Can someone direct me to a good computer voice gen.The kind that lets you chose male/female, tempo, export to .wav and all that jazz to use in songs.Like the "I may be paranoid, but no android " part in the radiohead song paranoid android.
Thanks
Party

dont know if its still around but shit-talker was pretty good ...

... as far a tempo goes i think you might need to get the calculator (or your audio editors timestretching function) handy

slainte rob

Fruity Loops speech is worth checking out.
Included In the fruity Loops Package.
I have done some neat wav exports and tunes with it.
Paul

afair the fl-studio speech IS shit-talker (with a few less options) ???

slainte rob

http://www.analogx.com/contents/download/audio/sayit.htm



-Link

Yeah! I'm totally all for Sayit!
It's free and uh... robotesque.
There's no tempo sync as it's a stand-alone program, but it exports .wav files.
I usually edit the individual words into small files and throw them in Battery (any rompler would do the trick) so I can trigger the words in time with the tune.
Hope this helps.

it slicing up syllable samples isn't complicated enuff for you, i have a voice vsti called synger available as freeware.. you program notes on the lower keyboard, and phonemes on the upper keyboard, so knowledge of/a guide to pronounciation is useful.

it can sing tho.. check 'honey' on my music page.

Current thread on similar topic HERE.

xoxos wrote:

it slicing up syllable samples isn't complicated enuff for you, i have a voice vsti called synger available as freeware.. you program notes on the lower keyboard, and phonemes on the upper keyboard, so knowledge of/a guide to pronounciation is useful. it can sing tho.. check 'honey' on my music page.

hey guy, why advertise and then it's not available?
where is synger ???

freur6969 wrote:

hey guy, why advertise and then it's not available? where is synger ???

it took me less than 2 min. to google it. i'm downloading it right now.

Cantor from Virsyn that's supposed to be out any day looks very promising too.

Devon

cos i forget i took them down, but "off the record" for the last few months i've been mentioning that they're all in the same place.. just grab a linkee that works and change the .zip file name.

DevonB wrote:

Cantor from Virsyn that's supposed to be out any day looks very promising too. Devon

Thanks Devon, wasn´t aware of this one!
Checked out the demo-video, and the best I can say is that the german guy´s got a very charming accent )
But the vsti sounds like any other artificial speechsynth back to C64 SAM, to me, but of course it´s great that we for the first time have a VSTI that in an easy way will let us use such in music.
I think 300 $ is quite expensive though, taking into consideration that SayIt for example is freeware, and that for the consumer all you pay for is a Sayit that´ll be able to sing and controlable from inside a vsti host. For some it´s a must having this of course, but it´s not really earthquaking technology, which is also btw. why that it hopefully will be very light on CPU.

@ simple tip

speech through vokko or mda vocoder gives you more midi freedom as speech alone in fruity
you can use any robovox source and apply harmonics using any vocoder with any carrier out there (many free)

remco

That's what I use most of the times: http://www.readplease.com/

Then I record the output with Abobe Audition, and play the .wav with a basic sampler (like Wave!) or use SR-202 to trigger different phrases/samples. Then come the effects (delay, reverb, vocoder, whatever).

Sayit! is useful just when is exactly that kind of voice we want, not very flexible.

I've tried d/ldng synger with two browsers, changing the path so that it ends with /PLUGINS/synger.zip, and I get an error. Any suggestions, xoxos?

Thanks for the other plugins, by the way - I will check out a number of them.

Cheers,

e.

This is webpage I have occasionally used to get machine spoken wavs. You cannot choose pitch or tempo, but their speech engines quality is good, or maybe I should say superb.

http://www.research.att.com/projects/tts/demo.html

Male and female voices, U.S. and british english, spanish, german, french.

Metanol wrote:

This is webpage I have occasionally used to get machine spoken wavs. You cannot choose pitch or tempo, but their speech engines quality is good, or maybe I should say superb. http://www.research.att.com/projects/tts/demo.html Male and female voices, U.S. and british english, spanish, german, french.

Did you use them in any songs of yours? Would be interesting to hear!

Btw, the guys over at www.fonix.com have some nice sounding speech synthesis products too. There used to be an online demo where you could type anything that it would read aloud, but I can´t find it anymore. Anyway, they do have a demo of one of their products at:

http://www.speakthis.com/

If you take the tour, it´ll do some speaking for you.

