[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: espeakf experiences (was Re: Best advice for new user)
As the maintainer of espeakf, I would be very interested in hearing
from those who use, or have used, espeakf version 0.2. What do you
like about it? What bugs you? ...
Over all, I like it. In particular, I like having different voices.
But I cannot read Info with it: speaking stops at the end of every line.
Thus, the Introduction node in the Emacspeak documentation sounds like
this where I have put a space between each line to cause an
In the past, screen reading programs have allowed visually impaired
users to get feedback using synthesized speech. Such programs have been
commercially available for well over a decade.
Second, a speech rate change does not take effect immediately.
I change the speech rate using C-e d r . For the change to take
effect, I must switch out of the buffer to which I am listening and go
into another buffer and then switch back.
Third, the program should have some easy way to determine version. I
had to look in the README file to discover that I am using version 0.2
I should be able to switch to the espeakf directory and run a shell
command such as
or in Emacspeak, I should be able to evaluate this expression:
(shell-command "./espeakf.pl --version" nil nil)
The next problem may not have so much to do with espeakf as with my
overall use of Emacspeak.
How do I get a report on which voices I have available under which
conditions? And how do I set them?
My language here invokes the distinction between reporting a value or
set of values, as Dired does by listing the contents of a directory,
and an action that changes something, as when I use the d and x
commands in Dired to delete a file. A "reporter" is always harmless.
An "actor" may or may not be dangerous.
In this case, I want to know which voices I have: this requires a
report. And perhaps I want to change one: this requires an action.
In regular GNU Emacs, I can discover which typeset faces I have for
info-xref by typing
This reporter displays visually my various type faces.
I can act, which is to say, I can customize a face in one of two ways:
either by editing my .emacs file, which is what I normally do, or by
using the Customize facility.
However neither list-faces-display nor the Customize facility work to
report or change a voice in Emacspeak. What should I do?
Back to espeakf. Espeakf is not documented sufficiently.
I don't find anything when I look for `espeakf' or `Festival' in the
Emacspeak documentation. (I know that I should not expect any at this
time; but that is where I look.)
Since there was no documentation in the Emacspeak manual, I look in
the espeakf distribution. But that lacks documentation, too, except
for a little in the README file.
In the README file, the author expects the reader to know how to
program in PERL. That is reasonable for an alpha program, as this is,
but the PERL script itself requires too much knowledge and expertise
on the part of the user.
The PERL script should contain remarks which state which of the
variables in it that the README file says
You probably have to customize
With this documentation, you make espeakf more available to people are
poor PERL programmers and who don't know much. You increase the
number of people who can use espeakf and who can help you work on it.
This action does not help users who are not programmers, but at this
early stage in development, it widens your pool of support.
Espeakf does not produce as good sound as good as I expect.
Although eflite generates only one voice, its voice has better
quality. What should I do to improve the espeakf sound quality?
By default, espeakf and eflite generate different volumes. On my
system, eflite works best when I have it set to a volume of 85, which
is the value to which I default. I play music with that volume.
espeakf works best with a volume of 90.
I am not asking you to change your default settings, but please tell
me how to discover and then to change the volume inside of Emacspeak.
Finally, a general Emacspeak issue: what is a good way to set the use
of auditory icons in my .emacs file? I cannot find a good way to do
that. At the moment, I am using this kludge, which depends on the
local setting of the *scratch* buffer even though I am setting the
(if (eq emacspeak-use-auditory-icons nil)
(emacspeak-toggle-auditory-icons (quote (4))))
Incidentally, espeakf and eflite produce what sounds to my ear like
equivalent speech with different speech rate settings. I expect
that similar speech rates sound similar. They do when humans speak.
In my .emacs file, I include this kluge, which sets Emacspeak to
produce speech that sounds as if it is generated at the same rate by
the two programs:
(cond ((equal (getenv "DTK_PROGRAM") "/usr/local/src/espeakf/espeakf.pl")
(dtk-set-rate "330" (quote (4))))
((equal (getenv "DTK_PROGRAM") "/usr/local/src/eflite-0.1.1/fs"
(dtk-set-rate "270" (quote (4))))))
Robert J. Chassell Rattlesnake Enterprises
http://www.rattlesnake.com GnuPG Key ID: 004B4AC8
To unsubscribe from the emacspeak list or change your address on the
emacspeak list send mail to "firstname.lastname@example.org" with a
subject of "unsubscribe" or "help"
Emacspeak Files |
Unsubscribe | Search