Discrimination of Musical Instrument Tones Resynthesized with Piecewise-Linear Approximated Harmonic Amplitude Envelopes

Harmonic amplitude envelopes of eight sustained musical instrument tones were approximated using piecewise-linear segments to investigate how the discrimination of approximated tones varies with the number of breakpoints and instruments. Using a two-alternative forced-choice paradigm, listeners judged whether approximated tones sounded different than originals. Apart from basic statistical analysis, various error measurements were evaluated based on the results of the listening tests. Relative-amplitude spectral error and relative-amplitude criticalband error were found to be the best error metrics for predicting discrimination, accounting for about 81-82% of the discrimination variance. Strong correlations were measured between discrimination scores and modified spectral incoherence based on the three strongest harmonics, with breath noise in the flute and bow noise in the violin appearing to cause increased discrimination between original and piecewise-linear-approximated sounds.