A system for generating voice source signals that implements the transformed LF-model parameter control

Abstract

This paper describes a system which fully implements the transformed LF glottal flow model, incorporating also the often-overlooked k-factors. A problem with the original proposal is that the global waveshape parameter Rd, central to the transformed model and used to predict some of the other parameters, cannot be directly controlled. Instead, a stylisation of the model was used to indirectly control Rd. However, this approach can yield substantial errors in the Rd value of the pulse, thus undermining the usefulness of the model. To overcome this problem, an iterative algorithm is presented, which ensures that for a given Rd input value, a pulse with that Rd value will be produced. Using this new Rd control, it transpired that some of the original parameter predictions give rise to combinations of values that are incompatible with the LF model. Modifications to the original predictions of the Rk parameter were incorporated to ensure model conformity.