Aristech Open Issues

[pragmatrix]

From #35

• Determine whether the credentials authentication is likely to be necessary in future or whether we can reliably just use apiKey
• Is there a silence timeout and if so, is it configurable? Does the silence_timeout field in EndpointSpec have any effect?
• When using the example, if the default microphone settings are used ( as opposed to explicitly using 16kHz) does the conversion function which is currently used get in the way? (audio::into_i16) i.e. does not using it improve the performance of the example?

Additionally:

• Is caching of the voices needed in aristech-synthesize?
• Billing?