When everything was the good old days, or I mean when Eloquence was the biggest kid on the block, it was fast.

The closest to good old Eloquence is Espeak. But I miss how fast Eloquence could read.

The very newest versions of Espeak, at least those in NVDA, can go twice as fast as Jaws at the top speed.  I believe the new stand-alone command line version implemented Libsonic as well, and my biggest complaint with orca is that you can't get this functionality there in anything resembling a sane way--my attempt broke the sound system permanently, and I'm not going to use Vinux, sorry.  That, and you can't set intonation/prosity.
I've got 63% with rate boost, and rising.  I used to think 75% with jaws was fast, but not anymore.  I'm just turning it 1% faster every couple days, and I'm doing it because I can.  I switched screen readers and decided that I would do it as part of the whole different (in my opinion better) environment and self-improvement type of thing, and expect to max it out sometime by or in the fall.  I never slow it down--that would defeat the purpose.  This includes books, instant messaging, everything.
I quite literally can't, this isn't the can but doesn't like to can't here, function as slow as Jaws's slowest speed anymore.  I can't function below 50% with jaws, end of story.  People get frustrated with me because I have to turn it up even for simple, quick things when not on my own computer.

Well, I have been an Eloquence user since the day I learned about how to use a computer. I remember a time when 25 percent was comfortable for me, and 30 was too fast. Nowadays I can't stand anything below 50. Normally I hang around 75 to 80. I slow it down to 60 or so when I'm reading something unfamiliar, especially if it is a longer text. I can understand quite good at 90, and if I concentrate, I can understand 100, but I can't stand the sound of too fast speech. Even if I can understand it, it is unpleasant to listen to above a certain rate. That rate is different for different synths. With Eloquence it's around 95. With E Speak, it's around 20 with rate boost. Above that, it is speaking so fast that consonants are clipped, and the synth sounds sick, especially with Espeak. Eloquence isn't so bad, I just don't see why I really need to go the additional 20 or so percent, since I don't want a synth that speaks so fast that I don't have time to understand the meaning of the words it's saying. I find it interesting that Keynote gold never got mentioned... probably nobody uses it anymore but I still like using it on occasion, since i still have it for the time being, on my M power. I stick to rate 10 or so with that, I can understand it at top speed but I just don't like listening to it since it sounds like it's struggling to speak that fast. So rate 10 is a nice and comfortable speed for me that I don't have to touch, and when I use it, I'm most likely reading something huge. I'd use the braille display for other things since I am a proficient braille reader. Not nearly the best, and probably not even great, but enough to turn the speech off unless I specifically want it.
As for newer synths like Nuance, Ivona, and others... I hate speeding those up because the quality of the voice suffers tromendously, even with 5-10 percent adjustments. Male voices especially dont' sound pleasant sped up. Neospeech is the only modern synth I enjoy having sped up. Back when AT&T natural voices were state of the art, I do remember playing with them a bit, and thought they weren't too bad, but can't really remember it that well. Same goes for talking books; I generally keep the speed at normal, however most daisy players have pretty good speed changing algorithms, so the distortions they introduce aren't nearly as severe as they could be, and listening to sped up books is actually something I don't mind doing if I am pressed for time or if the narator is a 90 year old who reads slower than the slowest of Eloquence users.
I have a few friends who kept Eloquence at its default rate, and I got pretty much the answer as luke did. They say, "I can understand it that way". Then again some sighted people can't even understand Eloquence at that rate. My dad, for instance, had trouble understanding it, and his computer even had expensive speakers... making Eloquence sound like it was playing over a decent theater system.

That is a very interesting topic.
I sometimes set Eloquence up to a hundred percent on NVDA to do things I know I am doing and I know what to expect, and it makes things easy to read in some cases.
You see, sometimes I may know what I am doing. For instance, if I want to really skim through text with a rough outline of what it sounds like, I might speed my speech synthesiser as fast as it will

go so it will sound almost gibberish, but there is a little bit of a sense of what it is saying, it is just not direct. When you read using text-to-speech, and you spell a letter one at a time just by hitting the cursor keys, do

you call that reading because you are listening to the individual letters and you are putting them into words? I just find that interesting. I might be able to read through it slowly, or I might be able to convert it into

Braille and read it using text and word prediction because I use that a lot when I do not want to be hit with sudden emotions. I want to let it sink in slowly rather than listen to it barrelling at me with full force.
Even when I have speech at a hundred percent and I read messages I cannot identify what someone said. I have to slow it down to hear what it said, or spell letter by letter and then put it together.
On a side topic, does listening to speech at eight hundred words per minute affect your daily speech rate in general? I noticed that people have a tendency to get used to their spech syntheesiser's fast rate that it affects their regular speech pattern, thus making them hard to understand. Is this true?

hi,
when i started using a computer back in 2007 and was introduced to jaws, i simply didn't knew that the rate of speech can be increased or decreased, lol.

when my computer teacher pointed that out to me, i experimented a bit trying to keep it around 30. but that was pretty fast for me at that time.

slowly, i adapted myself to the increasing rate of eloquence, and now i work with a rate around 65. no more, no less.

whether its reading a website or a book, playing a game or checking email, it remains the same as it is clear to understand, and getting it more faster leads to confusion at least in my case.

however, now i can't even bear a moment of speech at the slower 15 percent, so much so that it irritates me when i've to install jaws from scratch ever.

as for nvda, i keep it a bit slower.

i also tried experimenting with sappy using jaws, while espeak was ok to comprehend at fast rates, ms marry and mike sounds weird after increasing therate.

regards

I use Eloquenjce at 85% these days, slowing down on occasion for unfimiliar or important material. It's almost gotten to the point where I cannot stand anyting under 50%, and the default rate of Eloquence in JAWS for me just sounds plain awful. I cano only stand to speedup human SAPI voices so much. After a certain point, they start sounding extremely jerky and clipped.

