5.4.3 Soft Templates based on the Voicing

Hard Template is sometimes too restricted to find regular rhythm pattern words so that we use a rather loose template, the Soft Template that is categorized according to the presence or absence of voicing (V, U) as described in Table 5.1. In Japanese, a vowel followed by a voiced consonant in the same syllable is generally shorter than an equivalent vowel followed by an unvoiced consonant of the same place of articulation. Moreover, vowel duration is more influenced by the preceding consonant than the following. At any rate, it is clear that Japanese vowel sounds are much affected by the voicing of preceding and following consonants. Typical examples of the Soft Template for two, three, and four mora words are shown in Figure 5.13, Figure 5.12, and Figure 5.14, respectively. Although it categorizes consonants into only two groups, they also show a stable standard deviation tex2html_wrap_inline4064 .

   figure1738
図 5.12: Rhythm patterns by Soft Template of 3-mora word

   figure1744
図 5.13: Rhythm patterns by Soft Template of 2-mora word

   figure1749
図 5.14: Rhythm patterns by Soft Template of 4-mora word


next up previous contents
Next: 5.5 Experiments Up: 5.4 Rhythm Pattern Templates Previous: 5.4.2 Hard Templates based

Jo Chul-Ho
Wed Oct 13 17:59:27 JST 1999