Recriação do timbre de guitarra usado nas gravações da banda Aurora Dark através da técnica de modelagem de perfil no software NAM (neural amp modeler)

Abstract

This Final Paper focuses on the attempt to recreate the different guitar tones used by the band Aurora Dark during the recordings of their first three songs, through profile modeling comparing it with conventional methods, which involve use of microphone and amplifier, present in the original recording. This experience report aims to detail the tone recreation process using neural network modeling, highlighting the benefits and limitations inherent to this emerging technology. The analysis includes the sound quality, versatility and usability of the results obtained. The underlying purpose is to make pertinent information available to musicians, producers and researchers interested in incorporating models generated through neural networks into their musical productions and mainly, in the creation of guitar tones. By exploring the consequences of this new approach, we seek to understand its impact on the big picture of musical production. This study not only documents the procedure adopted, but also serves as a guide for those interested in adopting this approach in the context of their practices, providing relevant information about the possibilities and challenges that permeate the use of neural networks in the search for the excellence of the sound.