Computers are still pretty dumb (as far as I know :D) and still suck while trying to do things like interacting using human languages for instance. But there are certain fields that computers have shown pretty interesting “capabilities” too. The text to voice/speech conversions is a great example.
If you’re willing to pay some decent amount of money and use Windows or Mac OS X, then there are pretty awesome proprietary text to speech engines available. They may not be as advanced as some of those premium (and proprietary ones) tool, but there are few noteworthy open source projects too.
If you’re looking for an ease to use text to speech utility that can be used in GNU/Linux (Ubuntu in this case) then Gespeaker is a tool that you can try.
This is actually a graphical front-end that uses the “espeak” (command-line based speech synthesizer, cross-platform tool) as the engine. But because Gespeaker is a GUI, anyone can use the “espeak” engine easily to do things like to add few additional voices, easily add texts, record etc.
Main features…
*. A simple UI.
*. Comes with inbuilt “Male” and “Female” voices (the ones come with “espeak”). You can manually adjust voice settings such as:
1. Change the pitch.
2. Volume level.
3. Speed.
4. Delay between words.
*. It even lets you translate between a huge list of languages too!. For instance, you can input some texts (say written in English) and can simply use the output language to “French”.
*. If you don’t like the quality of the built in voices (I wouldn’t blame you), then you can use another enhanced voice synthesizer project called “Mbrola” (more below).
*. Use ALSA or Pulse Audio for the audio output.
*. Enable Text Wrap (which is useful when copy & pasting a large portion of text/words).
*. Recordings are saved in the WAV format.
You can access few more settings via its “Preferences” window as well.
You can install Gespeaker in Ubuntu 12.04 Precise Pangolin, 11.10 Oneiric Ocelot, 11.04 Natty Narwhal, 10.10 and 10.04 by using the below command in your Terminal window.
sudo apt-get install gespeaker
If you’re dissatisfied with the quality of the built in voices then you can try the “Mbrola” voices which are slightly better. For installing them, use the below commands in your command-line window.
Update: Ubuntu 12.04 users don’t have to go through all this hassle and can install ‘Mbrola’ voices by simply using their proper package name.
sudo wget -O /etc/apt/sources.list.d/ubuntutrucchi.list http://www.ubuntutrucchi.it/repository/ubuntutrucchi.list
wget -O – http://www.ubuntutrucchi.it/repository/ubuntutrucchi.asc | sudo apt-key add –
sudo apt-get update
sudo apt-get install mbrola
Now we should have installed the Mbrola engine. It has few built in voices in different languages (American English, Italian, French etc) that you have to install manually. For example if you want the default Male voice in US English, then use the below command.
sudo apt-get install mbrola-us2
To enable this newly installed voice in Gespeaker:
1. Go to “Edit” -> “Preferences” and then click on “Mbrola Voices” tab. Under list, you should see the newly installed packages marked with a Greenish Tick.
However in my Ubuntu test machine (using the 11.10), Gespeaker did not detect any installed Mbrola voices because the default “Mbrola Voice path” was set to a wrong one.
You can easily fix this. And for that,
1. Click on the drop down menu next to “Mbrola Voices path” (in “Mbrola voices” tab) and open the below folder path (as shown in below screenshot).
/usr/share/mbrola
2. Yep that should do it. Now scroll down the languages list and under your newly installed language/s it should display a Green Tick.
For a complete list of available languages and their proper installation commands, please visit this page.
So as usual, as a final word: although as said before, when comparing with some of the more advanced text to voice engines, “espeak” certainly has some catching up to do.
But if you’re looking for a completely free, extremely user friendly and simple text to speech converter, then Gespeaker looks pretty good.
4 Comments
I have installed gespeaker 0.8.1 and mbrola voices, spanish in particular and used the tip above to change the right carpet of the voice to get the green tick available. Just as is typed above. All of that is great, but i cant use this "new" voice no matter what. Whenever I am in gespeaker, it remains the same original and bored voice and cant change it to the new one. Do not know what to do… any recommendations? Am i doing something wrong?
Hi 'Eduardo',
I'm using 'openSUSE' at the moment and don't have Ubuntu so I could attempt to test it again. However, after installing those voices have you navigated to the install location ('/usr/share/mbrola') in Nautilus file manager just to see that they're actually there?
If they are there, then once added in 'gespeaker', try clicking on the 'Refresh' button so it'll load the list again. Has it helped?
BTW, what version of Ubuntu are you using?
I’m in Debian Wheezy. I’d classify myself a pro Linux user. I thoroughly tried everything I can think of and no matter what, there is nothing showing up on the mbrola voices list in the mbrola preferences in either the absolute most recent version of Gespeaker for debian/Ubuntu (at this time), 0.8.1. I installed mbrola and I installed mbrola-en1, mbrola-us1, and mbrola-us2. Their folders (en1, us1, and us2) are directly under the /usr/share/mbrola folder.
Update:
I correctly set that folder in gespeaker’s preferences for mbrola and it then displays that as the folder and no I didn’t forget to click ok to save settings. , but when I either hit refresh or just close gespeaker and restart it, either way, the voices list stays blank. Mbrola package is installed and gespeaker confirms it’s detected. The default voices work, but they’re too robot-like.
Update:
And I forgot to mention the other things I tried after each and every one of the common sense things didn’t work. I tried different versions of Gespeaker, including that in the Wheezy repository and the latest in the Ubuntu repo, but none before .8 I think. Just for the ‘ell of it, tried also creating a voices subfolder under /usr/share/mbrola and copying the mbrola voces under that as well, because it looks like they used to be there in previous versions, and updated the preferences of Gespeaker temporarily for that, and yet still nothing in the mbrola voices list of Gespeaker. I could use suggestions. Other people are having this problem. I’ve seen people posting the identical problem. What can it be?
Hi ‘Daunn’,
I just installed ‘Gespeaker’ (0.7 version) in Ubuntu 12.04 (freshely installed) and it worked like a charm. I did not have to manually enter the ‘Mbrola’ voice path either (I installed the exact same 3 voices that you installed for testing).
However, I don’t know it it has anything to do with it, but I installed all of them in a single command. In other words, I did not get to open ‘Gespeaker’ before installing ‘Mbrola.
The reason I mentioned that was because, perhaps the existing configuration file of ‘Gespeaker’ could be conflicting with your manual settings?
So, have you tried removing the ‘Gespeaker’ settings file?
I know you know this, but just in case, you can use the below command to get rid of that existing configuration file of ‘Gespeaker’.
Then try re-opening ‘Gespeaker’ (also make sure undo the manually copied ‘Mbrola’ vocies too).
Could you let me know if it has helped you or not?
Note: My ‘Mbrola’ voices are located in ‘/usr/share/mbrola’ folder. And the individual voices are saved in three folders called ‘en1’, ‘us1’, ‘us2’ (and each folder has two files inside them).