Re: just a newbi problem, want to use festival, not flite with emacspeak

[tcross@xxxxxxxxxxx]

This can all be very confusing when you first start, so I'll try to
clarify some points. Apologies if I'm only stating information you
already know, but I suspect its probably a good idea not to assume any
level of prior Linux or emacs/emacspeak knowledge.

1. Speech dispatcher and emacspeak are two totally different systems
with no direct relationship with each other. 

2. To run emacspeak, you need emacs 21, speech synthesis - either
hardware, such as a dectalk express or a double talk or software
synthesis, either ViaVoice, Software Dectalk. There is also support
for flite (a lightweight festival and festival, but both of these are
not part of the core emacspeak distribution. I have used eflite and
some time ago tried festival, but to be honest, I didn't really like
them and they are not cleanly integrated into emacspeak in the same
way as the emacspeak "supported" synthesises. From memory, festival
requires custom patches to emacspeak. Therefore, I think of the two,
the eflite version is probably the best to go with if you must have a
free TTS, but note that I think you get what you pay for.. 

2. Speech dispatcher and speechd.el are a different minimalist
approach for providing speech feedback from within emacs. It has a
different goal to emacspeak in that it attempts to provide speech
support with minimal modifications to emacs and how it operates. For
this reason, I find its functionality isn't the same and I've had to
create a lot of my own additional customization to get good speech
support with some emacs modes. However, I have found it to be reliable
and at times have preferred the minimalist approach for some
situations. Its a matter of taste. I also think that speechd.el is
probably a little harder to use if you are totally blind compared to
emacspeak. However, if you have a little sight, this may not be an
issue. 

I don't run ubuntu, but I do run Debian and ubuntu is based on Debian.
You say you don't have emacs installed. This is the first thing you
will need to do. If you don't have access to the internet, then you
can install emacs 21 from the ubuntu install disks. You may be lucky
enough to fine eflite as well. 

With respect to your machine specs, I would say they are more than
adequate. Until a month ago, I was using a 1Ghz AMD Athlon with 256Mb
of memory and a SB Live sound card with absolutely no performance
problems using the software dectalk and ViaVoice. I also had no
problems using flite and festival or speech dispatcher. However, there
are some variables which can affect this, such as the sound card, disk
I/O and what other applications you may have running. Based on the
information you have provided, I would say that any of the software
synthesises would work adequately on your system. I use to run
software synthesis on a 600Mhz system with 128 Mb of memory and
performance was quite acceptable. 

Your comments regarding your ISP not supporting the protocol used by
linux and therefore you are restricted to internet connections through
windows doesn't make sense to me. Either your connecting via a modem,
in which case your ISP would support PPP and the PPP protocol is OS
independent. If you have some form of DSL, then you must be connecting
via a DSL modem of some kind and your computer is connected to that
access point via ethernet using TCP/IP, which is also OS independent. 

I would therefore suspect that if you cannot connect to the Internet
via your Linux system, then there is something misconfigured on your
system. It is possible your ISP is doing something really odd, but if
that is the case, I would find another ISP. Bottom line is that there
are no "protocol" differences that I am aware of between Windows and
Linux unless your ISP is using something very odd, such as a Windows
specific protocl, but I've never heard of that. If your ISP has said
that the problem is because you are using Linux and they don't support
that protocol, I would definitely get rid of your ISP as they don't
seem to know what they are doing and you will probably have constant
problems with them. 

Note also that if you can't find any other way, you can grab deb
packages from the ubuntu mirrors individually and download them onto
your windows box. if its a dual boot system, you can then boot into
Linux and install them using dpkg. However, you are likely to run into
dependency issues that will take ages to resolve, so I don't recommend
this approach unless you absolutely have to do it that way. 

