WebAIM - Web Accessibility In Mind

E-mail List Archives

Thread: clear text image descriptions and alt text with user-created content

for

Number of posts in this thread: 7 (In chronological order)

From: deborah.kaplan
Date: Tue, Mar 20 2012 7:09AM
Subject: clear text image descriptions and alt text with user-created content
No previous message | Next message →

Over the last year, I've noticed that a lot more posters on
social media sites have been putting image descriptions in clear
text under images. I'm pretty sure this started because Tumblr
makes it difficult and sometimes impossible to supply alt text,
and the large number of social justice types on Tumblr were quick
to respond to complaints from their friends who needed the alt
text. But I've certainly seen the practice spread to other social
media: blogs, forums, etc.

I'm all in favor of this -- a lot of times in web accessibility
we assume that the people creating content are web designers, but
in this age of social media, the majority of content most of us
consume is actually created by a slew of amateurs. When authoring
tools fall down on the job, kudos to that slew of amateurs for
providing as much accessible content as is possible. But I have
two questions that come out of this trend:

1. When there is a clear text image description, what should the
alt text on the image be?

2. In contexts (such as social media) where he clear text image
description is welcome or permissible, should be used even if the
authoring tool allows alt (and possibly longerdesc)? I can see
how it encourages good behavior among other content creators in
the social media world, and there are definitely people who
aren't using screenreaders who benefit from captions (such as
magnification users).

-Deborah

From: Jared Smith
Date: Tue, Mar 20 2012 8:24AM
Subject: Re: clear text image descriptions and alt text with user-created content
← Previous message | Next message →

On Tue, Mar 20, 2012 at 7:10 AM, < = EMAIL ADDRESS REMOVED = > wrote:

> 1. When there is a clear text image description, what should the
> alt text on the image be?

If the image description (i.e., caption) of the image conveys the
content of the image, the image should probably have alt="". I like to
teach that every image must have alternative text, but that
alternative text does not necessarily have to be in the alt attribute.
A caption is acceptable. Alt="" conveys that the image does not convey
content or the content is conveyed in nearby content (e.g., the image
caption).

> 2. In contexts (such as social media) where he clear text image
> description is welcome or permissible, should be used even if the
> authoring tool allows alt (and possibly longerdesc)?

I think this depends on the image and the image's context. In some
cases having an image and a text description would be redundant for
sighted users (imagine a photo of a cat with a caption of "cat"). In
this case, the alt attribute would be better. The key is to accurately
and succinctly convey the content while avoiding repetition or
duplication of content.

Jared Smith
WebAIM.org

From: Steve Faulkner
Date: Tue, Mar 20 2012 8:39AM
Subject: Re: clear text image descriptions and alt text with user-created content
← Previous message | Next message →

Hi Jared,
The problem i see with use of alt="" in this context is that if there is a
caption for something then that something needs to be identified.
In time I hope the HTML5 figure/figcaption elementes will provide the
appropaitae semantics for this situation.
In HTML5 it is conforming for an image inside a figure with acaption not to
have an alt attribute, on the presumption that the image is adequately
dscribed by the caption. In this case I would expect user agents such as
screen readers to announce the presence of the image "graphic -
caption:....".

Until such times I have suggested a pattern in examples in HTML5:
Techniques for providing useful text alternatives
http://dev.w3.org/html5/alt-techniques/#sec12
http://dev.w3.org/html5/alt-techniques/#m6

would appreciate your feedback on these.

regards
steve

On 20 March 2012 14:27, Jared Smith < = EMAIL ADDRESS REMOVED = > wrote:

