Discontinuity Removal in Concatenative Synthesized Speech

Concatenative synthesis concatenates segments of prerecorded natural human speech. It requires database of previously recorded human speech covering all the possible segments to be synthesised. Segment might be phoneme, syllable, word, phrase, or any combination. Concatenative speech synthesis is currently the most practical method for the generation of realistic speech. There mainly two types of issues in concatenative synthesized speech. These are spectral discontinuity and joining cost of concatenative synthesized sound. When the join between two speech units is clearly audible, it refers to discontinuity. The synthesized speech can sound very natural if the discontinuities at the concatenation points are inaudible. But when these joins are audible, their presence can be very frustrating to the listener and it also reduces the overall perceived quality of synthesized speech. The join cost in unit selection is used to represent the compatibility of two consecutive units. This may also determined as the weighted sum of subcosts such as: Differences in F0 and amplitude. Mismatch in different spectral parameters, like MFCC (Mel Frequency Cepstral Coefficients), LPC (Linear Predictive Coding Coefficients). Detection of spectral mismatch at joining points is done by following some steps. The knowledge of WaveFile Format is required. Waveform Audio File Format (WAVE, or more commonly known as WAV due to its filename extension) is a Microsoft and IBM audio file format standard for storing an audio bitstream on PCs.