Less Redundancy amongst speech cues in the output of a channel vocoder than in the speech signal at the input of the vocoder