Well, I doubt you can find that exact info, let's say you read a sentence containing the town name Llanfairpwllgwyngyllgogerychwyrndrobwllllantysiliogogogoch... That takes up quite some space, and also takes quite some time to say, and yet it's only one word...

These days I definitely can't stand anything less than 65% or so. On NVDA    I have Espeak setup to about 75% to 80% on average. When skimming web pages or emails I have Espeak cranked up to about 80%, but for programming I slow it down to say 75% just to make sure I don't overlook something important. If I am using Eloquence with NVDA I can go higher.

Well, so far as I can tell, I'm the fastest on this thread.  I wonder if that's actually true.  Since there's interest, I'll tell the secret (well, it's not really that big a deal).
Back in high school, I started playing muds, with an old command line client.  I didn't know about all this fancy sound stuff back then, and thought what I was doing was normal.  Long story short, Jaws was at like 30% or some such, so forget about keeping up.  I got fed up with this obviously, so I came up with a solution.  I'd turn Jaws faster by as much as I could possibly understand every Saturday, and would not turn it slower for *anything*.  Turning it slower to "enjoy" stuff is the biggest mistake people make for improving the speed.  You have to read large bodies of text with it turned to the highest point where you can understand it.  You may have to reread parts.  The more technical and full of big words, the better.  Throw in symbol level most or all, for extra benefit (I recommend most as the everyday punctuation setting, but not for this reason specifically).  The faster it gets, the harder you fight for another few percentage points.  I recommend finding novels and reading those.  I stopped at Jaws 75% for a number of years, switched to NVDA, and decided to keep going.  I didn't expect to get this far this quickly, though I definitely expected to get this far.
I'm on Espeak En-US variant none, but the specific voice shouldn't matter too much.  Pick one that you like.  Some may work better with libsonic, the method used by NVDA for speeding it up above 100%--it's basically overclocking speech, I'm not really sure.  Eloquence starts clicking when I go into rate boost, and the recent versions of the add-on floating around don't seem to have that functionality.  By clicking I mean *really* clicking.  Espeak doesn't have this problem, but it will sometimes stutter if you're on a computer from like 8 years ago when you get this fast.  none of the college computers and stuff that I've run it on have had that problem; it's only that my family uses a computer 5 minutes a day, if that, while at home, so the "family computer" is an epically old laptop (I think it's only 1 or 2 gigs of ram, it's only dual core, it tries to run Vista, you get the picture).  I've got my own, though, so not a problem.
The biggest point I can stress here is, don't ever turn it slower, ever again.  The only exception I allow myself is if i switch from headphones to speakers.  I do everything, including programming, at this rate, and that's what allows me to turn it faster in a few days.  In fact, I'll probably turn it up another percentage point today.  And a side benefit: Espeak will read properly capitalized sentences different than not capitalized sentences, which is a major boon to proofing.
As for affecting my speech patterns, yes, it kind of does.  Do it for long enough, and you will start talking faster.  I could just be a natural fast talker, it's hard to tell.  The truth of the matter is that I talk *really* fast,so...who knows.  That, and I tend to use larger words than those around me.  I think that it's like the rest of the world: we talk like those around us, but the person that talks to us the most is probably the computer.

Would any of you mind if I use some of what you said in your posts for a little research project I'm doing?  Of course I wanted to ask this before I just went and did it.

Anyhow, it seems my inference was not at all supported by other users' experiences.  It's pretty clear that many of you do keep your speech at a pretty quick rate, more so than I do.

I suppose my friend is just one of those people that just prefers the slow, very very slow speech.  I don't understand that one bit, but hey, if he gets work done and what nt, who am I to say something about it?

I've been finding more and more that faster rates in my Jaws and NVDA are helping a lot in real world situations.  For example, I couldn't understand a lick of Spanish early in the course of Spanish III I'm taking currently.  Now that I've increased my speech about 15%, it's really a boost in my comprehension for that class, which is great.  I'm not sure if anyone else has experienced that, but I think it's worth noting.

It's also helped me with my speeches in my Composition/Communications class.  I tend to not stutter as much, and to just go through a fact or statistic with an ungodly unpronounceable name.  I'm not sure if that's just natural skill, or if that's from hearing a fairly quick rate of voice, but either way, it's great.

Best Regards,

Luke

Very good point.

Everyone does work at different speeds, but inefficiency is a term that has near unlimited boundaries.  What I call inefficient, someone else may call fast-paced and runs smoothly for them.

It's just one of those things where it's perception based, and to point out here that I never really declared an issue with having a slower speed, although some may not care to be labeled as such.  So, I can understand where you're coming from.

Best Regards,

Luke

Noted.

I'll cut and edit either Jaws or NVDA reading a narrative of some sort, and use an application that will tell me the time it took.

I've been pronouncing last names incorrectly because I'm use to hearing a screen reader say them.  People always get on me about that because of social media, so I hear the name being said by Jaws and NVDA rather than correctly in person.

Best Regards,

Luke