What you are referring to is called a screen reader. It is very different technology. If we restrict ourselves to Windows XP, it includes what even Microsoft says is a very rudimentary screen reader. The reader comes with a minimal speech engine. Where things get a bit interesting is with the interface between those two parts. The screen reader communicates to the speech engine through an interface Microsoft calls SAPI. Eloquence also has a SAPI mode so it would be at least theoretically possible to greatly improve that reader just by hitching it to Eloquence. However, only the developers of the reader could do it.
Part of the issue that none of the three high performance speech engines, TrueVoice, Eloquence, and RealVoice, are exactly free. Ten years ago, the software development kit for Eloquence, basically the speech DLL and a few files to facilitate linking to it, was $1,000 No doubt that the SDKs for the others are even higher. Yet, that is just the tip of the proverbial iceberg. Any developer employing any of those great engines must then negotiate with the IP rights owners to share a percentage of the gross sales. I think you can see why most vendors resist going through that process, and apparently nobody is beating their heads down between their shoulders to drive home the reality that the quality is paramount. To me, it seems pointless to even have assistive or augmented communication technology if it doesn’t allow you to converse with the average Joe on the street. The indications I’ve seen say the speech technology owners don’t even expect a huge cut of any profits.
The value Xpress It adds in its light and flexibly intelligent user interface that makes Eloquence accessible to end users.
Sent: Wednesday, February 08, 2006 8:15 AM
To: Scott Royall
Subject: Re: FW: Left Coast vs. the Rest of Us?
The model most folks use now-a-days is to price software at entry level and make up the income difference in services / consulting.
Reading through your other e-mail this AM makes me wonder if you couldn’t use the product for more than intelligent interaction .. Could, for instance, it read to the blind?
Scott Royall <firstname.lastname@example.org> wrote:
Interesting comments from a blog reader. Yes, this company is doing
interesting things. And yes, the voice synthesizer they use absolutely
sucks. Just more evidence of what I’ve talked about in past entries. Even
forward-thinking medical research professionals are seemingly ignorant of
how far voice synthesis has advanced. You don’t even need much computing
power to get "near-human" voice quality, just good speakers and the right
software. Someone doing today what I did eight years ago would have a better
selection of synthesizer engines to choose from, since both AT&T and
Speechworks have other offerings of similar stature. Indeed, I think it’s
unfortunate that Speechworks bought out ETI, because they position Eloquence
as an entry-level alternative to their higher-priced RealVoice system.
Having done side-by-side comparisons of the two synthesizers, Eloquence
still seems to me to have a slight edge in diction and cadence.
As I hinted to before, I am seriously considering new pricing structures for
Xpress It. I still haven’t had any sales and my own financial clock is
winding down. Conventional wisdom holds that personal software should be
$200 or less. Of course, Aurora, the closest overall competitor to Xpress
It, is $250 with a much less capable vocal system. In all fairness, Aurora
is pitched as a general tool for disabled people, whereas Xpress It is
aggressively designed for verbal communication with complete strangers. The
latter is more for disabled people looking to interact with the world. I
suppose one option is to match Aurora on price and see what people do.
What do you think?
From: Carolm2345@aol.com [mailto:Carolm2345@aol.com]
Sent: Tuesday, February 07, 2006 11:22 AM
Subject: Re: Left Coast vs. the Rest of Us?
If you’re not already familiar with it, check out http://www.Cyberkineticsinc.com.
They had some coverage on CNN recently, and I couldn’t help but notice the
poor quality of the voice synthesizer they were using. For a non-techie
like me, their research is nothing short of amazing.
DJSloan .. Houston, Texas