Sound Synthesis
The sound synthesis is a whole of techniques allowing the generation of aural signals.
It is employed in several fields:
- At the musical level , it makes it possible to create new sound objects. In this context, the goal is not to reproduce existing sounds but rather to invent the new ones.
- On the level of the Telecommunications, it makes it possible to reduce the quantity of information during the transmission of an audio message: this one is then described by its parameters of synthesis which are the only transmitted data.
- On the level of the Virtual reality and the video games, the sound synthesis makes it possible to increase the feeling of presence of the listener-actor by managing the interactions between the actor and his sound environment. The actor acts in a direct or indirect way on the parameters of synthesis.
Various forms of synthesis
The whole of the sound syntheses can be classified in 4 categories:
-
syntheses by abstract algorithms . This category comprises the syntheses based on the use of various algorithms for the generation of the sound objects. It includes/understands the synthesis by frequency modulation (Synthèse FM), the synthesis by amplitude modulation (MY, ringmod), the Synthèse by nonlinear distortion (waveshaping), the Synthèse by distortion of phase, the synthesis by modulation of the width of impulse (PWM) etc the advantage of the syntheses by abstract algorithms lies in the weak costs of calculations and the low number of parameter of synthesis. By changing some parameters of synthesis, it is possible to obtain very different sonorities. The major disadvantage lies in the not very realistic aspect of generated sonorities (compared to the physical sounds of origins).
-
syntheses by reading of samples . This category comprises the syntheses based on the modification of sound recordings. Historically, the first form of synthesis by modification of sound recordings dates from the experiments of the Concrete music (Pierre Schaeffer): magnetic soundtracks were cut out then D assembled. Today this category includes/understands the sound Synthèse by sampling (sampling), the granular Synthèse, etc the Turntablism also forms part of this category. The advantage of this type of synthesis lies in the wide one and the richness of sonorities which it is possible to generate.
-
syntheses by models of signals . This category comprises the syntheses based on the description of the sound characteristics of sound sources perceived by the listener. It comprises the additive sound Synthèse, the subtractive sound Synthèse etc This category has the disadvantage of generating a great number of parameters of syntheses. However the algorithms of the treatment of the signal make it possible automatically to obtain the parameters of syntheses starting from recorded sounds (algorithms of analysis resynthèse).
-
syntheses by physical models . This category comprises the syntheses based on the description of the physical behavior of the sound sources. It includes/understands the Synthèse by guide of wave, the modal Synthèse, the Synthèse by solutions of equations physical, the Synthèse by assembly of elementary physical elements (mass-arises from the system Cordis Anima). The advantage of this category of synthesis lies in the direct link between the parameters of synthesis and the physics of the sound source. However the implementation of physical models is often heavy in calculations.
Tools
The first synthetizers combined seldom more than one type of sound syntheses (DX7 for FM, TB303 for the subtractive synthesis, etc). Currently, a great majority of the sound syntheses is accessible via software called modular synthetizers. This software makes it possible to combine freely various types of synthesis. The most used are: Max/MSP, Pure Dated, Synthedit, Reaktor, Absynth