For example, 1164 has this in the .lab file:
0.253154 121 /
0.253154 121 t_c
0.277975 121 t_b
0.314003 121 t
0.320234 121 t_1
0.355538 121 @
0.381497 121 @_1
0.459441 121 r
0.488449 121 r_1
0.571300 121 Q
0.662899 121 b_c
0.673225 121 b
0.712735 121 D_v
0.742110 121 D
0.811800 121 @
0.970952 121 m
1.092500 121 /
but this in the .pho file
; MBROLA output from /usr/prosynth/pgc104/xmlfiles/jh1164.xml
_ 100
t 67
3: 61 0 115 12 117 25 119 37 120 50 119 62 118 75 116 87 114 99 114
r 106 0 120 25 119 50 121 75 128 99 132
O: 82 0 132 12 133 25 133 37 134 50 134 62 134 75 132 87 128 99 124
b 101 0 128 25 114 50 103 75 100 99 102
D 68 0 102 25 106 50 113 75 116 99 112
3: 69 0 112 12 108 25 103 37 99 50 97 62 96 75 96 87 95 99 94
m 159 0 95 25 88 50 87 75 82 99 76
_ 100
I've therefore changed the two cases of [3:] to [@] and [O:] to [Q] in the
files to be used for testing.
I mention this just in case anyone else has overlooked it.
Paul
_ Paul Carter __________________________________________ pgc104@york.ac.uk _
Dept of Language & Linguistic Science | http://www-users.york.ac.uk/~pgc104/
University of York | telephone: +44 (0)1904 432660
Heslington, York. YO10 5DD | fax: +44 (0)1904 432673