CAK: Emergent Audio Effects from Minimal Deep Learning

TL;DR

This paper introduces CAK and AuGAN, novel techniques enabling a minimal deep learning model to produce emergent audio effects and discover unique transformations from just 200 samples, highlighting new possibilities in effect design.

Contribution

The work presents a minimal 3x3 convolutional kernel framework with a new conditioning mechanism and a redefined adversarial training approach for emergent audio effects from limited data.

Findings

Emergent audio effects achieved with minimal data and simple kernels.

Frequency-dependent transformations discovered through learned kernels.

Adversarial training used to verify control application rather than generate forgeries.

Abstract

We demonstrate that a single 3x3 convolutional kernel can produce emergent audio effects when trained on 200 samples from a personalized corpus. We achieve this through two key techniques: (1) Conditioning Aware Kernels (CAK), where output = input + (learned_pattern x control), with a soft-gate mechanism supporting identity preservation at zero control; and (2) AuGAN (Audit GAN), which reframes adversarial training from "is this real?" to "did you apply the requested value?" Rather than learning to generate or detect forgeries, our networks cooperate to verify control application, discovering unique transformations. The learned kernel exhibits a diagonal structure creating frequency-dependent temporal shifts that are capable of producing musical effects based on input characteristics. Our results show the potential of adversarial training to discover audio transformations from minimal data, enabling new approaches to effect design.