Skip to main content
. Author manuscript; available in PMC: 2012 Nov 27.
Published in final edited form as: J Acoust Soc Am. 1989 Aug;86(2):566–581. doi: 10.1121/1.398236

TABLE VI.

Phonemes accounting for the highest percentage of error in final position.

System Phonemes
(Percentage of error accounted for in each evaluation)
Natural t
(18.18)
d
(18.18)
θ
(18.18)
m
(18.18)
b
(09.09)
Prose 2000 V3.0 k
(38.46)
θ
(26.92)
m
(07.69)
z
(06.41)
v
(06.41)
DECtalk Paul k
(20.22)
n
(19.10)
ŋ
(15.73)
m
(10.11)
p
(08.99)
DECtalk Betty n
(16.11)
ŋ
(15.43)
θ
(14.09)
p
(12.75)
m
(08.72)
MITalk-79 n
(27.54)
ŋ
(19.76)
θ
(11.98)
m
(11.38)
f
(07.19)
Amiga 500 v
(16.75)
k
(16.23)
t
(14.14)
p
(13.09)
ŋ
(09.42)
Infovox v
(13.38)
f
(12.64)
m
(09.67)
d
(08.92)
n
(08.18)
Smoothtalker d
(21.37)
p
(12.88)
g
(11.47)
b
(10.26)
k
(08.65)
Votrax p
(18.32)
m
(15.59)
k
(13.12)
θ
(09.90)
n
(09.16)
TSI-Proto 1 n
(34.38)
t
(26.07)
m
(12.13)
ŋ
(06.29)
p
(04.49)
Echo k
(15.20)
p
(13.64)
d
(12.23)
n
(09.72)
b
(08.62)