AUDIO ARTIFACTS IS AN EXPLORATION into the world of audio visualization. By tying an audio input and its quantified data to coded conditionals and geometric shapes, I was able to generate artwork that created permanent footprints for otherwise ephemeral events. This project set out to create generative visualizations of peers saying their name. By capturing high resolution screenshots of spikes in a person’s voice and layering them into graphic compositions, each individual creates a visual translation of their name – each with its own personality.
Many variables are taken into effect when generating the graphics. For one, each audio clip is split into three seperate frequency thresholds which quantifies low, medium, and high frequencies into a value between 0 and 1 with accuracy to eight decimal places. With these three values, an RGB value is created – with lows tied to reds, mids tied to greens, and high frequencies tied to blues. By doing this, one can imagine how different types of voices can create colors that reflect the frequencies of each individual.
These variables, as well as other audio data such as volume peak and gain, are also tied to other code that rotates, scales, and distorts shapes and lines to create truly unique graphic translations for each voice.
One could imagine multiple real-world uses for an interactive piece such as this. By using live audio to dictate its movement rather than preset animation, the piece could be utilized for anything from a live music backdrop at a concert to creating audio driven experiences for the hearing impaired.
Try the interactive experience for yourself and download the file here.