Rockbox.org home
release
dev builds
extras
themes manual
wiki
device status forums
mailing lists
IRC bugs
patches
dev guide



Rockbox mail archive

Subject: RE: roadmap to blind Rockboxing (need an audio amateur/speaker)

RE: roadmap to blind Rockboxing (need an audio amateur/speaker)

From: Lee Marlow <lmarlow_at_yahoo.com>
Date: Thu, 26 Feb 2004 00:55:27 -0700

Maybe we could have mp3s of all phonetic "syllables" for a language and then
read words on the fly using an algorithm like soundex or metaphone. These
algorithms decompose text into its phonetic pronunciation.

Information and source code for Soundex and Metaphone can be found here:
http://www.creativyst.com/Doc/Articles/SoundEx1/SoundEx1.htm
http://aspell.sourceforge.net/metaphone/

This way new words, directories, songs, etc. could be supported without
additional offline text-to-speech processing. This could be the fallback
method if a dirname.mp3 file doesn't exist for the directory or a new menu
option appears that hasn't been recorded yet. It probably wouldn't sound
very good, but it might make be sufficient.

-Lee

-----Original Message-----
From: rockbox-bounces_at_cool.haxx.se [mailto:rockbox-bounces_at_cool.haxx.se] On
Behalf Of [IDC]Dragon
Sent: Thursday, February 26, 2004 12:23 AM
To: Rockbox development
Subject: Re: roadmap to blind Rockboxing (need an audio amateur/speaker)


> Can I ask what your resistance is to the phonetic idea? (That is a
> serious question)

Because this is pie-in-the-sky, versus for mp3 clips we have all the
infrastructure. mp3 clips from the buffer will take only minimum extra
resources, some slight changes to menu and screen code. A TTS engine is a
whole lot different.

> My push toward it is that within a couple of versions it could be
> multi-lingual; it would at least _attempt_ to read directory and MP3
> names,
> and the blind users (I guess) would happily rename all their files with
> appropriate phonetic names.

For this we can use talkbox, and you can generate the directory name clips
offline on a PC, using a text-to-speach script, if you like.

> From your post it seems the biggest hurdle would be playing samples <
> 1s

Hu? No, you can have clips down to 1 frame, which is in the milliseconds.
And latency is also no issue, either we have the clips in the mp3 buffer
(while not playing), or we don't.

Jörg

-- 
GMX ProMail (250 MB Mailbox, 50 FreeSMS, Virenschutz, 2,99 EUR/Monat...)
jetzt 3 Monate GRATIS + 3x DER SPIEGEL +++ http://www.gmx.net/derspiegel +++
_______________________________________________
http://cool.haxx.se/mailman/listinfo/rockbox
_______________________________________________
http://cool.haxx.se/mailman/listinfo/rockbox
Received on 2004-02-26

Page template was last modified "Tue Sep 7 00:00:02 2021" The Rockbox Crew -- Privacy Policy