Although the overlap-add based methods are widely used because they have a relatively low computational complexity, these techniques make no assumption of the nature of the signal they are applied to and in general apply uniform scaling rate to all segments of speech with out regarding if they are voiced, unvoiced or silence segments