The SampleFrequency
parameter specifies the sampling frequency of the audio (the frames produced by the module).
If you specify this parameter for the langid
and lidoptimizer
modules, IDOL Speech Server checks the classifier set sample frequency against the module sample frequency to look for mismatches. Setting SampleFrequency
to 0
(the default) disables this check.
When you have enabled the FFmpeg library, you can use a variety of audio formats with different sampling frequencies in the wav
and audio
modules. In these cases, the audio
and wav
module converts the audio to the specified sampling frequency.
In the audio module, this process applies only to file or binary data input types, not to streams.
To detect breaking glass using the audiosec
module, HPE strongly recommends that the audio sampling rate is 16000 Hz.
Type: | Integer |
Default: | 16000 (afp , afpfeature , audiopreproc , audiosec , audiotemplatedevel , audiotemplatescore , audiotemplatetrain , frontend , ivfile , ivscore , normalizer , and splitspeech modules)0 ( langid and lidoptimizer modules) (No default value for the audio , speakerid , stream , and wav modules) |
Range: | 8000 or 16000 |
Required: | Only for the speakerid module |
Configuration Section: | afp, afpfeature, audio, audiopreproc, audiosec, audiotemplatedevel, audiotemplatescore, audiotemplatetrain, frontend, ivfile, ivscore, langid, lidoptimizer, normalizer, speakerid, splitspeech, stream, and wav modules |
Example: | SampleFrequency=16000
|
See Also: | Sfreq (action parameter) |
|