Festival Speech Synthesis

Do you have a question? Post it now! No Registration Necessary

Translate This Thread From English to

Threaded View

I'm making my own minimalist embedded linux board, and I want to run
festival speech synthesis program on the board. Of course, there is no
sound card on the board, just CPU, RAM, and Flash so far.

I was wondering if anybody knows what kind of minimal hardware I need
to add so that I can hear the synthesized speech?

Do I simply need D/A converter connected to a GPIO port, then
amplifier, then speaker?

I guess I would have to write some kind of linux driver also that
festival can send digital speech samples to it, then the driver would
simply write that byte to the GPIO port to be converter by the D/A,
then amplified and then to the speaker.

Will that work? Does anybody have a clue?


Remember patient man ride donkey!

Re: Festival Speech Synthesis

Quoted text here. Click to load it

There was the Covox Speech Thing many years ago. It is a simple D/A
converter you put onto the parallel port. There was even a driver for it
(pcsndrv), but I don't know if it works under newer linux kernels. Speech
synthesis was possible through it, but the pc speaker made it possible, too.

The hardware itself is only an array of resistors and a capacitor and very
easy to build (I use it myself). It's capabilities are 8 bit mono, but the
frequency can be chosen as high as you like it (using my 286-12 I get around
22 kHz, but much more is possible using a faster CPU). All you need is an 8
bit wide output channel (or less).


Site Timeline