Differences between an untrained and trained singing voice lie in one or more of breathing technique, larynx settings or vocal tract settings. Measurement techniques in each of these areas are important to provide data on the singing voice, and accurate data are essential for natural personalized electronic voice synthesis in the future.