[DECtalk] Intelligibility/Listenability criteria

Damien Garwood damien at daygar.plus.com
Sun Jul 21 04:17:48 EDT 2019


Hi Don,
Here are my criteria:

1. Understandability
As a screen reader user who has to listen to speech synthesis on a 
constant basis while using a computer, understandability is first and 
foremost. If the synthesiser can't be understood, then you're not going 
to get the feedback you need. In my opinion, ESpeak ticks every box, 
except this, so I can't use it.
2. Responsiveness. Again, because the speech is reading everything for 
me, I don't want a synthesiser that acts sluggishly with any kind of 
latency, whether that be a second, or 50 milliseconds, whether through 
lack of performance optimisation or through audio silence. When I press 
a key, I want instant feedback. This automatically rules out most 
natural-sounding synthesisers.
3. Accuracy: It needs to be able to read text accurately for the 
language it is designed for. It's not enough simply to have a phonetics 
dictionary, but it also needs to be able to distinguish between words 
(Present noun versus present verb, for instance).
4. Flexibility: The voice timbres should be available to the user, and 
for the most part should adjust smoothly to the change. This is 
important if a user has specialist needs and cannot use the synth in its 
default state. Speed and pitch are definitely a must. Again, this rules 
out natural synths, since due to the nature of recorded samples they 
start to begin to sound unnatural if you attempt to adjust the speed and 
pitch. The bigger the change, the more unnatural.
Like Jason, I also prefer formant synths. My favourite by far is 
Keynote, which to me is the most understandable, but I do love DECTalk 
for its flexibility. I also like Eloquence and the synthetic version of 
Orpheus.
Cheers,
Damien.

On 21/07/2019 05:53 am, Don wrote:
> Hi,
> 
> Perhaps a bit off-topic for this list... if so, my apologies.
> 
> I'm looking for opinions as to how one evaluates the "effectiveness"
> of a particular synthesizer.  Said another way, how one decides that
> synthesizer A is "better" than synthesizer B.  Ideally, criteria that
> would allow you to rank a set of them!
> 
> I've been auditioning various synthesis devices and techniques
> to try to come to my own conclusions on this.  Then, hopefully,
> work backwards to come up with some objective criteria by which
> they could each be "scored" (even if that was done using bogus
> rating units).
> 
> "Intelligibility" is, of course, the prime issue.  "Listenability"
> coming into play for any prolonged use.  Finally, "naturalness"
> when it comes to extended use.
> 
> For example, the old Votrax units were intelligible -- once you
> learned their "accent".  But, listenability was rather poor... you
> quickly developed ear fatigue.  And, the idea of naturalness was
> never even considered!
> 
> With gobs of resources (hardware, software, processing power), you
> can achieve quite acceptable results.  This seems to be the approach
> most "modern" synthesizers -- and techniques -- adopt.  The real problem
> lies with limited resources attempting to handle unconstrained input.
> (If you know what you're going to be asked to speak, it's really easy to
> come up with a good presentation!)
> 
> Limiting the user's exposure to the synthetic voice can reduce ear fatigue.
> So, dealing with it for 10 minutes might be tolerable while 2 hours
> would be torture.
> 
> But, having to face the prospect of completely unconstrained input can
> tax even that brief usage.  "Dr. Jones' car -- bearing the license plate
> FTDKTR -- has been parked in front of his house on Jones Dr. since 12:34A
> this morning when his Polish butler finished polishing it."  Imagine you
> have no other way of inspecting the input text...
> 
> So, what makes a synthesizer "tolerable" or "intolerable"?  What is the
> "threshold of pain" when it comes to tolerating an underperforming
> synthesizer?
> _______________________________________________
> Dectalk mailing list
> Dectalk at bluegrasspals.com
> http://bluegrasspals.com/mailman/listinfo/dectalk
> 


More information about the Dectalk mailing list