On the Road: Familistère de Guise

The Location

What happens when you combine an industrialist, a political theorist and a social activist? You get someone like Jean-Baptiste André Godin. A French manufacturer of cast-iron cookers and stoves, Godin was also greatly influenced by socialist/utopian thinker Charles Fourier, which led him to a radical solution for workers and their families: the Familistère de Guise, (The Social Palace at Guise) in northern France. Built as a communal space, it was intended to be a self-contained locale for work, play and culture, and represented a state-of-the-art social experiment for its time.

Unfortunately, several of the buildings were damaged during World Wars 1 and 2, after which both the factory and the community moved into private hands. While the living quarters have become apartments and condominiums, the Social Palace has also been restored as a historical monument/museum which offers some unique opportunities for artists.

About This Project

Since the beginning of the restoration, the founders of the museum have had the desire to provide contemporary music in the different spaces the building offers. The architects and artists who were selected by the founders for the restoration imagined a Cour Sonore (sound courtyard), a sound system that would be entirely hidden inside the structure of the building itself. The structure could be opened to composers and sound artists to create works in which no equipment or technology would distract the visitor from a clear vision of the architecture. Sound artists Jean-Christophe Desnoux and Manuel Poletti have created a gigantic meta-instrument within these constraints in The Social Palace, assisted by some clever fellow sound engineers.

The museum’s curatorial staff set several interesting requirements for the project, which were then discussed with and developed by the artists:

• The system should play music and sound every day without the need for manual control.

• The museum staff should be able to select any of the sound material at any time to play for visitors using a remote control.

• Some composed musical sequences constructed from custom recordings related to historical themes surrounding the Familistère should be played regularly throughout the day. These themes would include work (machines or manual activities), persons (individuals or crowds), and countryside soundscapes. A year-long set of original recordings was put together by Jean-Christophe Desnoux in order to assemble the sound collection the Familistère now has on site, with more than thirty thematic sequences composed for the project.

• Every day at specific times (such as the end of morning or afternoon work time), some original music - mostly composed by Jean-Christophe - should be played.

• In between each thematic or music sequence, a generative system should create some constantly renewing sound sequences from the original recordings. These sequences combine the different kinds of sounds and play all day long - interspersed with silences between each sequence - to create a sort of “sound fountain”.

• The system should be able to host external sound sources - from live musicians or audio sources for a guest dance performance, for example.

• The system should be able to host future works from other sound artists and composers, allowing a for the creation of a unique and growing contemporary music repertoire which would provide the basis of a form of nightly electro-acoustic concerts.

• The system should take advantage of this particular space and provide interesting audio spatialization effects, of course.

In 2009, 58 speakers (providing a 52.6 horizontal audio system!) were installed in the ground of the main courtyard of the Central Pavilion. The speaker system is interconnected by four kilometers of cables that connected the speakers up to a central control location that includes 32 amplifiers, 3 soundboards and one main computer. The entire software system was written in Max.

While Jean-Christophe was the original creator of the sound courtyard, Manuel Poletti joined him in the project as a Max programmer and audio spatialization specialist, creating both the software architecture and control. The required features took a total of three months to program.

Since its opening in 2010, the Max patch has played every single day, and has never crashed, while the staff still love their improvising Sound Courtyard - Even after seven years of daily and almost continuous sound, the staff of the Familistère de Guise still love their improvising Sound Courtyard!

Sound Spatialization

The music software projects multichannel sound into the courtyard - a total of 16 audio sources may be spatialized over 52 audio channels. Sound distribution of each of the 16 sources is achieved using a simple audio matrix along with additional reverb effects.

The audio routing of a sound source can be set using a graphical matrix that represents each speaker - either by hand, by using some dedicated tools like IRCAM’s Spat, or even a custom granular spacing model. These three models of audio spatialization may be applied to a particular sound source:

1. Free: you may just “place" a sound in the space (i. e. a speaker) by selecting one or several sliders in the matrix and setting the volume.

2. DBAP (distance based amplitude panning): This model uses an algorithm which is implemented in the IRCAM Spat Max objects collection. You can place a sound source virtually in the space using a graphical interface, and the DBAP model calculates the energy of each output (that is, each speaker) according to the distance between a speaker and the virtual source. Since every speaker is involved in this calculation, it results in a smooth distribution of the sound over the global space of the sound scene. You can then automate the virtual source using cyclic or random path patterns. Virtual speaker positions were calculated using a map of the real positions of the speaker output positions in the space.

3. RSPAT (random spatialization): The audio matrix of one sound source is automated using a custom semi-random algorithm that allows for sound output in one of many modes. These modes are best imagined by the nicknames that were given to some of the presets of spatial effects this tool may create: rain, granular, smooth air, travelator, earthquake, or thunder.

The natural reverb of the Pavilion Central is close to that of a cathedral - the ground is covered with tiling, and the ceiling is glass - with a reverb time of almost 7 seconds. An additional reverb effect is used to blend the dry sound from the speakers and the huge resonance of the real space. This virtual reverb is the one used by IRCAM’s Spat - a smooth, high-quality room effect that you may shuffle over any speaker setup (in the present case, two 52-out multichannel reverbs can be used: one that is short, and one that is very long).

Six subwoofers were added to the system and hidden in the walls of the courtyard to produce low-frequency audio. Distances between subwoofers in the space are considerable in the space, and bass frequencies are in general difficult to localize in space but this system required the spatialization the bass frequencies. A second output matrix is used to re-spatialize the 52 dacs into 6 low-frequency sound points using the same Spat DBAP algorithm mentioned earlier. If you move a sound source in the space, separated low frequencies will follow the movement and distribute the bass sound accordingly within the space.

The Future

As a part of the original design, this system - similar to an ancient grand organ model - is meant to become a central instrument for a group of composers who may feel inspired by the unique historical and architectural space found in the Familistère of Guise.

The staff is now planning to open a hotel in the Familistère, and they hope to produce some special evening events of electro-acoustic or live concerts to be experienced in the calm and emptiness of the space of the Cour Sonore. This initiative will take place across a 3-year period, with guest composers invited and “produced” by Manuel Poletti to create new pieces. Composer Cécile Le Prado was the first invitee, and her beautiful 20 minute-long voice-based piece premiered in June 2017.