Beschreibung:
<jats:p>Is it possible to synthesize the French horn well enough to pass for the real thing? Most commercial synthesizers and soundcards are limited to producing timbres that are ‘‘horn-like’’ or ‘‘clarinet-like.’’ It is much more difficult for a model to be flexible enough to allow different types of articulations, phrasings, and tone colors. This work presents a hybrid additive-wavetable synthesis model for the horn using genetic algorithm optimization. The parameters for the model include fundamental frequency, overall amplitude, vibrato amount, attack and release times, brightness, and articulation shape. The model gets brighter as the amplitude increases, and the brightness decreases as the amplitude decreases by scaling each partial with a set of exponentially related amplitude envelopes. The model is implemented in the Csound software synthesis language. After implementing the model, several solos and orchestral excerpts that are commonly required in auditions were synthesized. The model is particularly effective on Baroque and classical excerpts.</jats:p>