|
|
Subject: Re: A version of Speakup... for your cell phone? - msg#00083
List: linux.kernel.speakup
I would have thought it should be possible, but some tweaking may be needed
to get it working well. When you consider things such as mbrola has been
ported to symbian series60, and flite has been run on an ipod (I can't
remember the specific link, but it was part of the ipodlinux project).
Also it won't be the first phone to run linux, as I remember, when looking
how to connect my phone to my linux machine, I found some one who had tried
to put linux on a nokia 3650 (or similar).
Mobile phones are becoming so close to the specs of PDA's and such like, I
am now just waiting for the day when so much time has been spent on the
other features, that manufacturers forget to put the phone functions in.
From
Michael Whapples
Gregory Nowak writes:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Ok, that takes care of the CPU requirements, now what about the amount
> of RAM required, as it relates again to the resources available in a
> cell phone?
>
> Greg
>
>
> On Wed, Oct 04, 2006 at 11:23:10PM +0100, Jonathan Duddington wrote:
> > It shouldn't be a problem. I'm typing this email on a 200MHz ARM
> > processor, with eSpeak echoing words as I type.
> >
> > I have measured eSpeak taking 51 sec to process 6 min 3 sec of speech
> > output while processing text to produce a WAV file. That's about 14%
> > processor occupancy.
> >
> > I originally wrote it to run on a 12MHz ARM processor!
> >
> >
> > _______________________________________________
> > Speakup mailing list
> > Speakup@xxxxxxxxxxxxxx
> > http://speech.braille.uwo.ca/mailman/listinfo/speakup
>
> - --
> web site: http://www.romuald.net.eu.org
> gpg public key: http://www.romuald.net.eu.org/pubkey.asc
> skype: gregn1
> (authorization required, add me to your contacts list first)
>
> - --
> Free domains: http://www.eu.org/ or mail dns-manager@xxxxxx
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.5 (GNU/Linux)
>
> iD8DBQFFJES77s9z/XlyUyARAgyAAKC4j8ncBM98jByMtN0txg8F3alyowCgiDGm
> fMH/gxYg/8s2gQ5je3E4C3I=
> =PnXx
> -----END PGP SIGNATURE-----
>
>
>
Was this page helpful?
Thread at a glance:
Previous Message by Date:
click to view message preview
Re: A version of Speakup... for your cell phone?
Actually I don't think it would be to difficult. I suspect the
hardest part would be getting around the graphical user interface they
use on most of them. A friend has offerred to purchase one for me if
I'd port speakup to it but I haven't really had the time to do the
back ground research to find a suitable unit. if anyone has
suggestions with details and prices I will look into it.
I am not sure the trade off between needing to use a high end linux
cell phone/pda will work out any less expensive than buying one of the
subsidized phones with talx or mobile speak.
I don't think one would need to use speech dispatcher if memory would
be an issue. A small stub program could act as a go between from say
speakup to espeak. You could probably even modify espeak to open
speakups softsynth device directly and bypass using a stub program
completely.
I have to admit it is an interesting challenge, even considering I hate
telephones.
Kirk
--
Kirk Reiser The Computer Braille Facility
e-mail: kirk@xxxxxxxxxxxxxx University of Western Ontario
phone: (519) 661-3061
Next Message by Date:
click to view message preview
Re: A version of Speakup... for your cell phone?
In article <20061004233315.GA11387@xxxxxxxxxxxxxxxxxxxxx>,
Gregory Nowak <greg@xxxxxxxxxxxxxxxxxx> wrote:
> Ok, that takes care of the CPU requirements, now what about the amount
> of RAM required, as it relates again to the resources available in a
> cell phone?
I did a test on Linux, with the command:
speak -w test.wav -f longTextFile
This showed the speak process's memory usage as:
VmSize: 3.4 MByte (total amount of virtual memory).
VmRss: 1.5 MByte (total amount of physical memory).
I'm not sure why it's using virtual memory, but these figures seem
modest by today's standards.
Using the portaudio interface increases the VmSize to 28 MByte.
Previous Message by Thread:
click to view message preview
Re: A version of Speakup... for your cell phone?
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On Wed, Oct 04, 2006 at 07:45:15PM -0400, Kirk Reiser wrote:
> I don't think one would need to use speech dispatcher if memory would
> be an issue. A small stub program could act as a go between from say
> speakup to espeak. You could probably even modify espeak to open
> speakups softsynth device directly and bypass using a stub program
> completely.
>
You've got a point there Kirk, and it's one I failed to consider. I
guess I got too used to the way that speakup interfaces to software
synths now, to attempt thinking out of the box here. However, yes, I'll
admit that with speech-dispatcher/speechd-up out of the way, it does
look doable from the memory side as well.
Greg
- --
web site: http://www.romuald.net.eu.org
gpg public key: http://www.romuald.net.eu.org/pubkey.asc
skype: gregn1
(authorization required, add me to your contacts list first)
- --
Free domains: http://www.eu.org/ or mail dns-manager@xxxxxx
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (GNU/Linux)
iD8DBQFFJFqw7s9z/XlyUyARAmxcAJ4lk9Tgp/AnJXylhdv8CP9y4ruwbwCgiTcF
yhf2YGfSaDLi2w77lb8wL6s=
=9tmT
-----END PGP SIGNATURE-----
Next Message by Thread:
click to view message preview
Re: A version of Speakup... for your cell phone?
In article <20061004233315.GA11387@xxxxxxxxxxxxxxxxxxxxx>,
Gregory Nowak <greg@xxxxxxxxxxxxxxxxxx> wrote:
> Ok, that takes care of the CPU requirements, now what about the amount
> of RAM required, as it relates again to the resources available in a
> cell phone?
I did a test on Linux, with the command:
speak -w test.wav -f longTextFile
This showed the speak process's memory usage as:
VmSize: 3.4 MByte (total amount of virtual memory).
VmRss: 1.5 MByte (total amount of physical memory).
I'm not sure why it's using virtual memory, but these figures seem
modest by today's standards.
Using the portaudio interface increases the VmSize to 28 MByte.
|
|