A Documentation Framework for Sound in Time-based Media Installation Art

Amy Brost
The Electronic Media Review, Volume Five: 2017-2018


Perhaps because the visual arts predominate in museum collections, conservation documentation methods for audio in time-based media installation art are less developed than those for video. However, the aural elements are equally complex, both technically and creatively. This paper proposes a framework for documenting sound in time-based media installations, from the artist’s vision for the listener experience of the artwork to its realization in the exhibition space. For works involving sound reinforcement—electronic amplification of audio signals—artists often work closely with audio engineers or other sound specialists to make adjustments that shape the sound when the work is installed. This documentation framework was applied to one such artwork, “The Visitors” by Ragnar Kjartansson, as a case study. The framework and excerpts from the case study show how documentation focused on sound elements can yield unique insights about an artwork. For time-based media works that foreground sound, conservators must be conversant with sound physics, acoustics, audio engineering and sound design so that they can have productive conversations with artists and sound professionals in order to create meaningful documentation of the significant aural properties of an artwork for the future. In this way, conservators can approach time-based media installations not only attending to how they should look, but also to how they should sound.


Modern and contemporary art practice has incorporated sound in diverse ways. Some artworks incorporate incidental sounds from the surrounding environment into the work, while others must be acoustically isolated. Some artworks have kinetic, mechanical, electronic, or other activity that produces sound. Others employ unique modes of listening, such as bone conduction, or heighten awareness of hearing by exposing listeners to unusual sonic experiences, including anechoic environments or frequencies near the limits of human perception.

A subset of artworks with sound comprises those with pre-recorded audio. In the mid-twentieth century, sound recording technology became more widely accessible, and artists began to incorporate sound recordings into their practices. The documentation framework that follows is structured specifically for time-based media installations with digital sound recordings. However, some of the parameters in the framework would lend themselves to documenting diverse artworks with sound elements. There is much more work to be done across this spectrum.


Documentation is a primary conservation strategy for time-based media artworks. This is due to the particular nature of these works, which are inherently transitory. Unlike artworks in traditional media, time-based media works only exist when they are installed. When not on view, they may be physically dispersed across the museum’s various storage environments: files in a digital repository, sculptural elements in crates in the warehouse, and media equipment in the audiovisual storage area. In some cases, crucial parts of the installation are not in the museum at all—they are procured or constructed anew each time the work is exhibited. Artists, their collaborators, assistants, galleries, and museum staff are all vital sources of information essential to understanding how to faithfully present these artworks. By collecting and organizing this information using formalized documentation strategies, conservators work to preserve the integrity of artworks whose material nature is subject to change.

Furthermore, the artist does not make every decision that affects how the work is shown. Conservator Glenn Wharton has described it this way—that artists transfer some measure of “interpretive authority” to the museum (Wharton 2016, 27), so that the fully realized work is the product of collaborative decision-making. Each time the work is shown, the museum exercises this “interpretive authority” in a different way as diverse perspectives are brought to bear on the realization of the work. Within the parameters defined by the artist, museum staff may reconfigure or reinterpret the work in light of a wide range of variables. These may include the space into which the work will be placed; the curatorial vision for the installation and the exhibition; the work’s relationship to other works nearby; or the views and interests of artists’ estates or studio assistants. Museum staff may also continue to refine their understanding of these parameters with guidance from the artists themselves, who may offer their perspective from a distance or be very involved—even physically present—to oversee the installation. Especially vulnerable to change are any electronics in the work, particularly those used for display, as these technologies eventually become obsolete. Depending on their aesthetic and conceptual significance, playback and display equipment in the artwork may be treated differently in different artworks. Both the degree of variability acceptable to the artist and also the properties of the equipment that are essential for the faithful presentation of the work must be considered (Phillips 2012). In light of this inherent variability, time-based media installations evolve each time they are shown. Only robust documentation can give future conservators an understanding of how the work has changed over time and a basis for collaborative discussions on how best to present these artworks.

Media conservators at the Museum of Modern Art in New York have implemented an array of documentation practices developed internally and adapted from other institutions. Among them is the model developed by time-based media conservator Joanna Phillips at the Solomon R. Guggenheim Museum in 2011. As articulated in her paper “Reporting Iterations: A Documentation Model for Time-Based Media Art,” Phillips describes two complementary conservation documents: the Identity Report and the Iteration Report. (Phillips 2015)

The Identity Report aims to be a complete dossier of the entire work, using the artist’s instructions received at acquisition as a starting point. This report includes the work’s complete production and exhibition history, all the elements it comprises, considerations for placing the work in a range of environments, all acceptable modes of presentation to date, preservation concerns, technical aspects and important contacts. This is an organic, living document, and its evolution parallels that of the artwork.

