Frequency discrimination of stylized synthetic vowels with two formants.