Text to Speech

There is a fantastic freely-available (X11-like license) Text-2-Speech tool called Festival that you might check out. I believe the majority of the code is built in libraries, which at least in theory would make it easier to build into the Empeg. (i.e. just throw the appropriate hooks into the player software to call the Festival API calls. It's a simple matter of programming, you see...)

It's possible that whomever it was a couple messages back mentioning a T2S software that could mimic personalities was thinking of Festival. You can, if you are so compelled, build "phoneme libraries" from any voice you want to take the time to sample and take apart and Festival will then use that voice to speak to you. They still sound pretty computerized, but noticably less mechanical than Stephen Hawking's chair for example. :) (To help those crazy^H^H^H^H^Hdedicated enough to want to build their own speech libraries, there's a thing called Festvox to help you.)

(When looking up Festvox just now, I also ran across an interesting-looking project called Knight Rider.)

I agree with probably everyone else that were a feature like this available, it would have to have an "On/Off" option. I also agree that it could be extremely cool.

All I want is a car stereo that can talk to me in the voice of Hal from the 20xx movies. Is that really too much to ask?

:-)

-Mike

I'm sorry mike, I just can't do that.

Kureg

The last time I brought up the notion of Festival in Hugo's presence, he started muttering unkind things about memory usage. Apparently TTS in general is rather memory intensive. I think the intended solution was to be able to attach a bit of mp3 sound to each menu item. empeg would put together a set of files for the standard menus, and leave it up to the user to provide clips for playlists.

(Could be entertaining... the Pink Floyd playlist would be identified by a single heartbeat off Dark Side of the Moon...the mind reels...)

-Zandr
Mk.I #150
Mk.II #39

The last time I brought up the notion of Festival in Hugo's presence, he started muttering unkind things about memory usage. Apparently TTS in general is rather memory intensive. I think the intended solution was to be able to attach a bit of mp3 sound to each menu item. empeg would put together a set of files for the standard menus, and leave it up to the user to provide clips for playlists.

Yeah, but that's not really TTS, that's cheating! :)

Nope, Mem usage cannot be the point, remember they did that old "S.A.M. reciter" text 2 speech on a C64 and it also ran on my ATARI 600XL when it only had 16KByte of RAM ...
And its Quality was good enough, i even think those "roboter voices" are far more amazing than a "perfect female voice" ( under *some* circumstances i have to admit :-) )

My 0.02 Euro ...


Nils

( If i dig deeper in the dark regions of my mind i also remember text2speech on the Radio Shack TRS-80 with 16KBate that my best freinds brother brought to our village, and that led all of us kiddies to be computer freaks, maybe majorly because of that thing actually *SPEAKING* , so *never* underestimate the facination of a computer speaking to non-techie people ( and that's what we were at the age of 8 ). )

I'll grant that TTS is possible in small-memory machines. But Festival isn't going to be the solution:

To quote from http://www.speech.cs.cmu.edu/festival/manual-1.4.1/festival_30.html#SEC138

The system is too big. It takes a long time to compile even on quite large machines, and its foot print is still in the 10s of megabytes as is the run-time requirement. Although we have spent some time trying to fix this
(optional modules have made the possibility of building a much smaller binary) we haven't done enough yet.

So Festival has a ways to go yet...but I'm sure there are other solutions out there.


-Zandr
Mk.I #150
Mk.II #39

MP3 clips are much more likely to happen than full TTS - which *is* generally very memory intensive. Using low bitrate clips we can get a lot of stuff into (say) 128k - over 15 seconds. The general idea is to have a special playlist into which you can upload your own clips (but there will be empeg-provided ones too).

All this sort of stuff is waiting for the new codec architecture (which is coming soon). When we can mix two currently playing mp3s (ie, music & sound clip) then I'll be Mike/John/Peter's worst nightmare and will want crossfading too... ;)

Hugo

Hi,

Regarding text-to-speech, I once read about someone who'd come up with a text-to-speech system that mimicked famous personalities. For example, you could type a sentence and it would be read back to you by a voice similar to Jimmy Stewart or Jack Nicholson. Not just the voice and inflections, but even speech patterns, such as Jimmy Stewart's semi-stuttering style.

I'd opt for Joanna Lumley, the sexiest femail voice I can think of.

Regards,

_________________________________________
John, (MK1 114-20G, MK2 15-36G).

What would I do with T2S?

I'd write an app for my pager that would send all incoming pages out it's IrDA port. Then when I get in the car, I can run my new prog on my pager and set it so it's IR is inline with the empeg. Then I'd like to have a T2S prog on the empeg pause the music, read the message to me and unpause the music.

I've been trying to think of what else I could have my pager and an empeg talk about (via IR), and this is all I've been able to come up with (so far!).

--The Amigo

I'd write an app for my pager that would send all incoming pages out it's IrDA port. Then when I get in the car, I can run my new prog on my pager and set it so it's IR is inline with the empeg. Then I'd like to have a T2S prog on the empeg pause the music, read the message to me and unpause the music.


Although I don't use a pager, this is a cool idea. One thing I like about this BBS is the great ideas I hear for projects.

Alex Lear

A huge YES would be my response to a TTS or mp3/list response & menu reading. Empeg is fine now, but when I'm driving with 6" of slush and 60' visibility (at my usual 80mph of course) there's NO WAY I'm looking down.

-Zeke

just say you weren't paying much attention...