By contrast, the Iteration Report aims to be a comprehensive record of a single installation. It is made at a discrete point in time, and is a snapshot in the life of an artwork. This report lists everyone who had an impact on the appearance of the installation and their contributions. It also includes a record of all the physical parameters of the exhibition space and every significant element that was part of the installation. It includes an evaluation of the iteration, so that the new information incorporated into the Identity Report following the installation will be appropriately contextualized. As the work continues to be installed and installation instructions evolve, the specifics of each installation are recorded in a separate Iteration Report. These in turn inform updates to the Identity Report. The documentation framework includes prompts for collecting and reporting information related to sound elements. However, for those works that foreground sound as a central conceptual and creative element, a greater level of detail in documentation may be required. The below framework expands on the prompts for sound elements, which in turn can yield new insights about an artwork and also preserve the significant aural properties for future exhibitions of the work. In this way, conservators can approach time-based media installations not only attending to how they should look, but also to how they should sound.


The sonic environment of a time-based media installation artwork is the product of a complex system consisting of a sound file, electronics used to amplify or alter the sound, the space in which the sound is made, and the listener or receiver of the sound. Artists frequently have specifications for how their work should sound in the exhibition space, and they may work with audio engineers and other sound professionals to realize that vision. The first part of the framework (Part I) is for documenting the sonic identity of the work, which could be incorporated into an Identity Report. The second part of the framework (Part II) is for documenting a specific installation of the artwork and could be incorporated into an Iteration Report.


The understanding of the sonic identity of the work is often acquired through interviews or questionnaires sent to the artist at acquisition. Interviewing artists about the role that sound plays in their work requires thoughtful questions specifically focused on sound. Below are some questions based on pre-acquisition questions used at the Museum of Modern Art in New York (MoMA 2018), but with the focus shifted specifically to sound. Questions may include:

-What is the desired listener experience of this work?

-What are the typical sound challenges when this work is installed?

-What are the typical sound adjustments when this work is installed?

-What spaces are acoustically ideal for this work? Are any unsuitable?

-How should visitors listen to the work (e.g., speakers, headphones, etc)?

Artists, or certainly the sound specialists collaborating with them, may also be able to provide technical details and guidelines for how to faithfully install the audio aspects of the work. These questions can help draw out information about the master material being acquired for the sound elements, the rationale for equipment choices, and the step-by-step procedure to shape the sound in the space once the work is installed. Questions may include:

-How was the sound recorded? Please describe the production process, recording equipment, and the specifications of the files generated.

-What is the master audio file format?

-What file formats are used for exhibition?

-Do you have preferences regarding the playback equipment used for exhibition? If so, please describe.

-Do you have preferences regarding the sound equipment used for exhibition (e.g., speaker type/placement)? If so, please describe.

-Do you have guidelines or instructions for how to adjust the sound for the exhibition space that could be used to guide future installations?

-When the sound is adjusted during installation, how do you know when it is right?

With a combination of research into the work and answers to these kinds of questions, conservators can begin to organize the information as described below.

Part I: Sonic identity

A. Sonic environment

In this section, the conservator can articulate the significance and character of the sonic environment of the work based on an understanding of the artist’s instructions and intentions. While artworks range from thinly to thickly specified in terms of acoustic requirements, this section is a place to make a summary statement about the role that sound plays in the work creatively and conceptually, and in broad strokes, how the effect is achieved. Because artworks employ sound in diverse ways, a concise statement about the sonic environment raises acoustic awareness and engenders consideration of where and how the work should be situated to deliver the intended listener experience.

B. Sonic parameters

In characterizing the sonic parameters, it is helpful to begin with a survey of all the sound-producing and sound-influencing elements of the installation, noting the location, behavior, and significance of each one. Then, the desired behavior of the sound in the space is described. Finally, reception of the sound and the listener experience are considered. This includes describing the ideal listener experience, how the work is tuned, and the modes of listening appropriate for the work.

In addition to drawing from existing conservation documentation terminology, the below prompts are informed by the ‘source to sink chain’ created by Michael Cohen, a Computer Science professor in the Spatial Media Group of the Computer Arts Lab at the University of Aizu in Japan (Cohen 2016, 279). His source to sink chain describes an audio augmented reality system. In this way, the specific vocabulary developed to describe all aspects of a virtual sonic environment can be adapted to document an actual one. This is not an exhaustive list—Cohen’s model contains additional variables and conservators will no doubt encounter parameters to add—but the below starting point encourages careful examination of the sonic character of the artwork. As time-based media conservators continue to document sound, terminology will continue to develop that addresses the specific needs of artworks.

