Text to speech (idea)

---

Hello Friends,

i think about a text to speech plugin for pimatic. I found this npm plugin here:

https://www.npmjs.com/package/say

I think it’s possible to implement this in a rule:
IF $mailbox.status == 1 THEN say “The mailbox are filled”

If this possible?

@Frank-Herrmann
Just to clarify: You mean outputting the voice directly on the device pimatic is running at?

@Frank-Herrmann I like the idea of add ing a text to speech module, but we should make a broader review of available sollutions. What I don’t like about “say”: There is no README.

@Oitzu said:

@Frank-Herrmann
Just to clarify: You mean outputting the voice directly on the device pimatic is running at?

I mean to speak over a ALSA device, today i read an article about that:
https://www.pubnub.com/blog/text-to-speech-audio-broadcast-with-raspberry-pi/

i like the idea of text to speech output too.
but wouldn’t it be better to get the speech output on the device where you are operating the gui actually (as oitzu was asking above) ? if possible somehow ? my pi is located at the hall, so the speech option would only work either with longer cable or maybe some kind of bluetooth speakers especially for the pi usage in my case.

similar issue with the sound output has the mpd-player plugin in my opinion or any other plugin using sound outputs.

@leader21 said:

but wouldn’t it be better to get the speech output on the device where you are operating the gui actually

Agree, this is an important point.

@leader21 said:

but wouldn’t it be better to get the speech output on the device where you are operating the gui actually (as oitzu was asking above) ?

But how to output sound if there is no pimatic client?
Maybe ‘store’ the sound note till some client (or special user) is available?

is homekit and siri an option?

Please check this link, this wil generate a soundfile with spoked text. We can transfer to client or use the browser play audio method to sound out at a client.

http://www.js2node.com/tts/text-to-speech-for-nodejs

and here you can test a tts engine:
http://www.voicerss.org/api/demo.aspx

This thread is a littlebit old but i do this:

install pico2wave via packages.

Create file speech.sh put in this:

#!/bin/bash
say() {
pico2wave --lang=de-DE --wave=/tmp/test.wav "$*"; aplay /tmp/test.wav; rm /tmp/test.wav
}
say $* > /dev/null 2>&1

After this you can do via execute command with a rule like this:

WHEN LVRMoveDetect reports present and $SomeOneAtHome = 0 THEN execute "/home/pi/speech.sh Hallo, Sie haben den Alarm ausgelöst. Polizei ist alamiert."

There are many scenarios that you can do now like a window is open to long. Or put out the weather if change etc etc etc.

Via aplay, you can play also sound files.

Now you have a secure dog wake up all the time and the postman will runaway if you have a big loudspeaker on door.

I use my Sonos to do this with the node-sonos-http-api. It’s pretty awesome

please look at: https://forum.pimatic.org/topic/4233/pimatic-tts-text-to-speech-released

See also https://forum.pimatic.org/topic/3124/woli-control. It has a TTS notification feature for Android devices.

@heizelmann would you care to elaborate? It’s really motivating for developers when people using their software free of charge also contribute to the community by explaining what they would like to have as a free feature