Real Time Emulation of Parametric Guitar Tube Amplifier With Long Short Term Memory Neural Network

Numerous audio systems for musicians are expensive and bulky. Therefore, it could be advantageous to model them and to replace them by computer emulation. In guitar players’ world, audio systems could have a desirable nonlinear behavior (distortion effects). It is thus difficult to find a simple model to emulate them in real time. Volterra series model and its subclass are usual ways to model nonlinear systems. Unfortunately, these systems are difficult to identify in an analytic way. In this paper we propose to take advantage of the new progress made in neural networks to emulate them in real time. We show that an accurate emulation can be reached with less than 1% of root mean square error between the signal coming from a tube amplifier and the output of the neural network. Moreover, the research has been extended to model the Gain parameter of the amplifier.

💡 Research Summary

The paper addresses the long‑standing challenge of reproducing the characteristic nonlinear distortion of guitar tube amplifiers in a digital, real‑time environment. Traditional approaches such as Volterra series and its subclasses can theoretically model nonlinear systems with memory, but they quickly become impractical because the number of kernels grows combinatorially with the order of nonlinearity and the length of the memory. Identifying these kernels analytically is cumbersome, and the resulting models often demand prohibitive computational resources for real‑time use.

To overcome these limitations, the authors propose a data‑driven solution based on Long Short‑Term Memory (LSTM) neural networks. LSTM units are specifically designed to capture long‑range temporal dependencies through gated mechanisms (input, forget, and output gates), making them well suited for systems where the output depends not only on the current input but also on a history of past inputs—a hallmark of tube amplifier circuits that exhibit hysteresis, bias drift, and harmonic buildup.

Data acquisition and preprocessing
A classic parametric tube amplifier (12AX7 pre‑amp stage, 6L6 power stage) was set up in a controlled lab environment. The authors generated a diverse training set by feeding the amplifier with a mixture of sine sweeps, chord progressions, and pseudo‑random noise, all sampled at 48 kHz with 24‑bit resolution. Simultaneously, the input signal and the corresponding output from the amplifier were recorded using a high‑precision audio interface. The raw waveforms were normalized to the