1. Acoustic events/sources and influencers

a. Sound sources and influencers. List the sound sources and influencers in the work. Sound sources include speakers as well as physical objects in the space that produce sound, as well as environmental sounds, voice, music, auditory icons, etc. A sound influencer is an object or form that does not produce sound but intentionally affects it

b. Spatial relationship. Describe the spatial relationship between the sound sources and influencers (location, direction, distance) if specified, and note any instructions provided by the artist on how to approach the layout of the aural elements in the exhibition space

c. Behavior. List any specific behavior of each source or influencer (e.g., do they move; are they interactive, continuous, or intermittent; are they directional or omnidirectional, etc.)

d. Significance. List the conceptual significance of each sound source and influencer, and what aspects are mandatory or optional when the work is installed

2. Space/medium

Note any of the following sound behaviors in the work and their significance:

a. Directionality. Sound radiation pattern and desired effect

b. Obstruction/occlusion. Objects or strategies intended to diffract or bend the sound

c. Reflections. Echoes

d. Reverberation. Prolonged sound/resonance

e. Outdoor display considerations. Relevant aspects of sound propagation outdoors, if applicable

f. Medium (if other than air). Medium through which the sound propagates

3. Listeners/sinks

a. Ideal listener experience. Describe the artist’s ideal listener experience of the work (orientation and movement of the listener; location, direction, sound intensity/levels, presence or absence of low-frequency vibration, etc.). If the artist works closely with a sound professional, ask that person to describe this as well.

b. Tuning. Note any instructions provided by the artist on how to adjust or “tune” the audio for the exhibition space. Both technical detail from a sound specialist and subjective descriptions of the desired effects are valuable.

c. Additional listener experiences. If the experience of the work is fundamentally different depending on listener position, describe the key listener positions and the desired listener experience.

d. Modes of listening. Note the acceptable modes of listening (e.g., headphones, receiver/headset, loudspeakers, speaker array, or other)

Cohen also associates modes of listening with the listener’s reception of the sound (Cohen 2016, 298); for instance, headphones essentially provide a private or personal listening experience. Loudspeakers provide a multi-personal experience that may feel intimate yet communal, while large speaker arrays evoke modes of public address that may feel impersonal or even formidable. Artists often specify how their work is meant to be heard, and different modes of listening can reinforce meaning in different ways.

C. Media

  1. Inventory. List all sources of pre-recorded sound that are part of the artwork (e.g., magnetic tape, vinyl, digital file).
  2. Production history. Describe the production setting in which the sound was recorded. Knowing how the sound was recorded, including the recording equipment used and native format of the recordings, helps conservators understand the media that is part of the work.
  3. Characterization. Extract technical metadata with an application such as MediaInfo from the digital files and note file characteristics (e.g., filename, codec, container, sample rate, bit depth, bit rate, number and name of audio channels, duration). Examine the audio using software designed for audio (e.g., Adobe Audition) and appropriate listening equipment. Note waveform characteristics including dynamic range in dB FS (decibels relative to full scale) and frequency range.
  4. Condition assessment. Note condition issues with the time code where they occur. Software such as Adobe Audition can display waveform information for examination purposes to reveal any clipping, phase, or frequency issues. This or similar software may also be used for conservation treatment of audio if appropriate.

D. Sound System

  1. Specifications. List the equipment specified or required for playback and output. Include all equipment, whether concealed from view behind a wall or visible in the exhibition space. Note the status of the equipment (e.g., whether or not it has aesthetic or conceptual significance, whether it is part of the artwork or must be procured each time the work is shown, what variability is acceptable to the artist, and what properties are important)
  2. Cabling. List or diagram the signal path from playback to output.

Elaborating on methods of characterization and condition assessment is outside the scope of this paper, but understanding the technical characteristics and condition of the audio is essential for preservation purposes. When performed during the acquisition process, this examination can guide discussions with the artist about the significance of certain characteristics and audio artifacts and errors, whether or not treatment should be performed, and whether or not additional master material might be available to the museum or collector to ensure the highest quality display of the work in the future.

The information gleaned from characterization and condition assessment has numerous practical implications. The number and names of audio channels are particularly important because of the relationship to equipment and channel layout in the exhibition space. For example, if a file has two-channel audio but the installation floor plan indicates eight speakers, the artist may want to indicate which of the two sound channels should be output to each speaker. In other cases, terminology can cause misinterpretations. Artists may indicate that their work has “surround sound,” which may mean that the listener experience of the work should be immersive, or it could mean that the equipment required to show the work is an actual surround-sound system. If an audio file has a single audio channel and should have “surround sound,” then that same channel can be output to multiple speakers surrounding the listener. However, a 5.1 surround sound system requires six discrete audio channels within the digital file. Those six channels are typically named Center, Front Left, Front Right, Left Surround, Right Surround, and LFE (low-frequency effects, the 0.1 channel) to indicate which channel is output to which speaker in the exhibition space (Figure 1). Characterization and condition assessment must take place in order to have timely dialogues with the artist to avoid the confusion that may otherwise arise during installation.

