Screen-reader pronunciation quirks
Hi all
I recently read a document from the RNIB where they suggest using the phrase ‘Skip to main content’ rather than just ‘Skip to Content”. Why? Apparently screen readers can mispronounce the word ‘content’ by putting the emphasis on the second syllable. Test is in JAWS and you’ll see they are right. I presume that since it is only a short phrase it has little context for the words so it guesses incorrectly.
This got me wondering whether anyone has found similar quirks during their browsing or testing? Maybe even someone has built a resource we can all reference? Or may be I should do it? Perhaps such a resource might be useful as a starting point for those who can’t afford screen readers?
One other: I remember a few years ago that some speech synthesis software I was using would pronounce ‘e-mail’ correctly but not ‘email’ (quite rightly I’d say). It seems that JAWS does not have this problem (though I’m sure it has said ‘e dash mail’ to me!)
Anyone got any more? Any language accepted though beyond English I’m a typical useless Brit
Cheers
Kevin
Share: 
I don't think we should code/write for buggy software (with the exception of some adaptations for IE/Win due to its inexplicable ubiquity). If JAWS pronounces 'content' wrong in this context, it's up to Freedom Scientific to fix the bug. I'm quite sure that JAWS users are smart enough to understand what we're saying anyway.
When I tried pwWebSpeak, I found that it doesn't like things like using ’ instead of a proper apostrophe. It doesn't know how to pronounce it, so it makes a short pause. Contractions like "it's" would then be read as "it<short_pause>s". Well ... you should have heard it try to pronounce 'Åsnorrbodarna' ...
I saw a demonstration where a blind guy showed us how he used his screen reader. It was Window-Eyes, and the program was an English version, but it had a Swedish dictionary/library. Thus it pronounced everything as if it were in Swedish – including the English phrases that Window-Eyes sent it!
It took us a while to understand the gobbledygook, but I finally realised that it was saying 'page loading'. It pronounced it something like 'pah-geh loo-ah-ding'.
Tommy has left the building
Share:
Gez has a pretty comprehensive chart at Juicy Studio that catalogs the various assistive device quirks. May not be quite what you're looking for, but it's a good start …
Share:
| Nathan wrote: |
| Gez has a pretty comprehensive chart at... |
Thanks. I'll take a look.
Share:
Tommy, you've brightened my day already. Thanks!
JAWS can do the language swtiching thing. Hours of fun
And I'm pretty sure I couldn't pronounce Åsnorrbodarna (though I can type it).
Back to the subject...
| TOOLman wrote: |
| I don't think we should code/write for buggy software |
Is this really writing for buggy software? I'd see it as being quite helpful. Perhaps it should be classified under the 'until user agents...' category?
Maybe this is the only example but it was worth asking.
Cheers
Kevin
Share:
Don't worry, Kevin. Few native Swedes could pronounce Åsnorrbodarna correctly.
What I mean is, we should be careful trying to work around quirks in software. I wouldn't consider this particular thing to be a bug. What if we do a workaround, and then they fix the issue and our workaround stops working instead?
Also, let's not underestimate the intelligence of our visitors. (We shouldn't overestimate it either, but that's another issue.
)Tommy has left the building
Share:
I agree with Tommy. The onus is on the user(agent), to coin a phrase. It needs to learn how to speak better, just like software like Dragon Naturally Speaking need to learn how to listen better. For crying out loud, if you're paying £600 for JAWS, you are in a position to be able to demand improvements.
On the flipside, JAWS users know how it says "content"... they are they experts, and we shouldn't interfere.
Kajun
Share:
| Kajun wrote: |
| The onus is on the user(agent), to coin a phrase. |
That's more like counterfeiting ...
Tommy has left the building
Share: