Speech interface for ISY

[rusgrafx]

I'm trying to find a way to add voice announcements / TTS (text-to-speech) option to ISY99i.

The problem is - I don't really want to have an always-on PC/Mac running just for that. The whole idea of switching to ISY was to get rid of power-hungry PC system. But I can't see any other way of doing it.

You know, I'm looking at my son's talking and singing toys, musical door bells, etc. and I think that somewhere there might be some kind of device or embedded system that would allow me to do it. I studied Digi-Key catalog and searched on the web, but I can't find any solution as of yet. May be someone here can come up with a good idea?

The requirements are simple:
-low cost (under $100)
-quiet (no fans)
-very low power consumption (at least in sleep mode, which would be 99% of the time);
-should be IP network connected;
-should have mono or stereo line-level output;
-should be able to receive a string of text and voice it.

ISY supports Networking Module that can send IP requests to external devices, so I was hoping to use it somehow for that purpose.

At the very least it might be OK to have a set of pre-recorded messages (MP3, WAV) stored somewhere (like on networked HDD) that could be retrieved and played on some other device (aka DLNA or Apple's AirPlay).

Any ideas?

 

[Lagerhead]

I am also looking for a TTS appliance with a simple serial- or IP-type interface.

Dan was doing some work in this area late last year, AFAIK hasn't followed-up (last post was rather cryptic).
http://www.cocoontech.com/forums/index.php?showtopic=17657

 

[rusgrafx]

Dan was doing some work in this area late last year, AFAIK hasn't followed-up (last post was rather cryptic). http://www.cocoontech.com/forums/index.php?showtopic=17657

Yes, I remember that thread. It would be nice to know that someone is looking into this. Dan, any updates you can share with us?

Just for the reference, the Wikipedia lists several TTS chips under Dedicated Hardware section. Now all we need is an embedded system engineer who can make sense of it and design a product for general public. Here you have a ready-to-bake business idea to take to investors.

I was even thinking of re-purposing some of the talking dictionaries or ZYGO devices, but that as well requires some system engineering knowledge that I unfortunately do not have.

 

[damage]

you can use a seagate dockstar for something like this. low power, easy to hack, runs linux so you can install freely available TTS packages. just add a $1 usb soundcard from ebay. you can build on something like xpl-perl http://www.xpl-perl.org.uk/ to add network controllability. a few of us have used the dockstars for various projects: http://www.cocoontech.com/forums/index.php?showtopic=12745

 

[electron]

I kind of lost track of this project (so much going on), but will have to start this back up as it does look promising.

There are many free/cheap TTS solutions, but most of them sound really bad. IMO, if it doesn't sound as good as the AT&T/Neospeech voices, then I personally won't waste any time on it.

 

[rusgrafx]

...just add a $1 usb soundcard from ebay.

Damage, thanks, you gave me some food for thought.

I just realized that I already have an always-on HP Media Smart server running on my network. It lacks audio out, but an external sound card might be just what I need. This server already handles more duties than it was originally designed for (like SIP/Google Voice gateway, PlayOn server, etc.), so adding TTS should not be too much trouble for it (unless TTS doesn't install in Windows Server environment).

I'll give it a spin, I just need to figure out the control part. I heard a lot about xPL Perl, but haven't tried it yet. Thanks again.