Figure 1. 5.1 Surround Sound. Subwoofer not pictured. Kamina [CC BY-SA 3.0 (https://creativecommons.org/licenses/by-sa/3.0)] via Wikimedia Commons [https://commons.wikimedia.org/wiki/File:5-1-surround-sound.svg]

Compatibility issues between the audio files and output equipment can affect the quality of the sound heard in the space. The installation team must be made aware of the technical characteristics of the files in order to procure the appropriate equipment to play the audio. Media players may require files in a specific codec or prepared in a particular way (e.g., BrightAuthor for making BrightSign presentations for the BrightSign player and digital signage monitors currently popular for exhibitions). Many media players utilize compressed files; however, decks utilizing solid state hard drives for playback are becoming more accessible, making it possible for some works to use full resolution uncompressed files for exhibition as well as preservation. Equipment instructions from the artist can only be thoroughly understood by examining the digital files.

Another audio file characteristic is frequency range. Frequency range should be considered when choosing equipment to faithfully play the file. Different sizes of speakers create different frequencies, which is why mid-range speakers are often supplemented with small “tweeters” to create high-frequency sounds and large “woofers” to create low-frequency sounds. There is a full range of transducers available to create all the sound frequencies in the audible range that might be in the digital file itself, but not all may be required to install the artwork. For example, consider a cat’s purr, in the range of 20-40 Hz. While audible in real life, this frequency in a digital file can only be created with a subwoofer. Without a subwoofer, this frequency in the file cannot be reproduced or heard at all. Sometimes audio files provided by artists contain low frequencies, but the artist prefers not to have a subwoofer in the exhibition space. Perhaps these cases are analogous to matting decisions, where edges of a work may not be visible while on exhibition because the artist has specified an overmat. Questions about audio channel layout, output, and equipment choices are best asked or considered well before exhibition or loan. 

Part II: Installation Documentation

The below can supplement and contextualize the information collected in the Iteration Report, which includes exhibition file information, equipment used and its status, cabling and placement of equipment, equipment settings, and documentation of the decision-making process.

A. Acoustic environment

Details about the space and its acoustics are often needed to contextualize the settings on the sound equipment. Adjustments to the equipment are related to how the sound behaves in the space. This means noting the dimensions of the space as well as surface materials, such as masonry or sheetrock, the presence or absence of windows, furniture, carpet and upholstery, acoustic treatments, and so on. Each of these materials has different sound absorption properties at different frequencies of sound. A poured concrete gallery has very low sound absorption generally, while a room with drapery on the walls and heavy carpet will absorb much more sound. Details about the decision making process to arrive at sound settings can shed light on what aspects of the acoustic environment affected the audio output and necessitated those adjustments. For example, noting that low frequencies were minimized in one installation does not indicate that they always should be lowered. This could be a response to unique room acoustics. In another case, raising the low frequencies and adding as much bass as possible might be important to the artist for every installation. Arriving at an understanding of sound equipment settings requires an understanding of the acoustic environment as well as a discussion with the sound professionals involved in the installation and the team members involved in decision making.

B. Sound documentation

  1. Media information. Exhibition media characteristics, including digital file names and technical metadata, are captured in the Iteration Report, along with documentation of any decision-making processes that influenced those characteristics.
  2. Channel layout and cabling. Aural aspects of the work are easily overlooked in visual documentation of a specific iteration. Floor plans and diagrams showing speaker placement should also include which audio channel or file was output to each speaker. The audio signal path with settings and adjustments for each audio channel should be noted in cabling descriptions or diagrams.
  3. Live recording. In addition to documentation of the physical elements such as equipment, cabling and settings, additional strategies of live recording or measurement may be appropriate. For the project Inside Installations: Preservation and Presentation of Installation Art, Robert Hermann, in partnership with Reinhard Bek at Museum Tinguely, created the guide “Recording the Sound of Installation Art Objects.” (Hermann 2007) In it, Hermann suggests a rationale for when it makes sense to record, how to approach recording installations, and choosing appropriate equipment for the task. He suggests that recordings are not always necessary, but can be particularly valuable in several cases. The first is when artworks have sound as an integral element that can be influenced by the process of constructing the work, as with elements remade every time the work is shown (e.g., a bubbling liquid). A second case is when sound is integral and, while not influenced by construction, is particular to the exhibition space. These works may sound very different in different spaces, such as mechanical works installed in different environments, or they may produce a chorus of sounds impossible to experience except while the work is fully installed. Thirdly, recordings of mechanical or electronic sounds can also be a valuable reference for when repairs are needed. Lastly, while some works produce sound as a byproduct of their operation, if this sound is incidental, whether or not it requires documentation could be a question put to the artist. Hermann also explains how to choose equipment and recording positions, how to combine visual and sound recording strategies, and how certain types of microphones, including dummy head or binaural microphones, could be useful to document the listener experience. He raises the important point that plans must be made to preserve the recordings themselves. Their authority as references depends on successful storage, retrieval and faithful playback.
  4. Sound data collection. To measure actual sound levels, a representative listener position can be chosen and a sound level meter set to collect readings over time, correlated to audio file timecode or perhaps a video made simultaneously. This can indicate the dB SPL (dB of sound pressure level) experienced by a typical listener. However, this data should be contextualized with a rationale from the artist and documentation of the decision making process for installation. Seldom is such data collected in order to suggest that copying those exact readings is desirable for future installations, although in some cases that may be the goal—it depends on the artwork.
  5. Subjective listener experiences. To complement this data collection, it may be useful to collect subjective listener impressions as well, to account for psychoacoustic factors in addition to sound levels. Depending on the listener’s age and hearing condition, loudness and other sound perceptions may vary; whether a sound is pleasant or not can depend on the listener’s cultural background and personal experience. The stated objectives of the artist for the sonic identity of the work are challenging to execute for the same reason—descriptive terms often lack absolute meanings when it comes to sound. Much like paper conservators working to define the point at which undulations are actually cockling, audio installation tuning instructions and sound documentation often incorporate terms without clear definitions and boundaries. Sound may be ambient, immersive, aggressive, tinny, bright, boomy, muddy or one of many other descriptors. These terms often loosely correspond to technical characteristics better described with vocabulary describing frequency characteristics, harmonics, distortion, and the like. Terminology development in this area is needed to improve the quality of conservation documentation.
  6. Iteration-specific modifications to sound. The Iteration Report can capture what, if any, adjustments were made to the sound that should not be considered integral to the work in the future. Any iteration-specific modifications to modes of listening that differ in some way from prior understanding of the artwork should be noted (e.g., perhaps the work requires loudspeakers but the artist permitted headphones in a specific case). Noting those modifications and shortcomings will prove useful to the next team installing the work.


