If we remove the quietest of components from a frequency analysis, leaving only a percentage of the loudest or softest components, then we are left with a delicate interplay of sine tones. This does rather beautiful things to water and voice sounds. Here is an example:
Percentage of partials remaining: Indicates how many of the loudest or softest components are to remain to form the new sound-file.
Loudest / Softest popup: Indicates whether the loudest partials are to remain (the most common option), or whether the softest partials are to remain (unpredictable).