End-to-End Amp Modeling: From Data to Controllable Guitar Amplifier Models

TL;DR

Addresses the problem of reproducing a guitar amplifier's full control-range response with a data-driven, end-to-end neural model. It introduces an automated data-collection pipeline using a robotics setup to vary knob positions, play audio, and record outputs. It trains a conditioned LSTM-32 network with an ESR loss and compares it to an offline LTSpice SPICE model, showing comparable subjective quality in listening tests. Practically, the approach enables real-time, non-intrusive, data-driven emulation of boutique amplifiers across multiple controls.

Abstract

This paper describes a data-driven approach to creating real-time neural network models of guitar amplifiers, recreating the amplifiers' sonic response to arbitrary inputs at the full range of controls present on the physical device. While the focus on the paper is on the data collection pipeline, we demonstrate the effectiveness of this conditioned black-box approach by training an LSTM model to the task, and comparing its performance to an offline white-box SPICE circuit simulation. Our listening test results demonstrate that the neural amplifier modeling approach can match the subjective performance of a high-quality SPICE model, all while using an automated, non-intrusive data collection process, and an end-to-end trainable, real-time feasible neural network model.

Figures (6)

• Figure 1: Neural network amplifier model training scheme. A model $f$ receives control positions $\mathbf{c}$ as conditioning, and maps an input signal $\mathbf{x}$ to resemble a target output signal $\mathbf{y}$. A loss function $L$ scores the similarity of model output $\hat{\mathbf{y}}$ and $\mathbf{y}$ and provides a learning signal to adjust the model parameters to improve the resemblance.
• Figure 2: Data collection robot connected to an amplifier tina_patent_eutina_patent_us.
• Figure 3: Example pathfinding solution for the case of two knobs and 500 data points. The data points are shown as dots, and the line segments between them show the travelled path for a) random order, and b) sorted order using a TSP solution.
• Figure 4: Comparison of model output (blue) to reference (black) in time domain (left) and frequency domain (right).
• Figure 5: Output from the Matchless DC-30 model with various tone cut settings in time domain (left), and in frequency domain (right).