creepy thing of the day

[shateredsoul]

http://192.20.225.36/tts/speech/b9d0aaac664d373502f71bdd5ae2334a.wav


http://www2.research.att.com/~ttsweb/tts/demo.php

[Ginsu Victim]

Your wav file link doesn't work.

[drventure]

Kind of interesting, but it doesn't seem to sound all that much better than Microsoft Sam, with the pitch tweaked a bit.

[patrickl]

These voices have existed probably for almost a decade.

The first TomTom Navigator versions used computer created voices already (which sounded better than this though)

[shateredsoul]

true true, my old mac performa (ha ha hah) had this. It's a shame that new pcs don't have this feature built in. 

I bugged my wife with it for a while by only talking to her through the text to speech. Maybe i'm the reason it was creepy, probably.

[patrickl]

true true, my old mac performa (ha ha hah) had this. It's a shame that new pcs don't have this feature built in. 

I'm pretty sure every Windows PC since Windows 95 came standard with text to speech. Go to settings and then Speech (or something similar). You can pick the default voice.

It usually comes with standard crappy voices (and a poor engine) though. That Sam voice is appalling (Stephen Hawking bad). You can buy and install much better voices. Like the ones from AT&T, but also many others. I like the RealSpeak ones, but mostly because they supply Dutch voices too 

NextUp has a website with lots of them

[newmanfamilyvlogs]

I've always found the Vocaloid system by Yamaha rather intriguing:

http://en.wikipedia.org/wiki/Vocaloid

[patrickl]

Indeed Dr Sbaitso! Compared to the current engines that sounds pretty bad though. Much like the standard SAM voice.

That Vocaloid singing voice, yeah that's pretty imprsesive.

I just remembered Loquendo also had a an emotional TTS engine. You can find some demos on their website. never quite figured out where to buy their software or voices though. They seem to work with big corporations only or something. I think their voices cost hundreds of euro's.

[RayB]

I first used that AT&T one back in the late 90's. I was impressed with it then. I wonder why 13 years later it has not improved.

[ChadTower]

I learned to code on a Ti99 in the third grade basically so I could use the speech synthesizer and tell my friends to do obscene things in that voice.   It was really impressive back in the day to be able to tell someone "Michael Jackson called - he wants his jacket back".

Hey, it was third grade.   

[AtomSmasher]

http://www.xkcd.com/768/

[RayB]

I was impressed with it then. I wonder why 13 years later it has not improved.

Yeah, and we could say the same thing about fax machines, video chat, and all that other crap we were promised 50 years ago....   

Hold on now, webcams can now shoot high def 30fps (or higher?), we have the bandwidth to carry it, fax machines aren't limited to 300 baud anymore... etc. Not sure what your point is.

[patrickl]

There is no comparison to the speech syntesizers of the late eighties early nineties and the software that we have now. It's lightyears better now.