Abstract | In music production, it is common practice to apply dynamic-range compression to audio signals. Traditionally, the operator’s attention is drawn to the reduction in dynamic range and the sonic signature imposed by the envelope (settings) of the device, and the resulting distortions are familiar to studio practitioners. However, the non-linear characteristics of compression, combined with the interaction of these signals once summed, are likely to produce less familiar side effects, such as intermodulation distortion, manifesting itself as signal masking and other related artefacts. Comparative quantitative analysis of compressed simple and compound signal structures shows the products of this distortion to be realignment of harmonic structure, reduction of spectral and temporal clarity, and rearrangement of dynamic variances related to the rhythmic structure of musical signals. Although the rearrangement of the dynamic variances is expected (in that the variances are reduced), what is less expected is the extent to which amplitudes of certain individual components of summed signals are attenuated, effectively precipitating signal masking. This research shows that decreasing the number of signals interacting with each other whilst applying an equivalent amount of compression can reduce the intermodulation distortion and therefore improve the overall signal quality of commercial music. |
---|