Sonic fingerprinting:

In the following, a more in-depth description of the underlying technical concepts behind Queer Sonic Fingerprint is provided. For an even more detailed step-by-step explanation of sonic fingerprinting and evolutionary computing, see technical explanation below.

Every physical object responds in a certain way to excitation. We are familiar with this phenomenon from the difference between the reverb of a small room compared to that of a cathedral. In this project, the term sonic fingerprint is used to describe the unique acoustic properties of individual objects. Extracting the sonic fingerprint of objects involves either exciting them with a speaker and recording the result, or making a 3D-scan of them. The following two photos illustrate the process of recording the fingerprint of a wooden elephant and a steel thermos.

Using various mathematical operations, the fingerprint can be extracted and stored as a digital file. This allows for the use of the fingerprint as a filter on sounds unrelated to the object from which the fingerprint derived. For example, the fingerprint of a wooden object can be used to filter the sound of wind blowing, thereby creating a representation of what wind would sound like through the object. Because the fingerprints are stored as digital data, they allow for novel manipulations, such as merging two very different fingerprints or changing the individual components of the fingerprint's frequency spectrum, to allow for very subtle changes in the resulting sonic output. As an example, here we use the combined sonic fingerprints of the elepant and thermos pictured above to filter ambient sounds from a backyard. First the original audio:

And then the filtered version:

Queer sonic evolution:

This paves the way for using so-called evolutionary algorithms to generate new fingerprints. An evolutionary algorithm is a highly simplified model of biological evolution that can generate surprisingly complex behaviour through the use of a few simple mathematical models of genetic operators.

The fundamental idea behind QSF is that this novel approach to the morphology of sound is very suited for addressing queer and non-normative material relationships around museum collections. By extracting the sonic fingerprints of objects, either in the collection itself, or from storage, we bypass the visual identity of these objects for instead to focus on their acoustic properties. Approaching the sonic fingerprints as digital strings of DNA allows for the creation of new relationships through evolutionary processes such as recombination and mutation.

Where many evolutionary algorithms in the past have been biased by normative and neo-darwinist ideas around biological evolution and "survival of the fittest", our goal is to explore queer notions of kinship in which the problematic notion of "fitness" is replaced by relation and diversity.

Installation:

The result will be presented as an interactive audio-visual installation in which the objects themselves have been substituted by placeholders, for example archival boxes distributed throughout the exhibition space. Each archival box is equipped with a small speaker emitting the sound of a fingerprint. Using placeholders rather than the actual objects allows for audience interaction with the sonic ecology of the fingerprints: visitors to the exhibition can move and displace the boxes, each of which contains motion sensors. changing their spatial relations which in turn informs the evolution of the individual fingerprint. circumvent the exlusionary not only are handling concerns avoided; the placeholders can also signify contested property relationships of the objects, or even their absence from the collection, for example due to restitution. Through interaction with the objects, new fingerprints will evolve, changing the ongoing sonic identity of the installation. Yet, while the changes may be triggered by the audience, the detail of the sonic output will originate in the relations between, and the autonomous processes of, the fingerprints, and hence the objects, themselves.

QSF questions the norms and biases inscribed in objects by our visual culture. Throughout the lifetime of an object, its identity evolves through the interaction of social norms, biases and material decay. The sonic fingerprints evolve on a much shorter timescale and develop non-normative and queer traits through their digital autonomy. As such, the project will contribute to engaging audiences in contemporary museum spaces.

Queer Sonic Fingerprint: Technical explanation

Below follows an explanation of the process of creating and modifying sonic fingerprints.

In the following example, we use a so-called sine wave sweep as our mode of excitation. The sweep spans the frequency range between 20 Hz and 20 Khz and sounds like this:

A visualisation of the sweep looks like this:

This is what we call an impulse response , that is, a description of how a sound evolves over time. In this case we're dealing with an idealisation of the impulse response, since any physical propagation of the sweep, will colour its sound, and hence, its impulse response

We now play the sine wave sweep through a wooden elephant. The sound of the sweep being filtered by the material properties of the elephant sounds like this:

Notice the difference between the above visualisation of the impulse response of the original sweep and a plot of the impulse response generated by the sweep filtered through the elephant:

Now we play the ortiginal sine sweep through a double-walled steel thermos bottle:

Visualisation of its impulse response:

Through a mathematical operation known as the Fast Fourier Transform (FFT), we can now take the impulse responses of the two objects and turn them into frequency responses . As suggested by its name, the frequency response, rather than describing a sound's temporal character, describes its frequency content. That is, some frequencies will be more pronounced than others, which is exactly what defines a sonic fingerprint. That transformation is visualised below (impulse responses to the left, frequency responses to the right):

For the next step, we need a source sound. This can be anything, but here I am using the sound of snow melting in my backyard:

Since we can now represent sounds as frequency responses, through a process known as convolution we can superimpose the frequency responses, or sonic fingerprint, of the wooden elephant on the source audio. After this operation, we can perform and inverse Fourier transform, to turn the convolved frequency response into an impulse response that we can listen to:

Here we're using the thermos fingerprint and convolving it with the source audio:

Now we'll combine the fingerprints of the elephant and the thermos. We simply slice the frequency responses in two at an arbitrary point and combine the first slice of the elephant response with the second slice of the thermos response, and vice versa. The left column below shows the original frequency responses and the arbitrarily chosen slice point, while the right column illustrates the recombined frequency responses.

We now have two new responses with which we can convolve the source audio, giving us the following sounds:

What if we imagined these frequency responses as chromosomes? We could then conceive of the frequencies as genes, which, during the phase of chromosome recombination, would carry over certain sonic fingerprints of their parents. But a few genes would also sometimes undergo mutations that would give the new generations their own sonic identity. As such, what was originally fingerprints derived from actual physical objects, now beget new fingeprints that occupy a speculative sonic space. In this space new sonic fingerprints suggest a continually evolving population of virtual material configurations.

The next stage of the Queer Sonic Fingerprint project involves exploring the discipline of evolutionary algorithms (EA) to create diverse and dynamic sonic textures through the evolution of sonic fingerprints. Yet where many EAs in the past have been biased by normative and neo-darwinist ideas around biological evolution and "survival of the fittest", our goal is to explore queer notions of kinship in which the problematic notion of "fitness" is replaced by relation and diversity.