“The Visitors” (2012) by Ragnar Kjartansson (b. 1976) is jointly owned by the San Francisco Museum of Modern Art (SFMOMA) and the Museum of Modern Art (MoMA) [https://www.moma.org/collection/works/170621]. It was installed for the “Soundtracks” exhibition (July 15, 2017 – January 1, 2018) at SFMOMA [https://www.sfmoma.org/read/ragnar-kjartansson/]. This large-scale, nine-channel, hour-long video performance piece has been exhibited worldwide to great acclaim. The setting of the work is a stately, aged mansion in rural upstate New York. The artist gathered fellow musicians there in 2012 to perform an original composition with lyrics inspired by the writings of the poet and performance artist Ásdís Sif Gunnarsdóttir, his ex-wife. One of the themes in the work is the break-up of their marriage, giving the piece tremendous emotional range which has a corresponding broad dynamic range in terms of sound. Life-sized video projections of the individual musicians encircle the visitor, whose experience ranges from contemplation of solo performers to immersion in the music of the entire ensemble. The piece involves vocals and numerous musical instruments including two pianos, drums, electric and acoustic guitars, bass, banjo, accordion, and cello, as well as the sounds of the natural landscape, punctuated by two cannon blasts. The author visited SFMOMA and also conducted an interview at MoMA with Christopher McDonald, a sound engineer and also the artist’s Director of Sound, who was onsite at SFMOMA during the installation.

Shown dozens of times, the work is versatile in terms of where it can be installed. It has been shown in a range of architectural and acoustic environments varying in size and arrangement of video channels. The objective of the interview was to learn more about the master material, the requirements for the work, what variations were acceptable, and how Mr. McDonald approached the channel layout and final sound adjustments of the work. The below excerpts from the sound documentation of “The Visitors” have parentheticals indicating how information corresponds to the documentation framework described above.

Part I: Sonic identity

A. Sonic environment

Mr. McDonald approaches every installation as essentially “mixing a new album,” or at least approaching it like a reissue, where the goal is to be faithful to the original but do something new in the circumstances at hand. The exhibition space should be acoustically treated as described in the installation instructions with acoustic paneling or curtains (e.g, duvetyn) and if possible, carpet, which helps with sound absorption as well as light reflections. The studio will often provide advice on the environment based on information about the space. The space should be as acoustically isolated as possible. Each of the nine video channels has mono audio output to a single loudspeaker located at the top center of the screen, although larger exhibition spaces may require a second one at bottom center for adequate coverage. One or two subwoofers are used to provide subtle support. According to the installation instructions, the work should sound “like a symphonic work in a good concert hall.” When the sound mixes were done, it was the studio’s intention that the nine digital files played at the same level would create a generally good representation. Because that is almost never true in practice, the artist prefers to have McDonald personally tune the work each time it is installed. The work delivers an immersive, intimate and yet communal listening experience.

B. Sonic parameters

In terms of sound sources (1a), there are nine channels of audio in a single acoustically treated exhibition space, and either nine or eighteen loudspeakers (depending on screen size). If space does not permit all nine screens to line the walls, seven can be on the walls with two in the center back to back. Other configurations are possible in collaboration with the artist and studio.

