Abstract:

It is described a program for translating graphical and MIDI scores into digital audio such that orchestral instruments figuring in these scores are replaced by mathematical ones which are previously selected according to their timbral similitude after a human hearing criterium. These mathematical instruments are found in algorithmic orchestras which are computed by composition systems based on time-trees. In this way, a new sound rendering with good similarity to the original score is produced by a SOM-A chart having its note parameters coming from the corresponding MIDI file, but having its orchestra coming from an additive- instrument bank.