> On Tue, Mar 20, 2012 at 7:10 AM, < = EMAIL ADDRESS REMOVED = > wrote:
>
> > 1. When there is a clear text image description, what should the
> > alt text on the image be?
>
> If the image description (i.e., caption) of the image conveys the
> content of the image, the image should probably have alt="". I like to
> teach that every image must have alternative text, but that
> alternative text does not necessarily have to be in the alt attribute.
> A caption is acceptable. Alt="" conveys that the image does not convey
> content or the content is conveyed in nearby content (e.g., the image
> caption).
>
> > 2. In contexts (such as social media) where he clear text image
> > description is welcome or permissible, should be used even if the
> > authoring tool allows alt (and possibly longerdesc)?
>
> I think this depends on the image and the image's context. In some
> cases having an image and a text description would be redundant for
> sighted users (imagine a photo of a cat with a caption of "cat"). In
> this case, the alt attribute would be better. The key is to accurately
> and succinctly convey the content while avoiding repetition or
> duplication of content.
>
> Jared Smith
> WebAIM.org
>

From: Ryan E. Benson
Date: Tue, Mar 20 2012 11:09AM
Subject: Re: clear text image descriptions and alt text with user-created content
← Previous message | Next message →

Steve,

> Until such times I have suggested a pattern in examples in HTML5:
> Techniques for providing useful text alternatives
> http://dev.w3.org/html5/alt-techniques/#sec12
> http://dev.w3.org/html5/alt-techniques/#m6
I like example C in Alt Techniques #m6, and example B in #sec12. I
wonder if there was a way to tell here is the end of the caption
semantically. It took me a minute to figure out why you put (Photo 1)
at the end of the caption, then I realized by putting the alt text
again at the end, it would signal the end of the caption, such as when
people say start/end quote on the news.

--
Ryan E. Benson



On Tue, Mar 20, 2012 at 10:38 AM, Steve Faulkner
< = EMAIL ADDRESS REMOVED = > wrote:
> Hi Jared,
> The problem i see with use of alt="" in this context is that if there is a
> caption for something then that something needs to be identified.
> In time I hope the HTML5 figure/figcaption elementes will provide the
> appropaitae semantics for this situation.
> In HTML5 it is conforming for an image inside a figure with acaption not to
> have an alt attribute, on the presumption that the image is adequately
> dscribed by the caption. In this case I would expect user agents such as
> screen readers to announce the presence of the image "graphic -
> caption:....".
>
> Until such times I have suggested a pattern in examples in HTML5:
> Techniques for providing useful text alternatives
> http://dev.w3.org/html5/alt-techniques/#sec12
> http://dev.w3.org/html5/alt-techniques/#m6
>
> would appreciate your feedback on these.
>
> regards
> steve
>
> On 20 March 2012 14:27, Jared Smith < = EMAIL ADDRESS REMOVED = > wrote:
>
>> On Tue, Mar 20, 2012 at 7:10 AM,  < = EMAIL ADDRESS REMOVED = > wrote:
>>
>> > 1. When there is a clear text image description, what should the
>> > alt text on the image be?
>>
>> If the image description (i.e., caption) of the image conveys the
>> content of the image, the image should probably have alt="". I like to
>> teach that every image must have alternative text, but that
>> alternative text does not necessarily have to be in the alt attribute.
>> A caption is acceptable. Alt="" conveys that the image does not convey
>> content or the content is conveyed in nearby content (e.g., the image
>> caption).
>>
>> > 2. In contexts (such as social media) where he clear text image
>> > description is welcome or permissible, should be used even if the
>> > authoring tool allows alt (and possibly longerdesc)?
>>
>> I think this depends on the image and the image's context. In some
>> cases having an image and a text description would be redundant for
>> sighted users (imagine a photo of a cat with a caption of "cat"). In
>> this case, the alt attribute would be better. The key is to accurately
>> and succinctly convey the content while avoiding repetition or
>> duplication of content.
>>
>> Jared Smith
>> WebAIM.org
>>

From: Steve Faulkner
Date: Tue, Mar 20 2012 11:21AM
Subject: Re: clear text image descriptions and alt text with user-created content
← Previous message | Next message →

Hi Ryan,

> I wonder if there was a way to tell here is the end of the caption
semantically.

