Basic strategy: using a flexible, controllable signal, which has a range going from constant '1' ( 1 sample = no signal alteration) to noise (= infinite alteration; of course within techn. limitations), as the transforming cross-filter signal, in order to create and control as many possible sonical/musical effects of the cross-filter as possible. This is a real-time implementation, so that one can be in control of the many musical possibilities of the cross-filter, make smooth transitions between them etc., in real-time. signal flow
Cross-filtering the various timbres of one instrument with each other, in real-time. Variation is created simply by a delay and a time modulation algorithm (before cross-filtering the signal). Especially with new, exotic instruments like for example the knifetable this can lead to astonishing results. signal flow
Cross-filtering two voices, in this case an adult voice with a child voice - notice that the length of the response file (the child voice) first gradually increases before it decreases again towards the end.
some examples of real-time implementations and spin offs of Trevor Wisharts waveset technique in kyma (programming language):
Waveset timestretching (sample domain), by Camille Troillard - signal flow
a patch that features the old Reichian way of phasing, as well as a new approach to phasing, made possible by new technical possibiltities which come with a program-ming language like kyma. This new approach of phasing is based on a simple oscillator-crossfade-technique. In the music example it is audible as a kind of gradual shift, almost morph, in the accent patterns, which in this case are equal to the so-called result patterns. signal flow
This can be used as an alternative technique to create frequency and amplitude fluctuations/deviations. Furthermore, by using multiple noise delay modules, the illusion of a chaotic sound-mass can be established.
Example 1 (frequency and amplitude fluctuations/deviations) - signal flow