I wouldn't worry about a deb package for emacspeak. Emacspeak is a
very easy package to install. All you need is the sources and a
working emacs 21 installed. then it is simply a matter of doing a 

make clean
make config SRC=`pwd`
make emacspeak 
make install

from the root of the emacspeak source tree. I always build emacspeak
this way. The only other step depends on the TTS engine you have
chosen and you will need to check the documentation for that engine
for specific details. Note that there are both instructions and tips
on testing things are working in the top level Makefile for emacspeak.
Ensure you have read them and you perform the tests suggested at each
stage to verify individual components are working - for example, it is
important to test the TTS engine is working before trying to run
emacspeak - this is the most common mistake people make and failing to
do so can waste a lot of time why you track down where the problem
lies. In general, its important to go through the documentation. I
would strongly recommend going through the users guide that comes as
part of the distribution. It is quite likely some of the information
in this guide is a bit out of date, but the principles are still the
same and there is some very valuable information that wil help you get
the correct mental model of how all the parts fit together.  

In fact, I have three versions of emacspeak installed at any
one time

1. The last stable emacspeak release
2. A working CVS snapshot
3. The most recent CVS snapshot

I DO NOT recommend using the CVS snapshots when you are just starting
as they are not guaranteed to work reliably (which is why I have two
which I move between and why I have the last stable version on the
system "just in case". 

Finally, you are going to find yourself confronted with quite a steep
learning curve, especially if you are new to Linux. The effort
required is worth it in the long run, but it is very likely you will
need to resolve problems yourself much of the time. You will need to
do considerable reading and a fair amount of experimentation is
likely. Keep notes on what you do, make backups of files before you
change them. If you make a change and it doesn't fix things, go back
and undo your change - this will reduce the chance of introducing more
problems as a result of your changes. You need to be methodical -
don't change things based on nothing other than pure guesswork. If you
think changing something will fix the problem, actually try to
reason out why you think this and be concious of assumptions. Try to
verify any assumptions and be prepared to address problems form a new
direction and not be stuck in the same mind set. Probably the best
thing to do is assume that 80% of the time, if something is not
working correctly, it is because of something you have done
incorrectly or something you have misconfigured. Bugs are actually
quite rare in Linux and emacs/emacspeak. Finally, as is often the case
with open source software, expect documentation to be a bit out of
date. Many people enjoy writing and working on new software, but very
few enjoy writing documentation. Often the documentation is not great
and is usually a few version behind the current software version. 

HTH

Tim


krishnakant Mane writes:
 > On 23/09/06,  <> wrote:
 > >
 > > Just a few points which may be of interest in this thread.
 > >
 > > Festival is a very powerful and configurable TTS. It can support quite
 > > a few different voices and has a lot of parameters which can be
 > > tweaked to improve synthesis. The "out of the box" configuration is a
 > > general default that is likely to meet most users requirements.
 > > However, I found (quite some time ago) when playing with festival, I
 > > could get better performance tweaking some of the settings associated
 > > with festival and the voice tools. Unfortunately, it was a while ago
 > > and I dont actually remember what all the tweaks were.
 > all right.  do you at least remember was it used with speach-dispatcher?
 > 
 > > Festival supports voices from other sources and in many distributions,
 > > you will get different quality voices - for example, in Debian, there
 > > are both 8k and 16k voices for US/UK english. The 8k voices don't have
 > > the same quality, but will probably give better performance on slower
 > > machines with less memory.
 > I have 256 mb ram with a p4 1.4 ghz processor and an amd duron machine
 > with 128 ram.  what will be the performance on these two?
 > by the way, I have given up my crave for deb packages of emacspeak for ubuntu.
 > I emailed the ubuntu-accessibility team, but they seam more interested
 > now a days in orca.
 > it is good but right now I really need emacspeak.
 > can some one give me idea of what tools I will need to compile
 > emacspeak and where do I get all the sources.
 > I don't even have emacs and no emacspeak of course.
 > apt-get wont work at my place because the protocol my ISP has given is
 > not working on my linux system.
 > so I am in windows when I use internet.
 > please help me if any one can.
 > 
 > -----------------------------------------------------------------------------
 > To unsubscribe from the emacspeak list or change your address on the
 > emacspeak list send mail to "emacspeak-request@xxxxxxxxxxx" with a
 > subject of "unsubscribe" or "help"
 > 

-- 
Tim Cross
tcross@xxxxxxxxxxx

There are two types of people in IT - those who do not manage what they 
understand and those who do not understand what they manage.

-----------------------------------------------------------------------------
To unsubscribe from the emacspeak list or change your address on the
emacspeak list send mail to "emacspeak-request@xxxxxxxxxxx" with a
subject of "unsubscribe" or "help"