In firefox the figcaption element is mapped to the IA2 accessibility API
caption role, so screen readers can identify to users what caption text is
once they have implemented support.

regards
steve

On 20 March 2012 17:11, Ryan E. Benson < = EMAIL ADDRESS REMOVED = > wrote:

> Steve,
>
> > Until such times I have suggested a pattern in examples in HTML5:
> > Techniques for providing useful text alternatives
> > http://dev.w3.org/html5/alt-techniques/#sec12
> > http://dev.w3.org/html5/alt-techniques/#m6
> I like example C in Alt Techniques #m6, and example B in #sec12. I
> wonder if there was a way to tell here is the end of the caption
> semantically. It took me a minute to figure out why you put (Photo 1)
> at the end of the caption, then I realized by putting the alt text
> again at the end, it would signal the end of the caption, such as when
> people say start/end quote on the news.
>
> --
> Ryan E. Benson
>
>
>
> On Tue, Mar 20, 2012 at 10:38 AM, Steve Faulkner
> < = EMAIL ADDRESS REMOVED = > wrote:
> > Hi Jared,
> > The problem i see with use of alt="" in this context is that if there is
> a
> > caption for something then that something needs to be identified.
> > In time I hope the HTML5 figure/figcaption elementes will provide the
> > appropaitae semantics for this situation.
> > In HTML5 it is conforming for an image inside a figure with acaption not
> to
> > have an alt attribute, on the presumption that the image is adequately
> > dscribed by the caption. In this case I would expect user agents such as
> > screen readers to announce the presence of the image "graphic -
> > caption:....".
> >
> > Until such times I have suggested a pattern in examples in HTML5:
> > Techniques for providing useful text alternatives
> > http://dev.w3.org/html5/alt-techniques/#sec12
> > http://dev.w3.org/html5/alt-techniques/#m6
> >
> > would appreciate your feedback on these.
> >
> > regards
> > steve
> >
> > On 20 March 2012 14:27, Jared Smith < = EMAIL ADDRESS REMOVED = > wrote:
> >
> >> On Tue, Mar 20, 2012 at 7:10 AM, < = EMAIL ADDRESS REMOVED = > wrote:
> >>
> >> > 1. When there is a clear text image description, what should the
> >> > alt text on the image be?
> >>
> >> If the image description (i.e., caption) of the image conveys the
> >> content of the image, the image should probably have alt="". I like to
> >> teach that every image must have alternative text, but that
> >> alternative text does not necessarily have to be in the alt attribute.
> >> A caption is acceptable. Alt="" conveys that the image does not convey
> >> content or the content is conveyed in nearby content (e.g., the image
> >> caption).
> >>
> >> > 2. In contexts (such as social media) where he clear text image
> >> > description is welcome or permissible, should be used even if the
> >> > authoring tool allows alt (and possibly longerdesc)?
> >>
> >> I think this depends on the image and the image's context. In some
> >> cases having an image and a text description would be redundant for
> >> sighted users (imagine a photo of a cat with a caption of "cat"). In
> >> this case, the alt attribute would be better. The key is to accurately
> >> and succinctly convey the content while avoiding repetition or
> >> duplication of content.
> >>
> >> Jared Smith
> >> WebAIM.org
> >>

From: Ryan E. Benson
Date: Tue, Mar 20 2012 11:30AM
Subject: Re: clear text image descriptions and alt text with user-created content
← Previous message | Next message →

Steve,

> In firefox the figcaption element is mapped to the IA2 accessibility API
> caption role, so screen readers can identify to users what caption text is
> once they have implemented support.
Do you know what other browsers are planning to do?

