Terry Dowling in
<[log in to unmask]>:
>I need to find out the relative merits of text-to-speech packages, to run on Mac.
>My student has one eye (with good vision, but tendency to visual fatigue) and memory/ concentration difficulties, from a severe head injury.
>She does not seem to have language difficulties, as such.
>Her need is therefore for the basic screen reading, rather than bells and whistles.
>Since she may use it for extended periods, the main requirements are: good listen-ability, ie most natural sounding + good accuracy (wide vocabulary).OS X V10.2 has built-in text-to-speech.
>How does it compare to textHELP Type and Talk or Kurzweil 3000?
>Any experiences gratefully received!
>Terry Dowling
Hi,
If you are interested in a screen reader for the Mac, Alva group do a
screen reader package called Outspoken, which was designed for
visually impaired users, and is comprable to such products as Jaws or
Windoweyes.
http://www.alva-bv.nl/alvacorp/alva_corp_home.asp
I should warn you however, that if OSX is the OS 10 release, then
Outspoken will not yet function on this OS release, and you must use
the OS 9 release or earlier. As Macs are not my forte, however, I
cannot give technical reasons for this.
Furthermore, the output given in Outspoken is quite different from
that of other screen readers, such as Jaws, which takes what is on the
screen and presents it in a manner which is easily understood with VI
people, Outspoken will take what is on the screen and read it back to
the user. I have been told that it can be operated with someone not
looking at the monitor at all, but it is easier if you can look at the
screen to start with as you are getting used to the system.
As for voice quality; depending in what the student is doing, if the
student is doing work such as word processing etc on the machine, they
will probably find it easier to use a more robotic voice than a
natural voice, since the robotic voices are much faster than the
natural voices that come with Kurzweil etc. Furthermore, if your
student finds they get very used to the system after a while, the
robotic voices can be made much faster, whereas the natural voices, in
order to keep their natural state, must be played at a preset level,
or very near the preset level, to ensure that the voice remains
stable. Outspoken does not require any separate speech hardware; the
new versions run using the Mac sound support.
Like with any product, Outspoken has a time bombed demo available; use
it with the student as well as other solutions and find the best
solution with her.
Andrew.
--
Andrew Hodgson, Bromyard, Herefordshire, UK.
Email: [log in to unmask]
|