Fig. 14From: Convolutional neural networks for the classification of guitar effects and extraction of the parameter settings of single and multi-guitar effects from instrument mixesMean absolute error of the parameter extraction of the convolutional neural network across pitch of keyboard and bass for all time-frequency representations investigated. Only the two E tones were part of the training. Shown are, respectively, the extraction errors for the gain, tone, depth, and frequency parameter of the distortion and tremolo effect at −12 dB mixing volume (a, c, e, g) and at −3 dB mixing volume (b, d, f, h)Back to article page