones attached to the shoulders of the researcher, or set in the hair slightly above the ears.

We have bracketed our Hagia Sophia data with an experiment at Memorial Church on Stanford’s campus. This building also has a domed interior but with much smaller dimensions, different materials, and surfaces: mostly wood and stained-glass. Here we popped balloons and also played electrically engineered test signal, and we recorded the output using omni-directional microphones attached to stands. We replicated similar source and listener positions with both of the balloon pops and the mathematically engineered sound. In comparing the results of these two techniques we were able to understand better the dynamics of the balloon pop data as a means of establishing an impulse source versus a more ideal test signal. We ended creating a method of cleaning the disturbances in the recorded balloon pops and thus arrived at a more precise IR estimation (matching the results derived from mathematically engineered input sound).

© Jonathan Abel

Contextualized Psychoacoustics

Psychoacoustics, the science of human auditory perception, uses experimentation with human participants to measure the perceptual effects of acoustic conditions or events.

In this research, we use the term "psychoacoustically relevant" to denote acoustic phenomenon or characteristics that can generally be perceived by people. Relating the measured acoustic characteristics of spaces with their estimated auditory perceptual implications allows us to generate hypotheses regarding the sensual effects of contextualized sound. Existing psychoacoustic studies can be referenced to give weight to these arguments, and new experiments could be designed and implemented to further inform.

Psychoacoustic experimentation is typically engaged with abstracted stimuli in laboratory conditions, and the auralization models we are developing offer the opportunity to create simulations that make experimental possibilities more realistic. Leading from the work begun here, we see the opportunity to engage two new areas of psychoacoustic research: 1) experimentation using perceptually realistic simulations of acoustic conditions and musical stimuli; and 2) "ecologically valid" historical psychoacoustic experimentation -- that is, with realistic stimuli and/or conditions, and informed by historical knowledge.

© Miriam Kolar