From: Analysis of Salient Feature Jitter in the Cochlea for Objective Prediction of Temporally Localized Distortion in Synthesized Speech