View Full Version : Overlaying current track with text to speech

2009-03-21, 03:01

I enjoy my SB for years now. I have a sort of intelligent house which is able to produce text2speech currently broadcasted over a pair of speaker connected to the machine were squezzecenter is running on. Dont get confused. The current text2speech uses only the same machine and has nothing to do with slim at all.

I would like to use my SB setup to broadcast the text2speech in the future. In this regard a question. The text2speech engine I am using is able to produce a MP3 file.

Is it possible to "interrupt" the running title on the SB, play a text2speech MP3 file and than resume/unpause the music to play from the position it was before the text2speech message?

A even better solution would be to "overlay" the music with the text2speech engine like in the car when "paula" (the voice of my navigation system) says something and the sound volume of the music is a bit reduced. You got it?

Any hints?


Mick Seymour
2009-03-22, 02:02
Interesting idea. I presume the one in your house is called "Proteus".

What do you use it for?

2009-03-23, 12:15
if you are able to have it as an mp3 file, then you could write (or have someone write) a plug in that puts this mp3 file in your playlist after the current track (add next)...

I don't exactly know how SC streams, but if the file name and location remain the same, you could potentially have the TTS mp3 every alternating track, while a track is playing, the TTS mp3 is overwritten with the new text to speech data (i.e. new song announcement...)?

2009-03-24, 06:33
Would be very helpful for blind people as well.
We would need to have the track piped through an external app which would perform the mixing.
ecasound under linux would be able to do it in batch mode I guess.
What sort of speech engine do you use?

2009-03-24, 07:58
You say you have speakers connected to your machine. Are you running SqueezeSlave on the machine or something? I do that with VortexBox, run SqueezeSlave locally and use that as part of my multizone Audio. In that case you can just play the music through a soft mixer attached to the sound card. You don't need to do anything to SC or SrueezeSlave. This will not work if your using a hardware player such as a SqueezeBox.

2009-04-18, 12:32
Guys, tnx for you feedback. Was on a business trip and hat honestly no time to read your replies. Appologies.

@jaysung: The text2speech engine I use is swift. It has several voices in different languages. You can test it for free. It produces some nag text bevor your text is spoken. If you invest <$15 this nag text/speech goes away.

@agillis: I dont use softsqueeze on the machine as it is in the living room together with a squeeze box. Does not really make sense.

@Mick Seymour: No the machine is called appolon (the greek god of the light ;).

@matthijskoopmans: I dont want it to be played at the end of the song. it should "overlay".

2009-09-12, 07:25
* pull up *

Has anyone an idea how we can "overlay" some mp3 in a running stream?

2009-09-12, 09:10

Once you have the file and you can make your app launch an application\command line you could use what is written in this thread:


I don't know if it still works but can't think that it doesnt.


BTW it doesn't overlay but immediately plays the file you send to it.