With more training data that covers more word instances, you have higher possibility to reduce the DSAT (dis-satisfied part of the speech, for example, the glitches) ratio for the voice. Here are some ...