Note: Due to technical difficulties, the Archives are currently not up to date.
GMANE provides an alternative service for most mailing lists. c.f. bug 424647
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">
<META NAME="Generator" CONTENT="MS Exchange Server version 5.5.2654.45">
<TITLE>RE: [gentoo-accessibility] speakup and software speech revisited</TITLE>
</HEAD>
<BODY>
<P><FONT SIZE=2>Hi Deedra,</FONT>
<BR><FONT SIZE=2>I am far from expert, but I will give it a shot. </FONT>
</P>
<P><FONT SIZE=2>First, for speakup itself make sure that you have software speech compiled in as a module. You will also need to create the /dev/softsynth character device (I don't remember the major/minor node numbers offhand), will look those up and forward. I presume you have alsa already loaded/configured, if not do so. I then emerged flite (the only tts I got to work with speechd-up). I then emerged speech-dispatcher and speechd-up. Speechd-up at the time required an entry in package.keywords to emerge. I manually edited a startup script for flite since it doesn't have one. I then did</FONT></P>
<P><FONT SIZE=2>rc-update add speech-dispatcher boot</FONT>
<BR><FONT SIZE=2>rc-update add flite boot</FONT>
<BR><FONT SIZE=2>rc-update add speechd-up default</FONT>
</P>
<P><FONT SIZE=2>Use a kernel command line which includes speakup_synth=sftsynth.</FONT>
</P>
<P><FONT SIZE=2>Making speech-dispatcher and flite start at boot loads them before speechd-up which is the proper sequence, getting speechd-up before speech-dispatcher results in much pulling of hair. On my system speech starts around the time networking is being loaded. flite is/was the default tts for speech-dispatcher and I personally didn't have any luck getting it to use festival (which was loaded/configured and worked in gnome just fine). However I didn't spend a lot of time on it either. I found a dectalk express on ebay which i got and as soon as it arrived and worked I disabled software speech, except for festival which is still present for gnopernicus to use. </FONT></P>
<P><FONT SIZE=2>As a side note, for ease of troubleshooting I wrote a script that would flip speakup back to the hw synth I had so if speech didn't come up I could just log in and type hwsynth to get speech.</FONT></P>
<P><FONT SIZE=2>------- cut here ------</FONT>
<BR><FONT SIZE=2>#!/bin/bash</FONT>
<BR><FONT SIZE=2>echo "synth_keyword" > /proc/speakup/synth_name</FONT>
<BR><FONT SIZE=2>----- cut here ----</FONT>
<BR><FONT SIZE=2>Replace synth_keyword in the above with the appropriate speakup designation for your hw synth. </FONT>
<BR><FONT SIZE=2>**** important *****</FONT>
<BR><FONT SIZE=2>Also check some of the commands above, some used _ instead of - in places, and I don't remember if it was in the command names or the names of the log files. The critical things are to create the character device, to start speech-dispatcher *before* speechd-up and to start your tts with an init script during boot. </FONT></P>
<P><FONT SIZE=2>Regards,</FONT>
<BR><FONT SIZE=2>Cecil</FONT>
</P>
</BODY>
</HTML> |
|