--
Ryan E. Benson



On Tue, Mar 20, 2012 at 1:19 PM, Steve Faulkner
< = EMAIL ADDRESS REMOVED = > wrote:
> Hi Ryan,
>
>> I wonder if there was a way to tell here is the end of the caption
> semantically.
>
> In firefox the figcaption element is mapped to the IA2 accessibility API
> caption role, so screen readers can identify to users what caption text is
> once they have implemented support.
>
> regards
> steve
>
> On 20 March 2012 17:11, Ryan E. Benson < = EMAIL ADDRESS REMOVED = > wrote:
>
>> Steve,
>>
>> > Until such times I have suggested a pattern in examples in HTML5:
>> > Techniques for providing useful text alternatives
>> > http://dev.w3.org/html5/alt-techniques/#sec12
>> > http://dev.w3.org/html5/alt-techniques/#m6
>> I like example C in Alt Techniques #m6, and example B in #sec12.  I
>> wonder if there was a way to tell here is the end of the caption
>> semantically. It took me a minute to figure out why you put (Photo 1)
>> at the end of the caption, then I realized by putting the alt text
>> again at the end, it would signal the end of the caption, such as when
>> people say start/end quote on the news.
>>
>> --
>> Ryan E. Benson
>>
>>
>>
>> On Tue, Mar 20, 2012 at 10:38 AM, Steve Faulkner
>> < = EMAIL ADDRESS REMOVED = > wrote:
>> > Hi Jared,
>> > The problem i see with use of alt="" in this context is that if there is
>> a
>> > caption for something then that something needs to be identified.
>> > In time I hope the HTML5 figure/figcaption elementes will provide the
>> > appropaitae semantics for this situation.
>> > In HTML5 it is conforming for an image inside a figure with acaption not
>> to
>> > have an alt attribute, on the presumption that the image is adequately
>> > dscribed by the caption. In this case I would expect user agents such as
>> > screen readers to announce the presence of the image "graphic -
>> > caption:....".
>> >
>> > Until such times I have suggested a pattern in examples in HTML5:
>> > Techniques for providing useful text alternatives
>> > http://dev.w3.org/html5/alt-techniques/#sec12
>> > http://dev.w3.org/html5/alt-techniques/#m6
>> >
>> > would appreciate your feedback on these.
>> >
>> > regards
>> > steve
>> >
>> > On 20 March 2012 14:27, Jared Smith < = EMAIL ADDRESS REMOVED = > wrote:
>> >
>> >> On Tue, Mar 20, 2012 at 7:10 AM,  < = EMAIL ADDRESS REMOVED = > wrote:
>> >>
>> >> > 1. When there is a clear text image description, what should the
>> >> > alt text on the image be?
>> >>
>> >> If the image description (i.e., caption) of the image conveys the
>> >> content of the image, the image should probably have alt="". I like to
>> >> teach that every image must have alternative text, but that
>> >> alternative text does not necessarily have to be in the alt attribute.
>> >> A caption is acceptable. Alt="" conveys that the image does not convey
>> >> content or the content is conveyed in nearby content (e.g., the image
>> >> caption).
>> >>
>> >> > 2. In contexts (such as social media) where he clear text image
>> >> > description is welcome or permissible, should be used even if the
>> >> > authoring tool allows alt (and possibly longerdesc)?
>> >>
>> >> I think this depends on the image and the image's context. In some
>> >> cases having an image and a text description would be redundant for
>> >> sighted users (imagine a photo of a cat with a caption of "cat"). In
>> >> this case, the alt attribute would be better. The key is to accurately
>> >> and succinctly convey the content while avoiding repetition or
>> >> duplication of content.
>> >>
>> >> Jared Smith
>> >> WebAIM.org
>> >>

From: Steve Faulkner
Date: Tue, Mar 20 2012 11:36AM
Subject: Re: clear text image descriptions and alt text with user-created content
← Previous message | No next message

hi ryan, the hope is that other browsers will implement it in the same way
so that its interoperable.

regards
steve

On 20 March 2012 17:28, Ryan E. Benson < = EMAIL ADDRESS REMOVED = > wrote:

