Special Announcement: We have migrated to a new forum platform community.dsf.ninja. The old forum is locked, but it will remain open for reading. Read more.

Text-to-Speech Programs

hardware, software, tips and tricks
Forum rules
By using this "Production" sub-forum, you acknowledge that you have read, understood and agreed with our terms of use for this site. Click HERE to read them. If you do not agree to our terms of use, you must exit this site immediately. We do not accept any responsibility for the content, submissions, information or links contained herein. Users posting content here, do so completely at their own risk.

Quick Link to Feedback Forum

Postby futures_untold » Tue Jul 20, 2010 10:23 am

What is Text-to-Speech?

Text-to-speech programs allow you to type in text and create a spoken word audio sample. They usually have different 'voices' to choose from, and allow you to alter the speed and pitch the words are spoken at.

Here's a Google search for text-to-speech synthesiser.

------------------------------------------------

Programs

Two good ones are AT&T Labs Natural Voices and Cepstral Text-to-Speech.

You can use their online demo's which both allow you to save a wav file to disk.

Rivaling Microsoft Sam, Macs also have a built in text to speech program. (I'm unsure what it is called...PC user here.)

FL Studio has a text-to-speech plugin built right in! ---> http://flstudio.image-line.com/help/htm ... 0Synth.htm

Image

------------------------------------------------

Techniques

You may find it takes some tweaking to get the voices sounding right. Try editing the pitch and speed of the voice, and spelling words how they sound. Unusual placement of punctuation can also alter the tone and inflection of the voices, so try placing commas and full stops in different places to see if it improves the clarity of the pronunciation.

Entering one word at a time can often yield more natural results. with each word as its own sample, you can use your sequencer to adjust the timing between words more asily.

Many effects can be used to make the results of a text-to-speech synthesiser sound better. These include using chorussing, reverb, pitchshifting, timestretching and compression.

------------------------------------------------

Text-to-Speech vs Real Voice Actors

If you have a microphone, it will sound a lot more natural if you can record real people.

Writing a script before recording often helps actors to deliver better results. Try roping your mates and family into being voice actors for you. :)
Last edited by futures_untold on Tue Jul 20, 2010 11:36 am, edited 1 time in total.
User avatar
futures_untold
 
Posts:
4429
Joined:
Thu Dec 13, 2007 10:25 pm
Location:
London

Postby gnome » Tue Jul 20, 2010 11:03 am

Good stuff. You have recently been answering all my questions
User avatar
gnome
 
Posts:
4415
Joined:
Fri Aug 21, 2009 4:54 pm
Location:
Northern Ireland

Postby futures_untold » Tue Jul 20, 2010 11:26 am

Glad to hear that! :)
Last edited by futures_untold on Tue Jul 20, 2010 11:36 am, edited 1 time in total.
User avatar
futures_untold
 
Posts:
4429
Joined:
Thu Dec 13, 2007 10:25 pm
Location:
London

Postby wub » Tue Jul 20, 2010 11:35 am

AT&Ts one doesn't let you enter swear words :(


(NB: I know swearing is neither big nor clever)
wub
 
Posts:
34156
Joined:
Tue Feb 26, 2008 4:11 pm
Location:
Madrid

Postby Basic A » Tue Jul 20, 2010 11:50 am

:lol:

Right, right, get this one... seriously...

Open FL's text to speech...

Set note to whatever you choose...

voice to male...

now type.... yoy yoy yoyoyoyoyoyoyoy

Now resample

Now pitch down a full octave...

Now bitcrush. 8)
http://soundcloud.com/drjinx/raoul-juke-sous-le-pont-du-dr
http://soundcloud.com/teknicyde/targetneutralized

Image
:::::: Basic A. ::::: Teknicyde ::::: Dr. J!nx :::::
Phantom Hertz - Fentplates - Reboot Records - Cosmology - Applied Mathematics
User avatar
Basic A
 
Posts:
6037
Joined:
Tue Jan 19, 2010 11:53 am
Location:
Pittsburgh - You might know me as Teknicyde

Postby paravrais » Tue Jul 20, 2010 11:58 am

Another thread of win from you man! are all of these going to be added to the production bible at some point cos we don't want all of this great information to get lost in the nexus of dsf back pages..
User avatar
paravrais
 
Posts:
2869
Joined:
Wed Jan 07, 2009 12:31 am
Location:
Cambridge, UK

Postby futures_untold » Tue Jul 20, 2010 11:58 am

^^ Yes ^^ ;)

Basic A wrote::lol:

Right, right, get this one... seriously...

Open FL's text to speech...

Set note to whatever you choose...

voice to male...

now type.... yoy yoy yoyoyoyoyoyoyoy

Now resample

Now pitch down a full octave...