From a sonic standpoint (1b), the channel arrangement varies but is ultimately based on aural criteria. Two channels, Kjartan (bass and piano) and Doddi (drums) have subwoofers, so neither of them should be near an entrance. McDonald mentioned that the subwoofer is intended to provide a subtle support for the piece and would be too prominent if located at the entrance. Once inside the exhibition space, the low frequencies can be heard and felt, which help to envelope visitors in the music. At a few points in the piece, Ragnar, Gyda and Kristin Anna sing together in a kind of trio, and it is ideal to position this trio where they can be seen simultaneously. But because femininity is a central theme in the work, Ragnar does not want to be on any of the first screens visitors see when they enter. Having one of the women visible at the entry rather than the artist means that sometimes only two of the three members of this trio can be seen at once, which is acceptable. Because the drums can overpower delicate passages, that channel is not placed near Gyda or Kristin Anna. Ragnar’s is also a louder channel. The pianists can be on opposite sides of the gallery because the pianos are in the same room of the house, so both pianos can be heard in both audio channels.

In terms of behavior (1c), the speakers have a fixed position and are directional. There is a single speaker in the center above each screen. For larger screens, a second speaker may be added to the center bottom of the screen. The videos are experienced like a visitor going into and out of each room, and the positioning of the speakers reinforces that. As McDonald explained it, a speaker emits a cone of mono sound. In this work, that single speaker draws visitors in to face each performer at the center of the screen. The speaker is tilted down, so the tweeter is pointed at ear-level to the visitor. Visitors enter this cone of sound and eventually stand at its center. As they move away, they move out of that cone and into the next one, which envelopes them right in front of the next performer, which produces an effect akin to being guided through a series of rooms, but by sound. Incidentally, this is also how the horizontal spacing of the screens is done, by sensing where this transition occurs, and keeping the coverage across the entire screen. In the exhibition space, the voices seem to actually emanate from the performers themselves, which is the intention.

None of the equipment has aesthetic or conceptual significance (1d), and is chosen only for performance. There are no intentional sound obstructions or occlusions, and reflections and reverberation are undesirable and should be minimized (2c-e).

The ideal listener experience (3a) is self-directed. There is no preferred position to view the work, so there is no seating, and the visitor is encouraged to continuously move through the space.

Tuning (3b) takes great care and the channels are diverse in terms of dynamic range and frequency range. McDonald will spend five or six hours on this. The porch video has the sound of the outdoors, and the cricket sounds in this channel give the exhibition space the feeling of late afternoon in a country house in the summer. These cricket sounds should be heard across the gallery. According to McDonald, the artist always wants this channel louder. It also contains two cannon blasts. It is also where the piece ends—the performers stroll out of the house, off into the sunset together, but even though it is the climax of the piece in some ways, it should not be offset in importance in any way. No channel should be given more prominence than another. Because the porch video is loud, the channels near it cannot be too delicate, and it is bright, so it cannot be too close to other channels it might wash out with light spillover. McDonald noted that it is quite moving when visitors have to explore the gallery to discover where the performers disappear to at the end. One by one, the performers walk off-camera and the rooms become empty. Eventually the performers reappear together in the porch video and everyone in the gallery ends up in front of it. This creates a beautiful moment, and due to the volume of the video, it can still envelope everyone present. McDonald listens to the interplay between channels across the entire piece. There are certain “sections” where the character of the sound or music changes to something new, and each is tuned accordingly. He ensures that everything from room tone to the cannon blast is considered, so the visitor never loses their sense of being in this house. The visitor’s attention and movement are always responding to the performance as intended. In essence, when everything feels correct, the work feels spontaneous and natural, and the intentions behind the tuning process are inevident.

Additional/specific listener experiences (3c) are not specified. There are no preferred listener positions. The artist rather likes when people sit on the floor, move, sit again, and so on. It gives the gallery an intimate, bohemian feel that mirrors the feeling of the space the performers occupy. Seating, when provided, is to accommodate those visitors who may need a bench or seat to experience an hour-long work, but there is no conceptual importance to having seating in the space or placing it in certain locations.

Modes of listening (3d) are restricted. The nine loudspeakers in the room give the entire listening experience a multi-personal, communal feeling which in turn reinforces meanings in the work. A different mode of listening such as headphones would not be appropriate.

C. Media

Inventory (1). There are nine video files with mono sound, which include uncompressed master files as well as a set of compressed exhibition files. Either may be used for exhibition, depending on playback equipment available.

Production history (2). The production circumstances are visible in the videos themselves. Each performer has a single microphone and is recorded by a separate camera on a tripod. The performers wear headphones so they can hear each other during the performance. The technician can be seen turning the cameras on at the start of shooting and turning them off at the end. In post-production, a careful sound mix was done and every detail was painstakingly managed, down to the sounds of the flies. Flies can be seen at several points in the video, and it was determined to leave about five of them in the final sound mix.