> Steve,
>
> > In firefox the figcaption element is mapped to the IA2 accessibility API
> > caption role, so screen readers can identify to users what caption text
> is
> > once they have implemented support.
> Do you know what other browsers are planning to do?
>
> --
> Ryan E. Benson
>
>
>
> On Tue, Mar 20, 2012 at 1:19 PM, Steve Faulkner
> < = EMAIL ADDRESS REMOVED = > wrote:
> > Hi Ryan,
> >
> >> I wonder if there was a way to tell here is the end of the caption
> > semantically.
> >
> > In firefox the figcaption element is mapped to the IA2 accessibility API
> > caption role, so screen readers can identify to users what caption text
> is
> > once they have implemented support.
> >
> > regards
> > steve
> >
> > On 20 March 2012 17:11, Ryan E. Benson < = EMAIL ADDRESS REMOVED = > wrote:
> >
> >> Steve,
> >>
> >> > Until such times I have suggested a pattern in examples in HTML5:
> >> > Techniques for providing useful text alternatives
> >> > http://dev.w3.org/html5/alt-techniques/#sec12
> >> > http://dev.w3.org/html5/alt-techniques/#m6
> >> I like example C in Alt Techniques #m6, and example B in #sec12. I
> >> wonder if there was a way to tell here is the end of the caption
> >> semantically. It took me a minute to figure out why you put (Photo 1)
> >> at the end of the caption, then I realized by putting the alt text
> >> again at the end, it would signal the end of the caption, such as when
> >> people say start/end quote on the news.
> >>
> >> --
> >> Ryan E. Benson
> >>
> >>
> >>
> >> On Tue, Mar 20, 2012 at 10:38 AM, Steve Faulkner
> >> < = EMAIL ADDRESS REMOVED = > wrote:
> >> > Hi Jared,
> >> > The problem i see with use of alt="" in this context is that if there
> is
> >> a
> >> > caption for something then that something needs to be identified.
> >> > In time I hope the HTML5 figure/figcaption elementes will provide the
> >> > appropaitae semantics for this situation.
> >> > In HTML5 it is conforming for an image inside a figure with acaption
> not
> >> to
> >> > have an alt attribute, on the presumption that the image is adequately
> >> > dscribed by the caption. In this case I would expect user agents such
> as
> >> > screen readers to announce the presence of the image "graphic -
> >> > caption:....".
> >> >
> >> > Until such times I have suggested a pattern in examples in HTML5:
> >> > Techniques for providing useful text alternatives
> >> > http://dev.w3.org/html5/alt-techniques/#sec12
> >> > http://dev.w3.org/html5/alt-techniques/#m6
> >> >
> >> > would appreciate your feedback on these.
> >> >
> >> > regards
> >> > steve
> >> >
> >> > On 20 March 2012 14:27, Jared Smith < = EMAIL ADDRESS REMOVED = > wrote:
> >> >
> >> >> On Tue, Mar 20, 2012 at 7:10 AM, < = EMAIL ADDRESS REMOVED = >
> wrote:
> >> >>
> >> >> > 1. When there is a clear text image description, what should the
> >> >> > alt text on the image be?
> >> >>
> >> >> If the image description (i.e., caption) of the image conveys the
> >> >> content of the image, the image should probably have alt="". I like
> to
> >> >> teach that every image must have alternative text, but that
> >> >> alternative text does not necessarily have to be in the alt
> attribute.
> >> >> A caption is acceptable. Alt="" conveys that the image does not
> convey
> >> >> content or the content is conveyed in nearby content (e.g., the image
> >> >> caption).
> >> >>
> >> >> > 2. In contexts (such as social media) where he clear text image
> >> >> > description is welcome or permissible, should be used even if the
> >> >> > authoring tool allows alt (and possibly longerdesc)?
> >> >>
> >> >> I think this depends on the image and the image's context. In some
> >> >> cases having an image and a text description would be redundant for
> >> >> sighted users (imagine a photo of a cat with a caption of "cat"). In
> >> >> this case, the alt attribute would be better. The key is to
> accurately
> >> >> and succinctly convey the content while avoiding repetition or
> >> >> duplication of content.
> >> >>
> >> >> Jared Smith
> >> >> WebAIM.org
> >> >>