Request for Starting TTS

Function

After establishing a WebSocket connection with the TTS engine, the client can send a start TTS request to initiate TTS. If the client sends multiple synthesis requests over the same WebSocket connection, the WebSocket connection must be re-established for each request, as one connection can handle only a single synthesis request at a time.

Request Parameters

**Table 1** Parameter descriptions
Parameter	Type	Mandatory	Description
command	String	Yes	Set it to START to initiate the recognition request.
text	String	Yes	Text to be synthesized, which can contain up to 10,000 characters.
config	Object	No	Configuration information. For details, see Table 2.

**Table 2** config data structure
Parameter	Type	Mandatory	Description
audio_format	String	No	Audio format header. Option: pcm. Default value: pcm.
sample_rate	String	No	Sampling rate. Options: 16000 and 8000. Default value: 8000
property	String	Yes	For details, see Table 3. For speakers with high-quality pronunciation, the charge is made every 50 characters.
subtitle	String	No	Whether to generate timestamp information. Leave this parameter blank if not used. Range: word_level: text-level timestamp. phoneme_level: phoneme-level timestamp.

**Table 3** Range of property for speakers with high-quality pronunciation
Parameter	Value	Type	Use Case	Sampling Rate (Hz)	Audio Format
Ahmed	arabic_dh_male	Virtual human	Arabic	8k/16k	pcm
Aisha	arabic_dh_female	Virtual human	Arabic	8k/16k	pcm
Ahmed	english_dh_male	Virtual human	English	8k/16k	pcm
Aisha	english_dh_female	Virtual human	English	8k/16k	pcm

Example

{ 
    "command": "START",
    "text": "Nice to meet you.",
    "config": 
    { 
        "audio_format": "pcm", 
        "sample_rate": "16000", 
        "property": "english_dh_female"
    }
}