Dreaming in 1’s and 0’s

AI is the Bridge Between Art and Science

The use of Artificial intelligence within sound sourcing

AI shows us the way potentially to the future of sound design, from which we feed the algorithm a picture or scene, and the AI interprets that and produces sounds based upon the input.

For this area I explored the difference between two images fed to an AI. These can be seen below (and yes, one of them is just the colour white).

The thought process here was that often when creating a sound, or looking for sound sources, we already have the end product in mind with limitations already designed. By using these two images I fed a website (melobytes.com) the information, and it produced two files both rich with random sounds. The advantage of this is that the AI focuses on the limitations within its algorithm, leaving more creative expression with the material that is produced.

The beauty of being fed random sounds was that I could allow myself to creatively explore how I could transform the sound. First, I used a Paul stretch to really expand upon the length and content to really delve deeper into the makeup of the sound. This was followed by the use of a reverse to really help make the sound unrecognisable from the source. Next, I applied a large reverb and set the signal to only apply the wet signal, leaving me with just a phantom of the former sound. Creatively, by this point, I had decided that with the Paul stretch I wanted to create something almost ethereal in nature to evoke emotions of endless space and time. The final step was to normalise the volume of what was now just the wet reverb signal to that of -1db. The steps used can be shown below.

Below are more examples of me exploring the world of AI generative audio. With no end media in mind I experimented with the same sound source to create many different sounds.