Author: Alexandre Porres
Download full paper: Media:Dissonance Model Toolbox in Pure Data.pdf
This paper discusses the implementation of Psychoacoustic Models of Dissonance and its creative applications. Terhardt’s theory of dissonance perception is a departure point , it involves the perceptual attributes of Sharpness, Roughness, Tonalness, Root Relationship & Pitch Commonality. This is part of a PhD research that aims to the development of computational tools for interaction in real-time (live electronics: composition, performance or improvisation) based on a Dissonance Model.
The main problem for this dissonance theory becoming more available and usable in composition or Live Electronics is that the field of research is in constant research and much debate. There is not a final word on how each parameter contributes to the perception of Dissonance and there are still other concepts that lay out the division of Terhardt. A review of the theory is then part of the research.
Some perceptual tests are also being undertaken on the current stage of research. They shall provide data to be compared to the results from the implemented models. The application of this theory in creative computer tools for real time is still very incipient. The contribution provided is that the theory and discussion on its application becomes clearer and more available to composers. Most of the usage of this theory has been applied in computer aided composition, but not much has been done for real time applications in Live Electronics.
The model of sharpness is readily available as a Low Level Descriptor, which are more common, and have been gaining accessibility for composers. One creative possibility that is provides is to match sounds with similar characteristics, concatenative synthesis and other similar processes. Dissonance Models for real time applications can work in the same fashion, and expand this process. So the Dissonance Model can also work in conjunction with other Low Level Descriptors, and other creative possibilities shall arise by the pace this research becomes more available to creative musicians.
One model of Roughness is proposed here and has been previously exposed as well as some creative applications in PdCon07. Further applications will now be presented that also involve other tools such as the Phase Vocoder and a patch that arbitrarily changes the partials of a spectrum. This is in line with the technique of Spectral Mapping of Sethares.
This research hasn’t found the entire Pitch Commonality model of Parncutt  implemented for real time applications, and this has been done on this stage of research. The implementation of the model is described as well as some modifications that expand its capabilities are presented.
4th international Pure Data Convention 2011 Weimar ~ Berlin