[DECtalk] DECtalk TTS licensing

Don Text_to_Speech at GMX.com
Mon Aug 30 17:03:14 EDT 2021


On 8/30/2021 11:15 AM, Karen Lewellen wrote:
> And that, is at its best, provided by properly designed screen reading
> software.  Rich with choices so that the *individual* using the tool can
> decide  how much, which I likewise prefer, or how  little  inflection is
> presented.  Its like speed and  pitch.

The more you expect a user to "adjust", the harder your product will
be to use, effectively.

Developers often think they are being so clever by making all sorts of
adjustments possible.  The *cleverer* developer figures out what the user
NEEDS to be able to adjust and makes GOOD choices for everything else.

Imagine being able to adjust the air-fuel mixture in your vehicle,
while driving.  Or, the ignition timing.  Or, the sensitivity of the
steering, brakes, accelerator.

Chances are, you'd not muck with any of these -- because you
wouldn't be qualified to do so, skillfully.  Yet, the developer would
have expended effort to make those adjustments possible.

And, likely added bugs to the product in doing so!

Effort translates to cost and delivery schedule.  How long would you be
willing to wait for those things to be implemented and tested?  And,
what to pay for them?

While *you* may not find them desirable, someone else might.  Or, should
that someone else have a negative opinion on the adjustments that you relish?

Each design decision exposed to the user also cripples future
choices available.  E.g., if you like being able to adjust
*specific* characteristics of a voice (e.g., breathiness),
then that means future voices must be breathiness-adjustable,
as well.  Else you have to explain to users why that capability
has been elided!

The original KRM had just three knobs to adjust voice:
- volume
- pitch
- rate

And, because they were infinitely variable positions, it was
unlikely that you would get them set in exactly the same
places that you'd used, yesterday!

Yet, "good enough"!

> still, what is helpful to remember  may be this.  In many cases adaptive tools
> are extensions of, or even substitutions for human  function or activity.
> There is a difference between a simple tts, and  the many options a properly
> created screen reader, or voice navigator and the like provide.
> Unfortunately too many use the terms interchangeably creating the idea that tts
> equals screen reader, which is not true.




More information about the Dectalk mailing list