“…Sundberg ͑1982͒ states that steady-state sung vowels retain a degree of intelligibility up to 520 Hz ͑or even 1000 Hz in consonant-vowel-consonant ͑CVC͒ context͒. The problem of vowel perception at high F 0 has been recognized by many authors ͑Carlson et al., 1975;Carré and Lancia, 1975;Traunmüller, 1981Traunmüller, , 1990Bladon, 1982;Ryalls and Lieberman, 1982;Klatt, 1982;Darwin and Gardner, 1985;Gottfried and Chew, 1986;Assmann and Neary, 1987;Beddor and Hawkins, 1990;Perec, 1991;Assmann, 1991;Hirahara and Kato, 1992;Hirahara, 1993;Rosner and Pickering, 1994;Hirahara et al, 1996͒. Two qualifications must be made. The first is that in natural speech vowels are rarely sustained.…”