Text this: Revisiting the speaker discriminatory power of vowel formant frequencies under a likelihood ratio-based paradigm: The case of mismatched speaking styles.