Sl-c3200 Text To Speech Program

[kwy]

Does anyone successfully install/"migrate" the "Text to Speech" program from Sharp ROM to Cacko ?
Where can I locate the program?

FYI. The "Text to Speech" Program comes with SL-C3200.

[adf]

Does anyone successfully install/"migrate" the "Text to Speech" program from Sharp ROM to Cacko ?
Where can I locate the program?

FYI. The "Text to Speech" Program comes with SL-C3200.
[div align=\"right\"][a href=\"index.php?act=findpost&pid=151486\"][{POST_SNAPBACK}][/a][/div]

 try flite.

[HoloVector]

I wasn't able to find an IPK but I did manage to copy it over to Cacko.  Here follow the recipe in my Blog

http://www.tyrannozaurus.com/?q=node/215

[malik]

try the opera multimodal package, its much better than flite

malik

[speculatrix]

try the opera multimodal package, its much better than flite

malik
[div align=\"right\"][a href=\"index.php?act=findpost&pid=151716\"][{POST_SNAPBACK}][/a][/div]

I didn't realise Opera Multimodal was anything except a web browser, can you explain a little more please, like how to get it, how to use it?

[HoloVector]

try the opera multimodal package, its much better than flite

malik
[div align=\"right\"][{POST_SNAPBACK}][/a][/div]

I didn't realise Opera Multimodal was anything except a web browser, can you explain a little more please, like how to get it, how to use it?
[div align=\"right\"][a href=\"index.php?act=findpost&pid=151723\"][{POST_SNAPBACK}][/a][/div]

I did a search on this.  This what I found out - the browser has speech recognition and and text to speech functionality but, Text to Speech only works when you go to special websites.  The Multimodal package includes a sample website for you to test it out.  The Multimodal browser ipk is quite a hefty package and many Zaurii have trouble installing it.  IBM offers it for free if you create an ID with them and answer a bunch of survey questions.  Here is the the link to it:

[a href=\"http://www-306.ibm.com/software/pervasive/multimodal/]http://www-306.ibm.com/software/pervasive/multimodal/[/url]

Don't forget to follow Cresho's 7.55 install instructions here to fix it for the C series:

https://www.oesf.org/forums/index.php?showt...=15584&hl=opera

[Jon_J]

Sorry for off topic.  
The above IBM website has some interesting info about Netfront.
If you click the Netfront link, scroll down, on right side, you can eventually find this page.
http://www.access-company.com/products/net...tsdk/index.html
There is a trial SDK (software development kit) for Linux and GTK+ version of Netfront 3.2
I'm not very smart in this area, but it seems there could be a way to get a newer version of Netfront when this SDK is released.
I know it says it's for x86, but couldn't an arm version be made also?

[malik]

hello out there,

you can modify the sample files for your needs or write a txt2(whatever format) shell script doing this job. try it and you will hear the difference.

i didnt know about problems of this package with cxxxx models as i used it on a tosa. i tried it yesterday and the speech functionality worked out of the box. with cacko lite 1.23 there is also an opera 7.55 package but the libraries for speech recognition and tts seem to be missing.

malik

[ztep]

I'm working on a text-to-speech applet in order to use it on the next version of qpeGPS.

Although the tts applet included in qpegps sourceforge page is based of flite, my version will use mbrola and is based on mbrdico.

The advantage of this approach is that you can chosse the language between French, Arabic, English (US and Uk), Dutch and Spanish.
 
I have modified the command line version that works quite well and I'm working on the GUI version.

The GUI version will be a applet. When you click on the icon, a window is displayed (similar to memory applet or advanced battery applet in cacko) that allow you:
- Write text to be converted to speech
- Select the file to convert
- configure the tts options (language, etc)
You also can use the qCop interface to send text to the tts applet.

I'm still planning the applet, so, any idea or sugestion is welcome.

ztep

[malik]

I have modified the command line version that works quite well and I'm working on the GUI version.

[div align=\"right\"][a href=\"index.php?act=findpost&pid=151879\"][{POST_SNAPBACK}][/a][/div]

very nice! i tried mbrola before with the german tts txt2pho in combination with qemu as the
sources for txt2pho are not available. can you share your work, i.e. where can i download
the command line versions? thanx

malik

[ztep]

I have modified the command line version that works quite well and I'm working on the GUI version.

[div align=\"right\"][{POST_SNAPBACK}][/a][/div]

very nice! i tried mbrola before with the german tts txt2pho in combination with qemu as the
sources for txt2pho are not available. can you share your work, i.e. where can i download the command line versions? thanx

The command line version is only a working version to test code for the final applet. I have no plans to release a command line version.

You can try the oficial web of MBRDICO project ([a href=\"http://tcts.fpms.ac.be/synthesis/mbrdico/]http://tcts.fpms.ac.be/synthesis/mbrdico/[/url]). You can compile almost directly for the zaurus (some changes on Makefile).

I remember you that german language is not contempled in mbrdico, but in the same page there are utils to create news tree files for diferent languages. All you need is a phonetic dictionary (that is a text file with pair word - phonetic transcription with the phonetic simbols used by mbrola). Perhaps you can found a free german dictionary on internet and build the tree file.

ztep

[Jon_J]

I wasn't able to find an IPK but I did manage to copy it over to Cacko.  Here follow the recipe in my Blog

http://www.tyrannozaurus.com/?q=node/215
[div align=\"right\"][a href=\"index.php?act=findpost&pid=151538\"][{POST_SNAPBACK}][/a][/div]

I got it working on my C3100.
I just copied the files from my new C3200, using HoloVector's instructions at tyrannozaurus.
I was having a problem at first getting it to launch from an icon in Qtopia, but rebooting my C3100 fixed this small problem.  

[speculatrix]

I got it working on my C3100.
I just copied the files from my new C3200, using HoloVector's instructions at tyrannozaurus.
I was having a problem at first getting it to launch from an icon in Qtopia, but rebooting my C3100 fixed this small problem.  
[div align=\"right\"][a href=\"index.php?act=findpost&pid=152483\"][{POST_SNAPBACK}][/a][/div]

wa-hay, well done!

any chance of attaching a .tar.gz to the forum and we can see about ipk'ing it?

it'd be cool to get speech working on the Z. it'd be neat to add speech to kopi so that it beeps and says "meeting at ten oclock with jim".

[Jon_J]

Anyone know if any different voices can be downloaded for text to speech?
Any English documentation on this version?
How would you integrate it into another program? like the example speculatrix mentioned above.

[HoloVector]

Anyone know if any different voices can be downloaded for text to speech?
Any English documentation on this version?
How would you integrate it into another program? like the example speculatrix mentioned above.
[div align=\"right\"][a href=\"index.php?act=findpost&pid=152557\"][{POST_SNAPBACK}][/a][/div]

I hopped around the Nuance site looking for more voices but, I couldn't find any.    

Like most Zaurus things documentation is scarce.    

Has anyone translated the Japanese help file to English?

You can feed a text file to Text to Speech from the command line and if you have Auto Play Setup checked off under the Option menu it will automatically say what is in the text file.  However, it will fully open the TTS GUI and you will have to figure out how to exit the program automatically after the the text is read.