Working in the Front Line Massed violins & digital sound

Sidebar: Massed violins & digital sound
Why should the sound of massed violins should be so susceptible to processing damage? Given that all links in the chain are non-linear to some degree, either in terms of transfer function or in terms of delayed energy storage, the potential for audible damage to a given musical signal is proportional to its complexity. The intermodulation products for just a few tones look horrifying on a spectrum analyzer display. Imagine a musician sounding two notes on his violin: high-order harmonics are predominant, with perhaps 15 significant components. Then add the full orchestral complement of up to 20 players. Each will have a fractionally different tuning; in any case, their sounds are subject to differential delays en route to the microphone. Add in the contribution of floor reflection, a further delay. Add in the cumulative sound in the reverberant field.

The resulting massed string sound might well possess in excess of a thousand significant, recordable harmonic constituents. The build-up of multiple cross and intermodulation energy of an enharmonic nature is clear enough, and may well explain the very evident distortion heard on reproduced strings in the audible range. It is highly significant that the direct feed from a recording microphone to a good pair of monitor amplifiers and speakers shows very much less of this familiar audio distortion.--Martin Colloms