No, I havent made any finished tracks, just toying with them. I am one of those guys who have about thousand nonfinished tracks or pieces of tracks collecting digital dust.

Metanol wrote:

No, I havent made any finished tracks, just toying with them. I am one of those guys who have about thousand nonfinished tracks or pieces of tracks collecting digital dust.

Ok.. that sounds awfully familiar!

ha, this all reminds me of a few years ago,

i did a remix of "Even If" by Method Man.

i used an old text-to-speech program (probably worse than is available from those linked to in this thread) and typed all the lyrics, chunk by chunk, recording it to tape. Then i rerecorded it all back into my computer, and chopped it all to time (took ages!) then played a synth with it thru orange vocoder. Ha! it was a lot of fun... maybe i'll do something similar again some day.


www.angelfire.com/mech/haydxn it's still uploading at the moment but a link will be there soon just incase anyone's interested!

epiphanius wrote:

I've tried d/ldng synger with two browsers, changing the path so that it ends with /PLUGINS/synger.zip, and I get an error. Any suggestions, xoxos? Thanks for the other plugins, by the way - I will check out a number of them. Cheers, e.

you missed the other element, eg. giant lump of cheese on end of nob.. made so many of the bastards i forget that i called them.

synger is a vst effect.. my first phoneme thingy, but switching phonemes via cc# in my host had too crap timing for speech.. so what you're after is syng, not synger.. sorry. will maintain webpage before next decade

http://www.dancetech.com/file-library/PLUGIN/syng.zip

I just stumbled upon Windows Narrator under Start>Programs>Accessories>Accesibility. Reads aloud any active window with several voice/pitch/speed options.

Be careful when you use screenreaders and stuff like that. Many come with licenses that do not allow you to use the output for anything but listening to. Like the AT&T demo web page, if I recall correctly.

haydxn wrote:

ha, this all reminds me of a few years ago, i did a remix of "Even If" by Method Man. i used an old text-to-speech program (probably worse than is available from those linked to in this thread) and typed all the lyrics, chunk by chunk, recording it to tape. Then i rerecorded it all back into my computer, and chopped it all to time (took ages!) then played a synth with it thru orange vocoder. Ha! it was a lot of fun... maybe i'll do something similar again some day. www.angelfire.com/mech/haydxn it's still uploading at the moment but a link will be there soon just incase anyone's interested!

Wow, I just got reminded too now!

I did something like that too, like 8-9 years ago, using a speechprogram and then recorded every syllable into my AKAI S-900 sampler that I sold long long time ago. Then I ran it through the vocoder of my BOSS SE-50 effect, and fortunately didn´t have to stretch the samples, cause I could just play them at a lower speed, and that way make them fit when needing more sustain
It was a dance version of Elton John´s "Goodbye Yellow Brick Road", btw )

Been toying with a singing synth for a long time. Have got various none connected parts working but putting them together is very hard. Using phonetic language (english) mapped to ascii code and transcribed for midi control.
Done in C++ not synthedit I got sounds like homer simpson on dope! but tuned to midi key note pitch.
Of no use to anyone now but maybe in a year or so. I still need to setup many vowel/consonant samples.
The logic is:- samples-->Build words phonitically from ascii text-->voccoder-->pitch shift-->midi (for timing and tuneing).

Maybe it work. Maybe not. You never know till you try.

Regards
Etric

FL's speech generator is 100% ShitTalker. Also TalkIt ("for kids" as it says) is using the very same algorhythm.
IMO they are rather good when it comes to computerlike sounding voices, especially when you run them through further tools such as vocoders.
Actually, while it's a bit more of fun to have TalkIt and the likes doing the speech job for you, I kinda agree with Remco in that it's the easiest way to just record yourself and mangle things up through a vocoder of some sorts. That way you won't have any problems regarding the rhythm.

no wonder - file is synger2. you try keeping them all straight - freeware page updated

I just saw the avi demo for Virsyn Cantor.... is it just me, or does it sound an awful lot like Delay Lama? Either way, I'm sure I'd use it before touching vocaloid, from what I've heard...


Vorn.