Characterization (3). The master files at acquisition were ProRes 422 HD video files with 48 kHz, 16-bit PCM dual-channel mono audio, along with H.264 exhibition files with AAC audio. The files were made in the post-production house in Germany that color corrected the video, so the condition assessment (4) revealed no errors. The full technical characteristics of each file were noted in the condition report. However, to enable full resolution presentation of the audio in the future using increasingly available solid-state hard drive playback methods, the studio masters with 48 kHz, 24-bit audio should be used. Through the interview process the museum learned that it should request these files from the studio. Any BrightSign files used for exhibition that are now being provided by the studio are made with the 24-bit stems, but for solid-state hard drive playback, a new ProRes master with 24-bit PCM audio is needed. The quality difference to the playback of the sound between compressed and uncompressed is akin to that between standard definition and high definition video, but ability to discern the difference comes with developing the necessary perceptual skills with respect to sound.

D. Sound system

Specifications (1) for the speakers are Genelec 8030A loudspeakers and QSC K subwoofers, but comparable models are acceptable. The Genelecs specified are high-quality loudspeakers chosen for their coverage, detail and presence. The subwoofer is required, because without it, there are low frequencies in the audio file that would be lost entirely—the loudspeakers cannot physically create those frequencies. The specified media players are BrightSign XD1032, one per channel, with files prepared as BrightSign packages with AAC audio (a codec for compressed audio). Between the BrightSign player and speaker, there is a digital-to-audio converter with an equalizer and some gain settings that McDonald will set when he tunes the work onsite. He will leave the exhibition team with notes on what the settings are. As an alternative to playing compressed audio, solid-state drive playback on Blackmagic HyperDeck would enable the full resolution audio to be used. While tremendous attention is paid to equipment chosen for each installation, none of it is considered to be aesthetically or conceptually significant. It is chosen for performance, and other types and models delivering comparable performance may be substituted.

Cabling (2) varies based on equipment and cables appropriate to the space. Cables are chosen for their ability to effectively carry the signal from equipment closets to projectors and speakers in a large gallery.

Part II: Installation Documentation

Documentation of the acoustic environment (A) consisted of manufacturer specifications for carpet and acoustic treatments, as well as the dimensions and physical features of the gallery space. The size and placement of acoustic paneling was noted in floor plans and diagrams created during the documentation process. Note also that sound absorption is increased when the gallery is full of visitors, as it often was, with people moving freely through the space and sitting on the floor. Documentation of the channel layout and cabling (B1) was illustrated with diagrams and descriptions. The channel layout met the various criteria described in the interview with McDonald as to how the channels were arranged according to aural considerations (Figure 2).

Figure 2. Diagram of the Channel Layout of “The Visitors” at SFMOMA. Illustration by the author.

The Iteration Report provides a section to document the decision-making process for exhibition file preparation and the technical specifications of the files actually used in the exhibition, which in this case utilized BrightSign players and files prepared by McDonald.

The museum performed several types of live recording to document the installation (B2). SFMOMA created 360-degree documentation that was also edited and posted to the museum’s YouTube channel [https://www.youtube.com/watch?v=Igof7jUIaaE&feature=youtu.be], as well as live sound recording to document the typical visitor experience. Sound level meter data (B3) was not collected because the piece is not tuned according to such data, but in hindsight, the following might have been interesting to measure: dB SPL at the peak dB FS of each file at a fixed distance from the center of each screen, and approximate or average dB SPL at certain timecodes, such as the climaxes of the music, while moving through the gallery. However, such data may be misused in the future if not properly contextualized.

The author noted subjective experiences while in the installation (B4) and considered expanding this exploratory technique in the future by interviewing others and annotating a storyboard to associate their impressions with particular points in the work. Some key impressions were that, because the work loops and does not have set screening times, visitors may step into the work at any point and then must get their bearings. The persistent cricket sounds, while shifting from foreground to background depending on the music, heighten the impression of being in an open country house in the summer. The subwoofers can be felt as well as heard, and the first cannon blast is physically a shock if you do not see it coming. There is a communal sense of discovery throughout the experience of the work.

There were no iteration-specific modifications to the work in terms of sound (B5). For the installation at SFMOMA, McDonald personally tuned the work, and the artist visited as well, so the screen size, placement, acoustic treatments, and equipment all aligned with the artist’s specifications and installation instructions. As such, this installation was considered successful by the artist, the studio and the museum.

McDonald indicated that the process of channel layout determination and fine-tuning of “The Visitors” had never risen to the level of conscious method prior to the discussion for this case study. Through the interview and dialogue with a conservator, many of these aspects of the sonic character of the work surfaced and could be documented for the first time. Although these aspects will surely evolve, the unique value of this dialogue for the preservation and future presentation of “The Visitors” is evident.


