Problem with module_tts_output() function

[Luke Yelavich]

On Tue, Jul 11, 2017 at 06:14:05PM AEST, payman shaykhmehdi wrote:
> Hi,
> 
> I am writing an output module for speech-dispatcher. The module works fine
> in initialization, loading, synchronization, etc. But the problem is with
> playing the generated samples. I'm using  module_tts_putput() utility
> function for this purpose, But the resulted audio is very noisy. The point
> is that if the samples be saved in a wave file, that wave file is fine and
> there is no noise in the samples.

Is the noisy audio intelegable at all? In other words, can the speech be 
understood even through the noise?

> I have assigned the AudioTrack fields according to  the parameters of my
> synthesize engine. Is there any specific settings/configurations that I
> have missed and causes this noisy output?

One thing that immediately comes to mind is the endianness of the audio. Make 
sure you are correctly setting the audio endianness, as this has an effect on 
the playback of the audio, as well as the volume manipulation that Speech 
Dispatcher performs based on the desired volume setting.

Luke
-- 
Please check out my Patreon campaign and spread the word.
https://patreon.com/lukefoss