This paper describes an investigation of the comparative performances of formant and channel vocoders as expressed by segmental level intelligibility properties. There are a number of ways in which the performance of speech processing systems may be assessed from the standpoint of the human observer. This study examines basic segmental intelligibility patterns as a measure of the effectiveness of vocoder system in encoding phonological information in the acoustic signal with sufficient robustness to allow it to be decoded with minimal assistance from high level linguistic context.