Now bitcrush. 8)


Nice! haha :)
User avatar
futures_untold
 
Posts:
4429
Joined:
Thu Dec 13, 2007 10:25 pm
Location:
London


Postby gnome » Sun Jul 25, 2010 2:14 pm

http://www.sitepal.com/ttswidgetdemo?mId=36631391.3
Check this one out. You can even add breathes
User avatar
gnome
 
Posts:
4415
Joined:
Fri Aug 21, 2009 4:54 pm
Location:
Northern Ireland

Postby Depone » Sun Jul 25, 2010 2:45 pm

Im loving the original badman Atari ST Speech
User avatar
Depone
 
Posts:
3526
Joined:
Fri Sep 18, 2009 4:49 pm
Location:
South-West UK

Postby sully_harmitage » Sun Jul 25, 2010 4:28 pm

^^
is there anything that sounds like that on a modern platform?
sully_harmitage
 
Posts:
302
Joined:
Wed May 14, 2008 4:10 pm

Postby Depone » Sun Jul 25, 2010 8:33 pm

sully.harmitage wrote:^^
is there anything that sounds like that on a modern platform?


just get one of the many atari emulators, and search for the patch. twas easy last night, annnnd im on a mac
User avatar
Depone
 
Posts:
3526
Joined:
Fri Sep 18, 2009 4:49 pm
Location:
South-West UK

Postby sixth sense » Sun Jul 25, 2010 9:08 pm

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

edit: someone might of posted this on already
User avatar
sixth sense
 
Posts:
505
Joined:
Fri Jul 02, 2010 5:38 pm
Location:
the other side

Postby sully_harmitage » Mon Jul 26, 2010 11:30 am

Depone wrote:
sully.harmitage wrote:^^
is there anything that sounds like that on a modern platform?


just get one of the many atari emulators, and search for the patch. twas easy last night, annnnd im on a mac


atari emulators are not good for my blood pressure or mental well being.
this does the trick though!!
http://dl.dropbox.com/u/1942830/stspeech.exe
sully_harmitage
 
Posts:
302
Joined:
Wed May 14, 2008 4:10 pm

Postby Ldizzy » Wed Sep 15, 2010 8:59 am

Basic A made me laugh for 3 minutes straigt. do u take donations?


Mac Os built in Text-to-speech is actually called text-to-speech, but the crookers have turned its coolest voice into a hipster phenomenon :(
Sharmaji wrote:2011: the year of the calloused-from-overuse facepalm
User avatar
Ldizzy
 
Posts:
1651
Joined:
Wed Feb 03, 2010 10:47 am

Postby jaydot » Wed Sep 15, 2010 9:23 am

This thread could be verry interesting
User avatar
jaydot
 
Posts:
5860
Joined:
Sat Dec 05, 2009 11:34 am
Location:
Your place or hers?

Postby Ldizzy » Wed Sep 15, 2010 10:45 am

jaydot wrote:This thread could be verry interesting


http://soundcloud.com/user1193462/texttospeech

1. Create a .txt file. no other extension will work... again this is for mac.

2. In it, put your text-to-be-read.

3. Save it somewhere logical and note down its path.

4. Open your text to speech window in the system preferences menu and choose your favourite voice.

5. Open Terminal and type in :

say -f "...txt" -o "...aiff"

the ''...'' are to be replaced by the exact location of your file on the harddrive and the destination path of your aiff file.

6. press enter, you robot-laboring motherfucker

http://soundcloud.com/user1193462/speech9


good innit?
Sharmaji wrote:2011: the year of the calloused-from-overuse facepalm
User avatar
Ldizzy
 
Posts:
1651
Joined:
Wed Feb 03, 2010 10:47 am

Postby Xince » Sat Feb 16, 2013 1:53 am

Is there anything on the VST side?
I'm too lazy to record from a software into my DAW.
User avatar
Xince
 
Posts:
35
Joined:
Wed Nov 21, 2012 10:59 pm
Location:
Kungshamn, Sweden

Postby mthrfnk » Sat Feb 16, 2013 10:45 am

Xince wrote:Is there anything on the VST side?
I'm too lazy to record from a software into my DAW.

You'll go far.
mthrfnk
 
Posts:
2731
Joined:
Tue Jun 14, 2011 11:05 pm
Location:
UK

Postby rockonin » Sat Feb 16, 2013 5:04 pm

I've used the Google Translate text to speech tool in a song. Shame you cant choose different voices.
Image
https://soundcloud.com/rockonin

ehbes wrote:I'll remember that when City wins the league :W:
User avatar
rockonin
 
Posts:
3515
Joined:
Tue Oct 16, 2012 5:05 pm
Location:
Buttoned Up

This forum is locked. Continue the discussion by creating a new topic on The New Dubstep Forum.
Next

Return to Production, Hardware & Technical

cron