Appearance
Voice Cloning
With the speechify-1
engine, it is possible to clone custom user voices.
It can currently be done via the <speechify:clone> SSML tag - This method involves using a specific SSML tag tailored for the speechify-1
engine.
Using the <speechify:clone> SSML tag
To clone a voice using an audio prompt, you must provide a source
attribute. The source
attribute is required and can be in the form of either a Data URI (e.g. data:audio/mp3;base64,yourbase64
) or a direct link (e.g. gs://some-bucket/your-file
or https://example.com/prompt.mp3
).
For examples and detailed information, refer to the SSML documentation.
TIP
Ensure the audio prompt you provide a high quality ~5-10s prompt for optimal voice cloning results.