PING

| summary | goals | process | technical | credits |

PROJECT SUMMARY

PING is a collaborative online sound experiment which seeks to invoke this latency as a means of modelling the 'resonant' characteristics of a multi-site audio transmission network. The event will embrace the impossibility of true simultaneity in transmitted communication, and twist the effect of this phenomenon from undesirable interference into an expansion of creative means.

By shattering the commonly-held desire for simultaneity, what is gained is the opportunity to explore the resonances of multi-site performance spaces in the same way as a singer might wander through the space of a new hall in order to exploit its specific acoustic behaviour - in this case, however, the space presents the possibilities of a complex reflection which is directly tied to the existence of each listening point, and the incorporation of other artifacts (which might otherwise be labelled a loss in quality) from the process of transmission.

For more information visit the PING website or email Alex Geddie.

The final transmission of the experiment will blend live digital sound from Le Fresnoy, reflections of this sound received from the listening stations (which then re-enter the loop of communication), recorded musical sounds which have been pre-processed using impulse and response data from the first phases of this experiment to imitate the past resonances of the space, this same data in its raw form, and live impulses and responses which record the resonance of the performance space as it exists during the performance (encoding by extension the possibility of the future usage of this data).

PROCESS

In computer networking, to 'ping' a location is to determine whether it is reachable by sending one or a series of echo requests to a host location and 'listening' for replies. The term is borrowed from the jargon of sonar, the system long used aboard submarines to detect threats by actually listening for reflections from short bursts of sound (the aptly-named 'pings') actually diffused through the surrounding water.

The term 'latency,' when used in the context of a signal-transmission network refers to the delay between the time of a signal's entry into the network and the time of its reception. Generally considered undesirable for the purposes of communication, this delay is inherent in transmission, and is a characteristic which is independent from the data-transfer rate of a network.

The experiment consists of two preparatory components, and a final performance component.

PHASE 1
IMPULSES // LATENCY MEASUREMENT >> online, multi-site
The first phase will require the participation of as many sites as possible, hopefully three in addition to Le Fresnoy. Each site will establish its own outgoing audio stream, and 'tune in' to the streams of all incoming streams from the other sites - the incoming streams will be fed into a mixer, so that each site may control to what degree each incoming signal becomes part of the retransmitted signal. During the first phase, each site will be asked to manipulate the mix at their listening station for the purpose of gathering useful data for the next phase.

This component will consist of exploring all the possible global mixing permutations. The input for each test during this phase will be a single, short impulse. Given that the network forms an endless loop, each test will be concluded when the incoming signal from all the sites become a dense feedback. The complete mix received at the 'central' site will be recorded for use in the analysis phase.

PHASE 2
RESPONSES // CONVOLUTION & COMPOSITION >> offline, ROUBAIX
Convolution is a broad term used in musical digital signal processing to describe the modification of a signal. It can be used to describe the characteristics of almost any type of modification performed on the sound such as filtering, and most notably, reverberation. The measurement requires the comparison of the impulse to a file called the 'response' file. For example, with a good recording of a single impulse sounding inside a large basketball court, this response file could then be used to process any sound to make it seem to have sounded in the same setting.

A new set of sounds will be generated by using the response information gathered in phase one for inclusion in the final performance. The impulse files used during this generation of new material for the final performance will consist of a wider variety of digital musical sounds, each of which will make its own unique use of the range of collected response data. The generation of new sound by this convolution method is extremely processor-intensive and may not occur in real-time.

Additionally, through the analysis of the response data, an overall sketch of the shape of the final performance will take form, so that a plan for the movement of mixing settings at each of the sites involved in the final performance may be created. A loosely-defined composition will take form.

PHASE 3
FINAL MELTDOWN // WAS IT LIVE OR WILL IT MEMOREX? >> online, multi-site
The final event of the experiment will be the performance, where all the response files from the previous phases, freshly processed sound files, and live sounds, impulses and reflections will be combined according to the composition. The role of each player/listener is to be aware of aural cues (which will be discussed before the performance) and instructions received via chat software for the manipulation of the mixing controls. If it is not possible equipment-wise for the site to tune in to the final mix emanating from Le Fresnoy during the performance (or keep this mix separate from the re-transmitted signal), the performance will be re-broadcast immediately after its completion.

Hardware: 2 machines (Mac/PC)
Software: streaming software
Access: preferably high speed
Installation requirements: mini-mini cable.
Supported platforms: Mac/PC

PROJECT CREDITS

Artist: Alex Geddie

Le Fresnoy: National Studio of Contemporary Arts (Lille, FRANCE)

Ryerson University (Toronto, CANADA)