VP-76: When keyboard focus lands on video, screen readers don't say reasonable things

Description

Jaws+FF:
When focus lands on the video, Jaws says "Use Jaws key plus alt plus R to read the script of text." Note that pressing Ins+Alt+R does nothing.

NVDA+FF:
When focus lands on the video, NVDA says "application grouping section."

VoiceOver+Safari:
When focus lands on the video, VoiceOver says "group."

Jaws+IE9:
When focus lands on the video (and Jaws is in application mode), Jaws says "video"

NVDA+IE9:
When focus lands on the video, NVDA speaks everything in it that's text: the time codes, the content of the language menus, the caption, etc (regardless of whether NVDA is in focus or browse mode)

Environments

NVDA, Jaws and VoiceOver with FF and Safari and IE9

Comments

• Alexey Novak commented 2012-08-21T13:24:52.825-0400

  in my https://github.com/anvk/videoPlayer/tree/FLUID-4748 aria-label is been added so that at at least Safari and FF have some kinda proper response from screen readers.

• Alexey Novak commented 2012-08-21T14:16:34.900-0400

  Safari + VoiceOver = says "Infusion Video Player with 4 items. Application."

  FF + NVDA = says "Infusion Video Player grouping. Section."
  IE + NVDA = reads out the whole content of the videoPlayer. If you tab backwards instead of reading the whole content of a videoPlayer a screen reader reads "video grouping"
  Chrome + NVDA = says "blank". Even if you add aria-label to all div's including <video> tag Chrome refuses to read out the label for a videoPlayer

  FF + JAWS = "group box". Adding aria-label to any of the div's would not change anything.
  Chrome + JAWS = "Insert F1 L. Video player button". No idea where this text even comes from
  IE + JAWS = silence

• Michelle D'Souza commented 2012-10-05T11:50:26.022-0400

  We discussed this pull request in the channel and decided that we should try for a different solution:
  http://wiki.fluidproject.org/display/fluid/fluid-work+IRC+Logs-2012-10-05

  It would be a better user experience for the user if the title of the video is read out when the video container gets focus.
  Let's attempt to put a title on the video element and use labelled by to have it read.

• Anastasia Cheetham commented 2012-10-10T16:34:53.642-0400

  Notes: both a 'title' attribute and an 'aria-label' attribute on the video element will be read by Jaws and NVDA (not VoiceOver). The 'title' attribute is read twice, for some reason. Because of this, I feel the 'aria-label' is a better choice.

  Another note: the aria-label is read, but nothing is spoken that actually indicates that this is a video. I recommend prefacing the title with something like "Video: " so that you'd here "Video: Mammals" for example.

• Michelle D'Souza commented 2012-10-22T10:37:59.508-0400

  Merged into project repo at 436e45412c61e0ed25152d74970643545d64cc13

• Ravish Tiwari commented 2014-12-10T06:22:19.743-0500

  it seems that this issue is still not resolved.
  this can verified on Fluid Player demo site: http://build.fluidproject.org/videoPlayer/demos/Mammals.html
  With IE, when Keyboard focus lands on Video Player, it says:
  grouping, video grouping and completely ignores aria-label provided.

  verified using:

  • Firefox Developer Edition : 36.0a2 and NVDA 2014.4 : Working fine
  • IE 11 and NVDA 2014.4 : reading grouping, video grouping and completely ignores aria-label provided.
  • Chrome 39 and and NVDA 2014.4 : Working fine