Sony ICD-UX81 User Manual

Page of 112
ICD-UX71/UX81/UX71F/UX81F 4-114-023-11(1)
GB
U
tili
zin
g Y
ou
r Co
m
pu
te
r
Creating a user file adapted 
to your recorded speech 
(Training)
What is “Training”?
Before you can transcribe 
recorded messages using Dragon 
NaturallySpeaking, you must create a 
user file specifically adapted to the sound 
of your recorded speech. This procedure 
is called “training.” 
When training is complete, Dragon 
NaturallySpeaking creates a user speech 
file containing information about your 
recorded speech. 
Refer to the operating instructions 
or the help supplied with Dragon 
NaturallySpeaking, as some parts of 
training are the same as those required 
for dictating directly to the computer 
using a headset microphone. 
Even if you have already created a user 
speech file with a headset microphone, 
you should create a separate speech 
file for the IC recorder to get better 
recognition accuracy. 
Tips on using the IC recorder
When transcribing messages recorded 
using the IC recorder, the following 
tips will help you obtain maximum 
recognition accuracy. 
Set “REC MODE” to “ST,” “STSP,” 
“STLP” or “SP” in the menu. 
When creating a user speech file 
(training), set “REC MODE” to the mode 
you normally use for voice recognition 
software. 
Set “MIC SENSITIVITY” to “LOW (L)” 
in the menu. 
Hold the IC recorder firmly and 
start recording. 
Hold the IC recorder so that the 
microphone is about an inch (2 - 3 cm) 
from the corner of your mouth, not 
directly in front of it. This minimizes the 
sound of your breath in the recordings, 
enabling maximum recognition 
accuracy. 
Note
Avoid recording in a noisy place, such as on a 
subway platform, in order to obtain maximum 
recognition accuracy.