Talk-to-Speech Improvments

[ncborges]

I've done a request sometime ago for some improvments on TTS engine, and the answer was: It could be possible.
wish to ask Dimitri if are you taking consideration on teh request?

Now playing announcement, example to follow:
<?xml version="1.0" encoding="utf-8"?>
<Table>
  <Settings>
    <Tracks2Speak>3</Tracks2Speak>
  </Settings>
  <Message>
    <Voice>0</Voice>
    <tts>It's %now-hour and %now-minute minutes now, and we are listening %artist, %title.</tts>
  </Message>
  <Message>
    <Voice>1</Voice>
    <tts>You're listen the Brazilian Number one station on the web! That's great! Now, %artist, %title!</tts>
  </Message>
  <Message>
    <Voice>0</Voice>
    <tts>Another great song from %artist, right here on WRTTS, stay tuned for more great music!</tts>
  </Message>
  </Table>


On this example <Tracks2Speak> means how many musics must be played to have the announcment.
<Voice>0</Voice> choose the voices you have installed to announnce the desired text or song. WIndows always have the first voice as zero, then second voice will be one, third voice 2, etc

The oportunity t choose between the voices and the automatic rotation of messsages will be a lot appreciated.
Thnaks so much and congrats. TTS is working great on the latest version. Still having 120 tries,

 

[djsoft]

If those are standard Windows Speech API tags, then those are already supported in RadioBOSS.

 

[ncborges]

No they are not standards. As well, at the monet, you have no option to secelct the voice you want speak your txt if you do not change it manunally in teh Speech session of Control Panel. The idea of this XML is change the voices by the text command <voice>1</voice>. For fully atomation when you cannot be 24 hours on the Playout server. At the moment, if you use a .txt as an event, the TTS will announce the latest music played and not the music that text is overlaping the introduction. At the moment there is no way you put a TTS as event and it announces the music is coming, even it is running over the introduction, will announce the latest palyed song if it has passed the time scheduled.

 

[djsoft]

According to Microsoft Speech API docs: https://msdn.microsoft.com/en-us/library/ms717077(v=vs.85).aspx - you can set the voice requirements from XML tags. RadioBOSS supports all those tags, so it should work.

 

[ncborges]

Let's give a go and test then. Thanks for while.

 

[djsoft]

Let's give a go and test then. Thanks for while.

Let me know how it goes. Please note there's a bug in RB 5.4.3.1, sometimes it doesn't read files when using TTS (this will be fixed in the next update).