For time-based media works that foreground sound, conservators must be conversant with sound physics, acoustics, audio engineering and sound design so that they can have productive conversations with artists and sound professionals in order to create meaningful documentation of the significant aural properties of an artwork for the future. Documentation of time-based media installations can be aided by instruments and technology but is primarily an intellectual exercise informed by a thorough understanding of the artwork and its needs, conservation ethics, and modern and contemporary art history and practice. In a recent interview, Jim Coddington, former Chief Conservator at MoMA, said of conservators, “we must recognize, indeed celebrate, that we are humanists, not scientists.” (Rivenc 2017, 23) As reports are drafted and supplemented with the interviews, images, data and recordings that will guide future decisions, it becomes clear that comprehensive documentation is not inherently comprehensible or useful. Coddington goes on to make the important point that documentation of lasting value is purposeful and ultimately applied to the treatment and presentation of artworks. As such, the selection of strategies for documenting a complex artwork merits careful consideration. This framework represents a step-wise method for documenting sonic identity and the significant aural properties in time-based media art installations that have pre-recorded digital audio. Taken section by section, it mirrors the process of arriving at an understanding of how the artwork should sound, and how it should be situated within an exhibition space—a unique acoustic environment with its own innate qualities.


This research would not have been possible without the participation of Christopher McDonald, Director of Sound for Ragnar Kjartansson; Martina Haidvogl, Associate Media Conservator, SFMOMA; and Joshua Churchill, Media Technician, SFMOMA, all of whom were extraordinarily generous with their time. The author’s graduate internship supervisors provided invaluable guidance that helped shape this research: Jim Coddington, Kate Lewis and Peter Oleksik, The David Booth Conservation Department, Museum of Modern Art (MoMA), NY; and Joanna Phillips, Solomon R. Guggenheim Museum, New York. The Andrew W. Mellon Foundation provided essential support through the Fellowship in Media Conservation at MoMA (2016-2017).


Cohen, Michael. “Dimensions of Spatial Sound and Interface Styles of Audio Augmented Reality: Whereware, Wearware, and Everyware.” In Fundamentals of Wearable Computers and Augmented Reality, Second Edition, edited by Woodrow Barfield. Boca Raton: Taylor & Francis Group, 2016, 277-308.

Hermann, Robert. “Recording the Sound of Installation Art Objects.” Guide produced under B3 Activity: Documentation of Movement and Sound led by Reinhard Bek, Museum Tinguely. Inside Installations: Preservation and Presentation of Installation Art project co-organised by Netherlands Institute for Cultural Heritage / ICN, Tate (UK), S.M.A.K. (Belgium), Restaurierungszentrum Düsseldorf, Foundation for the Conservation of Contemporary Art/ SBMK (The Netherlands) and Museo Nacional Centro de Arte Reina Sofia, Madrid. June 2004–August 2007. Accessed March 7, 2019. https://www.yumpu.com/en/document/read/7803633/recording-the-sound-of-installation-art-objects-inside-installations

The Museum of Modern Art (MoMA). “Pre-Acquisition Questions.” Resources for the workshop Getting Started: A Shared Responsibility—Caring for Time-Based Media Artworks in Collections at The Museum of Modern Art, New York, June 11-15, 2018. Accessed March 7, 2019. https://www.mediaconservation.io/s/_MoMA_Pre-Acquisition_Questionsdocx.pdf

Phillips, Joanna. “Shifting Equipment Significance in Time-Based Media Art.” In The Electronic Media Review, Volume One, edited by Jeffrey Warda, Fletcher Durant, and Gwynne Ryan. Washington DC: Electronic Media Group, American Institute for Conservation of Historic and Artistic Works, 2012, 139-54. Accessed May 31, 2018. http://resources.culturalheritage.org/emg-review/wp-content/uploads/sites/15/2016/07/Vol-1_2010_Ch-6_Phillips.pdf

Phillips, Joanna. “Reporting Iterations: A Documentation Model for Time-Based Media Art.” In Revista De História Da Arte, 4 (2015): 168–177. Accessed September 25, 2017. http://revistaharte.fcsh.unl.pt/rhaw4/RHAw4.pdf

Rivenc, Rachel. “Interview: Jim Coddington, formerly/retired Agnes Gund Chief Conservator at the Museum of Modern Art.” Triennium 2014-2017 Newsletter 3, Modern Materials and Contemporary Art working group, International Council of Museums Committee for Conservation (ICOM-CC), 21-24.

Wharton, Glenn. “Reconfiguring Contemporary Art in the Museum.” In Erma Hermens (ed.) Authenticity in Transition: Changing Practices in Art Making and Conservation. London: Archetype Publications, 2016, 27-36.


David, Gary D. and Ralph Jones. The Sound Reinforcement Handbook. Milwaukee: Yamaha, Hal Leonard, 1989.

Kelly, Caleb. Sound in Art. London: Whitechapel Gallery; Cambridge, Mass.: MIT Press, 2011.


Amy Brost
Assistant Media Conservator (2017-present)
Andrew W. Mellon Fellow in Media Conservation (2016-2017)
The Museum of Modern Art, New York