Hi
I have used and demo'd Dragon V6 on a desktop using a Talkmic - and to
my surprise (given the criticism it has received), it has been
excellent. Of 6 people who have tried it so far (inc me)on a previously
created voice file, only one (who has a strong Welsh accent) found it
to be unsatisfactory. However, after he enrolled his voice, recognition
was, again, spot on.
That machine is at Milton Keynes Access Centre - so it might be worth
asking Phil Satchell if there's anything significant/different about the
PC setup.
Regards
Peter
Michael Trott wrote:
> Dear David, Bernard, and all others interested. The mystery continues ...
> Both machines were purchased from one of these suppliers! But I use different
> microphones, the most reliable uses a USB mic, the least relaible uses direct
> input to the sound card, so was betting that's the problem. However, then I
> read Bernard's comments below:
>>From Bernard: >>>>Hmmm, I have Dragon 6 installed on a desktop and I've tried
> it with a
> pile of mics and several hours of training (not to mention a few
> reinstalls). It's shaky from session to session, including transcribing
> the same audio files saved on the HDD. It just reminds me of OCR a few
> generations ago. I wish it would work with some stability: there's
> buckets of applications for it if it were reliable. That's why I've
> played with it so much.
> Regards, Bernard <<<<
>
> So, I'm back to thinking Dragon 6 is wobbly just like some other popular
> programs.
>
> My copy seems to have regular problems with words like "gone" which it often
> interprets as a "," and "full stop" which it almost always prints as words,
> (Yes, I do have the English language version installed and have even trained
> it to put a "." It seems happy to put a full stop when I say "period".
>
> What really annoys me is that Dragon 5 on my laptop is so reliable but I feel
> that I ought to demo V6 during an assessment as this is what the student's
> are going to use. So, I usually show them V6 "warts and all".
>
> Thanks anyway.
> Mick Trott
>
> In a message dated 27/09/02 08:46:58 GMT Daylight Time,
> [log in to unmask] writes:
>
> << I would recommend running Dragon only on a computer that has been designed
> and tested for very low 'electronic noise'.
> iANASYST and MicroLink are well aware of the importance of this and take
> trouble to ensure that the laptops & desktops they supply have minimal
> 'electronic noise'. If this level is high then the software cannot readily
> differentiate between the auditory input and its own noise. Bit like trying
> to follow a conversation in a noisy pub.
> David
> David Grant, PhD.,
>
> ----------
> >From: Michael Trott <[log in to unmask]>
> >To: [log in to unmask]
> >Subject: Re: digital audio notes and transcription
> >Date: Thu, Sep 26, 2002, 7:23 pm
> >
>
> >In a message dated 26/09/02 12:46:37 GMT Daylight Time,
> [log in to unmask]
> >writes:
> >
> ><< Unlike contributors to another thread on this list, I have found
> > Dragon takes a lot of training to get anything like a useable level of
> > accuracy, particularly if you are a using a highly technical or
> > specialist vocabulary. >>
> >
> >I have Dragon 5 on a 1Ghz laptop and it seems to give almost perfect
> accuracy
> >anywhere I demonstrate it, regardless of the room accoustics and despite
> >minimal training (one reading of the one training text). However, I also
> have
> >Dragon 6 installed on a fixed location desktop that is much faster and has
> >more memory and it is very unreliable despite fixed location, longer
> training.
> >Anyone else found similar problems? If not I can only guess that the
> >microphone is inadequate.
> >
> >Mick Trott
> >>
>
>
>
--
01527 500324
[log in to unmask]
www.study-pro.com
Dyslexia Consultancy and Resources
|