I’m going startup a radioshow / podcast type of thing, and I have a bit of a strange request. My voice sounds pretty dumb on record, so I want to use a robot voice type of thing. I could put my own voice through an effect or vocoder, but I’d rather do a speak and spell type of thing. Anyone know a website where I can type in some text and have a Steven Hawking type of voice speak it back? Perferably it would have a few different voices, because I’d rather avoid the cliched SH voice.
BTW, I’ll be using the voice to introduce the show, and between songs, etc.
http://www.research.att.com/projects/tts/demo.html?
I think Windows comes with a synthesizer as well.
Do you know where I’d find that?
Control Panel > Speech, type in what you want it to say and hit Preview. Or if you want it to read something longer, put it in a Notepad window or something and Start Menu > Accessories > Accessibility > Narrator and highlight your text. But mine only came with one voice installed and it does sound like the cliched Stephen Hawking voice and personally I think the voices on that site do a much better job of sounding, well, human.
If you can get hold of a Mac, they had a couple of voices that straddled the line between Hawking/drunken Norwegian and rather life-like. It’s possible you could have a well-modulated but slightly robotic voice. I don’t know what’s available under OS X, though – I’ve got an antiquated system.
To follow up on Daithi Lacha’s comment: actually there’s now a great many voices (screenshot) for the Mac. (And here’s where a couple of them come from.) “Bad News” sounds kind of like Marvin!
There are many, many text-to-speech (TTS) voices available. There are dozens of companies that sell them, and many that are free. Microsoft has quite a few free ones in different languages as part of their SAPI (Speech API) development kit. As well, many companies offer online demos, such as AT&T’s Naturally Speaking, given above. Interestingly, AT&T does not sell that product anymore, choosing instead to use it for their own products.
[hijack] ScanSoft, a big player in the industry, has been on a buying spree over the last few years, culminating in the purchase of Nuance (their major competitor in the telephony sector). After the “merger,” the new company chose to go with the Nuance name. I suspect that between the products they own, and the ones they license from other companies, they hold the majority of the speech recognition and TTS market share. [/hijack]
You might also try out Loquendo, who have some high quality TTS voices as well.
Speakonia is a free text-to-speech program for Windows that will record directly to a .wav file; it uses Microsoft speech technology, so it includes the standard Windows voices, but also adds some new ones - including robot voices.
Do a web search under "text to speech"and “free” to find the many programs that use “MS Agents”. Some do certain things better. I use Deskbot now and used Cyberbuddy before that to read stories. http://www.deskbot.com/
