Page 1 of 1

Screen readers and the <figure> and <figcaption> tags

Posted: Mon Mar 25, 2019 5:45 am
by Daniella.LP

One important piece to consider regarding EPUB accessibility is how HTML 5 is being used--or not being used--by assistive technologies. Do JAWS, NVDA, Narrator, VoiceOver and Chrome support the <figure> and <figcaption> tags? We should test this.

In the EPUB Testing folder in Dropbox you will find a file titled "figcaption-test.html". This file contains alternative text in addition to a caption. please test this out with your screen readers and note how they treat these tags, for example if they don't read them. In short, what we need to know is whether the screen reader read the text that is in a <figcaption> tag.

Please post your replies here, and note which screen reader (including version number) and which browser you used to test this.

Many thanks!

Re: Screen readers and the <figure> and <figcaption> tags

Posted: Mon Mar 25, 2019 8:14 am
by Danny
Good morning Daniella and team,

I just tested this on Google Chrome Version 72.0.3626.121 (Official Build) (64-bit) with NVDA Version 2018.3.2. NVDA read both the Alt text (describing the two houses with their paintings) along with the figure caption (discussing the famous site). In addition, I was able to review the Alt text by character, word and line. I could not get Narrator to read anything in Chrome.

I then performed the same test on Microsoft Edge. Here, NVDA and Narrator read both the Alt text and figure caption, but it was not possible to review the Alt text by word or character, though it was possible to review the heading at the top and the caption at the bottom in this way.

Finally, I opened the FigCaption test in Internet Explorer. NVDA performed the same here as it did in Chrome, reading both the alt text and figure caption, and allowing review of both elements by character, word and line. Again I couldn't get Narrator to properly review the page, however both the Alt text and figure caption were readable with the mouse cursor.

Top of the morning to all!

Re: Screen readers and the <figure> and <figcaption> tags

Posted: Tue Mar 26, 2019 1:17 pm
by Karoline
Good day everyone!
I just opened it in IOS. I am using IOS 12.4.1. So far, it has told me about the houses. The caption is Trulli in Puglia. It seems to read the text just fine unless I am missing something. It looks like an HTML document?

Re: Screen readers and the <figure> and <figcaption> tags

Posted: Tue Mar 26, 2019 1:36 pm
by Karoline
I took a look at this in Chrome Version 73.0.3683.86 (Official Build) (64-bit)
and JAWS 2019 1903.47 ILM. Both are the latest versions. I was able to hear exactly the same information that I did with the IOS test. The description of the houses and the caption.

Re: Screen readers and the <figure> and <figcaption> tags

Posted: Tue Mar 26, 2019 8:04 pm
I'm using NVDA 2019.1 and Firefox 66.0.1 and it reads the image like alt-text followed by the caption. as regular text in a document. According to
it looks like the figure and figcaption attributes are supported properly by Chrome/opera, Edge, Firefox, and Safari. The only webbrowser that's not supported is Internet Explorer. When I mean properly supported, I mean not only on the web browser side but the screen reader side as well with the accessibility API.

Re: Screen readers and the <figure> and <figcaption> tags

Posted: Sun Mar 31, 2019 8:34 pm
by Simon
So far, I have tested this with common Windows browsers using NVDA and narrator, as well as Safari on Mac OS and iOS using VoiceOver.

NVDA was able to communicate with every Windows browser and see both the alt-text and the caption. These included Chrome, Edge, Firefox, and even Internet Explorer.
Narrator inconsistently read the text in Microsoft Edge, and did not read anything at all in other browsers. I was once able to make it read the entire page including alt-text and caption, but could not review it by character.
The good news is that Windows insider builds are steadily improving Narrator's support for Chrome and other browsers, so I expect this to change soon.

With Safari on iOS, VoiceOver reads all page content.

With Safari on Mac OS, VoiceOver reads all page content, and indicates the start and end of the figure, which is not something I've observed in other browsers.

Overall, I would say the support for this is quite good. I have yet to test it in third-party browsers for the Mac. Third-party iOS browsers ust use the same gendering engine as Safari, so I expect no change but will still try.