E-mail List Archives
Thread: Dragon NaturallySpeaking
Number of posts in this thread: 7 (In chronological order)
From: Seth Kane
Date: Wed, Sep 16 2009 12:55PM
Subject: Dragon NaturallySpeaking
No previous message | Next message →
Does anyone use Dragon NaturallySpeaking during testing or do you primarily stick to Screen Readers only?
- Seth
From: J. B-Vincent
Date: Wed, Sep 16 2009 1:05PM
Subject: Re: Dragon NaturallySpeaking
← Previous message | Next message →
I'm occasionally asked to test with NaturallySpeaking. There's some overlap with screen reader testing (e.g., inaccessible event handlers won't respond to either JAWS or NatSpeak commands).
--Jane Vincent, Center for Accessible Technology
--- On Wed, 9/16/09, Seth Kane < = EMAIL ADDRESS REMOVED = > wrote:
From: Seth Kane < = EMAIL ADDRESS REMOVED = >
Subject: [WebAIM] Dragon NaturallySpeaking
To: " = EMAIL ADDRESS REMOVED = " < = EMAIL ADDRESS REMOVED = >
Date: Wednesday, September 16, 2009, 11:55 AM
Does anyone use Dragon NaturallySpeaking during testing or do you primarily stick to Screen Readers only?
- Seth
From: Julie Romanowski
Date: Wed, Sep 16 2009 1:15PM
Subject: Re: Dragon NaturallySpeaking
← Previous message | Next message →
Our team tests with Dragon and JAWS (screen reader). A screen reader
will catch accessibility issues that may affect blind/low vision users,
but will not catch many issues that voice recognition users may
encounter.
From: Moore,Michael
Date: Wed, Sep 16 2009 2:20PM
Subject: Re: Dragon NaturallySpeaking
← Previous message | Next message →
We also test mission critical applications using Dragon and the text to speech functionality of ZoomText to make sure that all of our users can access these applications. We have frequently found that code that meets accessibility and html standards will still not work properly with Dragon and ZT. Duplicating label content with the title attribute resolves most of these issues.
Mike Moore
(512) 424-4159
From: Steve Green
Date: Wed, Sep 16 2009 5:00PM
Subject: Re: Dragon NaturallySpeaking
← Previous message | Next message →
Yes, we test with NaturallySpeaking, and it invariably throws up a bunch of
issues that were not found through technical testing or testing with other
assistive technologies.
An increasing problem is its inability to interact with Flash content.
NaturallySpeaking could interact via the Flash API but Nuance seem to have
chosen not to bother. This leaves the mousegrid as the only means of
interaction, which is entirely unsatisfactory. Sometimes you may be able to
use voice commands to tab through the links but that is not reliable. This
problem is exacerbated by the fact that users usually have no idea what
technology is being used - a Flash movie often looks like text and images,
yet it fails to respond to the voice commands for those types of content.
On pages with more than one vertical scrollbar users sometimes have
difficulty controlling the one they want, not realising that this may be
determined by the position of the cursor or the place the mouse was last
clicked.
With forms we often encounter problems where NaturallySpeaking enters
prohibited characters such as currency symbols or commas in large numbers.
For instance, saying "two thousand pounds" will result in £2,000 or even
£2,000.00 being entered. These symbols can cause data validation errors, and
I have seen users totally baffled as to why an apparently valid number is
not being accepted.
Note that a designer or tester using NaturallySpeaking is unlikely to
identify many of these issues because they know too much about the
technologies and coding. You really need to test with real users and look at
the strategies they use for interacting with the content.
Steve Green
Director
Test Partners Ltd
From: deblist
Date: Wed, Sep 16 2009 6:35PM
Subject: Re: Dragon NaturallySpeaking
← Previous message | Next message →
On Wed, 16 Sep 2009, Seth Kane wrote:
> Does anyone use Dragon NaturallySpeaking during testing or do you primarily stick to Screen Readers only?
I use Dragon when testing. I also use it when coding, sending
email, or watching YouTube videos. *waves disabled developer
hand at you*
And I beg everyone to test with Dragon -- it has different issues
from screen readers, and plenty of screen reader compliant pages
are unnavigable with Dragon. Fewer these days, but it still
happens.
-deborah
From: Caleb Tang
Date: Thu, Sep 17 2009 7:00AM
Subject: Re: Dragon NaturallySpeaking
← Previous message | No next message
Dragon is in our standard user testing list and we often found it don't work
well on heavy CSS and JavaScript driven sites.
> On pages with more than one vertical scrollbar users sometimes have
> difficulty controlling the one they want, not realising that this may be
> determined by the position of the cursor or the place the mouse was last
> clicked.
Dragon uses JavaScript to insert HTML/CSS into pages when command such as
Link, Image, Edit box etc were called. By default Dragon will only recognise
links in first 200 HTML elements of the page (Although this can be changed
on the user files). Dragon may be slow in response on pages with more HTML
elements.
We recently starting to explore Windows Speech Recognition in Windows Vista
and it works pretty well with websites that doesn't on Dragon.
Caleb