| record | -length length -lsilence leading_silence -tsilence trailing_silence -stdev stdev -backoff backoff |
| length | Specifies the time-length of the recording in seconds. |
| lsilence | Specifies the amount of leading silence in milli-seconds. |
| tsilence | Specifies the amount of trailing silence in milli-seconds. |
| stdev | Adjusts the sensitivity of utterance detection levels. |
| backoff | Adjusts how much of the prior signal (in milli-seconds) is included once an utterance has been detected. |
NOTE: For the default settings of or for more information about the above options see the CSLUsh manual section, ``Removing leading and trailing silence,'' or the respective index listings in this manual.
record records the utterance, creates a wave object in memory that contains the utterance, and returns a handle to that wave object.
Returns a handle the wave object containing the recorded speech.
Italicized words are indexed in the CSLUsh manual, and can be referenced there for more information.
set recordedWave [record -length 10] # Record 10 seconds of speech,
# with default values for all other
# ``record'' options, and
# store it in a wave object named
# ``recordedWave''.
( ...use the recording as desired)
Nuke $recordedWave # Remove the wave object from memory when
# it is no longer needed ( \ldots neglecting
# to do this step can cause memory leaks).
NOTE: Nuke is listed in both this Appendix and in the
CSLUsh manual's Reference section.