Re: Tips for speed reading /listening with screen readers


Steve Nutt
 

Actually, Eloquence is beaten now for best pronunciation, by a synth that is not on Windows.  Google TTS beats Eloquence hands down for pronunciation.  I can’t find anything in the dictionary that it pronounces wrong, but I can with Eloquence.  For example, Samhain, a Pagan festival is pronounced Sowan with a hard OW as in Cow, Eloquence completely mangles that and pronounces it Samhain as spelled.

 

Google One, Eloquence definitely Nil.

 

All the best

 

Steve

 

From: nvda@nvda.groups.io <nvda@nvda.groups.io> On Behalf Of Gene
Sent: 16 July 2018 17:21
To: nvda@nvda.groups.io
Subject: Re: [nvda] Tips for speed reading /listening with screen readers

 

What you listen to makes a difference, headphones allow for faster listening with fewer words missed at the same fast speed.  But the most important thing is what synthesizer you use.  I don't know about all the newer synthesizers so I don't know if any are as good at fast speeds as Eloquence.  In general, they aren't.  Eloquence is, in my sampling of synthesizers, the easiest to understand at fast speeds, 350, 400 or more words per minute.  It pronounces more words correctly without correcting with a speech dictionary.  The voice sounds annoying when listening at a normal rate.  It sounds raspy.  This raspiness is significantly reduced when listening at a fast rate and I don't find it bothersome.  Perhaps at first, it might be and I might have found it so.  But I quickly got used to it.  At times, for pleasure reading of long sections of text, a book, a very long article, I use a different voice such as the Dectalk synthesizer. 

 

But how anyone can stand most of the newer synthesizers is beyond my comprehension.  They slur words when joining them, leave off certain sounds when joining words, pause at unnatural times, use unnatural inflections, and I don't recall if I've left anything out.  these newer synthesizers work by sampling a specific person saying thousands or more words.  Then the synthesizer joins these actually spoken words together into speech.  They can produce words out of phonemes of the sampled voice as well but much of their speech is playing prerecorded words in the right order.  Think of how you would speak if you had prerecorded words in your brain and you spoke by a system combining prerecorded words.  Inflections would be wrong, words wouldn't be joined properly, in short, you would sound like these newer kind of synthesizers. 

 

Gene

----- Original Message -----

Sent: Monday, July 16, 2018 10:23 AM

Subject: Re: [nvda] Tips for speed reading /listening with screen readers

 

I have found that what I am listening to makes a difference too.
If it is too fast, and there is a great deal of specifics to analyze, the information may need to be read at a slower speed.  If it is more conversational like, then a higher speed works because we can mentally predict the flow of the writing.
Glenn


-----Original Message-----
From: nvda@nvda.groups.io <nvda@nvda.groups.io> On Behalf Of Brian's Mail list account via Groups.Io
Sent: Sunday, July 15, 2018 2:58 AM
To: nvda@nvda.groups.io
Subject: Re: [nvda] Tips for speed reading /listening with screen readers

Unfortunately I think it depends on the person. I know people who regularly read so fast I simply cannot understand it. I also find that it gives me a headache if I set it too fast.
 I imagine it depends on the person, what their first language is and if they are listening in it and also the actual voice and synth in use.
 I never get anywhere near 95, but then I have no pressing need to most times.

I do find the more artifical voices better at spead than the so called human sounding ones probably due to the  whay certain sylables aare handled.

Brian

bglists@...
Sent via blueyonder.
Please address personal E-mail to:-
briang1@..., putting 'Brian Gaff'
in the display name field.
----- Original Message -----
From: "Sociohack AC" <acsociopath@...>
To: <nvda@nvda.groups.io>
Sent: Saturday, July 14, 2018 10:22 AM
Subject: [nvda] Tips for speed reading /listening with screen readers


Hello season users of screen readers!
Advice me on this. I'm a student and require to read large texts on regular
basis. Although, this is something I love doing, it would be very beneficial
if I could improve my reading, or shall I say listening speed, with NVDA. I
would like to retain the same level of comprehension I have right now at
higher speeds. I have gradually moved up my way to 95% without boost in
NVDA, so I know it could be done. But, I'm finding it difficult to move
forward. Also, beyond 95% and in boost mode Espeak MAx starts to flutter. It
is still very comprehensible, but the fluttering voice is annoying. Can you
guys suggest me ways to upgrade my listening game? Do I need to switch to a
different variant voice of Espeak or shall I change my synthesizer? Is there
a cap to how fast can we listen?
All suggestions are welcome!
Also let me know at what speed rates do you guys read your screen readers
on/

Regards









Join nvda@nvda.groups.io to automatically receive all group messages.