Neural networks have become ubiquitous with guitar distortion effects modelling in recent years. Despite their ability to yield perceptually convincing models, they are susceptible to frequency aliasing especially when driven by high frequency and high gain inputs. Nonlinear activation functions that create harmonic distortion also cause aliasing as the bandwidth of the signal is expanded beyond the Nyquist frequency. Here, we present a method for reducing aliasing in a pre-trained model via a teacher-student approach, where the teacher is the pre-trained model with its weights frozen, and the student is a copy of this with learnable parameters. The student is fine-tuned against an aliasing-free dataset generated by passing sinusoids through the original model and removing non-harmonic components from the output spectra. Our results show that this method significantly suppresses aliasing for both long-short-term-memory networks (LSTM) and temporal convolutional networks (TCN). In the majority of our case studies, the reduction in aliasing was greater than that achieved by two times oversampling. One side-effect of the proposed method is that harmonic distortion components are also affected. This effect was found to model-dependent, with the LSTM models giving the best balance between anti-aliasing and preserving the perceived similarity to an analog reference device.
