Acoustic Guitar Effects Series

Description

An attempt at an emulated parlour acoustic guitar sound, the NAM file is a result of passing the response WAV file from Tone3000 through an offline modeller I have written using Python, Numpy, Numba, et al, to process the sound and add in sound characteristics of an acoustic guitar. The simulation attempts to capture some of the following characteristics of a recorded acoustic guitar: * Room reverb * Bass bloom for lower strings/ open chords * Finger noise simulation induced during slides/ chord changes * Guitar body size (changes dynamics/ EQ) Due to the modeller being offline and reading WAV files it can analyse audio tracks to check, for example, changes in frequencies over time (randomly samples up to 32 sections of the WAV track) that would be associated with chord or note changes that might introduce classic finger slide noises in the heavier acoustic guitar strings. I am fairly certain a lot of these characteristics are not picked up very well by the NAM neural capture, however it is an interesting experiment to see how far the neural modelling can be pushed. In due course, if there is a enough interest, I will link to the Gitlab repository with my models/ simulators. Works best without pre-amp boost on your audio interface, and with a reverb plugin.

Makes and Models

Tags