- Installation
- Administration
- Programmer's Guide
- Grammars
- MRCP Server
- FAQs
The Speech Engine accepts audio only in certain formats. The sound formats listed below are members of an enum SOUND_FORMAT whch is used as an argument in loading the voice channel (LoadVoiceChannel or LV_SRE_LoadVoiceChannel) for decodes . The streamed audio format is set be setting the STREAM_PARM_SOUND_FORMAT to one of the below formats using StreamSetParameter or LV_SRE_StreamSetParameter.
ULAW_8KHZ
Format: mu-law
Sample Rate: 8 kHz
Bytes per Sample: 1 byte per sample
Memory: ~0.5 MB / min.
Notes: This is the standard domestic telephone format.
PCM_8KHZ
Format: PCM
Sample Rate: 8 kHz
Bytes per Sample: 2 bytes per sample
Memory: ~1 MB / min
PCM_8KHZ
Format: PCM
Sample Rate: 16 kHz
Bytes per Sample: 2 bytes per sample
Memory: ~2 MB / min
ALAW_8KHZ
Format: a-law
Sample Rate: 8 kHz
Bytes per Sample: 1 bytes per sample
Memory: ~0.5 MB / min
SPX_8KHZ
Status: Deprecated and non-functional since 9.0
SPX_16KHZ
Status: Deprecated and non-functional since 9.0