Tweak understanding for 1.2.3 and 1.2.5 #1790
Conversation
* turn the "During existing pauses..." sentence around, so as not to suggest that an exception exists * add mention of audio ducking (quack) * add a note for 1.2.3 that explicitly says that a lack of gaps is not an excuse/exemption
|
Any news on this @alastc ? |
|
any chance this could be considered/discussed at some point? |
|
IMO, this change to include audio ducking risks over-reaching the current guidance and requirements of 1.2.5. I am moving this PR back to the Drafted project column, until we have time to address that Response. |
Co-authored-by: Mike Gower <mikegower@gmail.com>
|
I will revisit this, as I may have run head-first into changes - particularly for the 1.2.3 case. Will reconsider. |
|
After discussion in the WCAG 2.x TF call, and mulling this over further, I'd suggest reworking my PR here to:
|
patrickhlauke
force-pushed
the
patrickhlauke-understanding-audiodescription
branch
from
2616bd9
to
3103888
Compare
|
That's an interesting logic - I have so far assumed that anything that passes 1.2.5 (even because there are no gaps to add AD without extending the video) would automatically pass 1.2.3 - but the case can be made that it bounces back to 1.2.3. |
|
@bruce-usab wrote above:
Can you explain where the normative text leads to that requirement? If there is support for the approach that if there are no gaps and 1.2.5 must therefore pass, but 1.2.3 would fail because a transcription would then have to be the place to include the missing visuals (say, to mention frequent cases, text insertions indicating speaker, or initial or final text "slides"), a Failure bridging 1.2.5 and 1.2.3 might then describe exactly that situation. |
simple syntax fix
Supported by TF, after discussion
|
@patrickhlauke I have incorporated the changes reviewed and supported on Friday's call. My inclination is to include this is in this week's set of changes, but wanted to check with you first. |
@detlevhfischer it follows from the definition of audio description and how audio description is traditionally/historically applied/implemented in other media (broadcast TV, VHS, DVD, and now streaming). Newscasts will often duck noisy background audio when the reporting is talking. The same thing could happen for audio description (narration) of what is causing the noisy background audio. If a video has arbitrary music playing over non-speech informative video (could be title cards or photo montage) then I would argue that is a failure against 1.2.5 because audio description could/should be provided by ducking the non-speech background audio. Edit to add, from definition Note 2, emphasis added:
If the video includes non-stop talking -- there are no gaps in dialog. So that means the video meets 1.2.5. There might be gaps in dialog but not gaps the audio track. Lack of gaps in the audio track does not automatically excuse lack of audio description. Gaps in the dialog where the audio track is pure decoration (i.e., not conveying information and not substantive content) are opportunities to add voice-over narration (i.e., audio description), as required by 1.2.5. Those gaps in dialog might be too short to be of any utility, but often they are long enough for substantive voice-over. |
|
I'd hold off sending this ... I've ended up with no time to look at this again, and I just want to make sure it properly reflects where we got to... |
✅ Deploy Preview for wcag2 ready!
To edit notification comments on pull requests, go to your Netlify site configuration. |
|
Made some further tweaks/changes, and think this is now ready to be looked over again by the WCAG 2.x group at least. On the parts that @mbgower had last added, one change I made was removing this
mainly, because the AAA one seems a red herring to me. IF content satisfies the AAA extended AD requirement, then by definition it would already have actual AD, so then the AA requirement would be satisfied and that whole sentence/bit would be pointless? Anyway, let's discuss this further and see if this now captures the essence of where we landed. (I also did a big cleanup of the indentation / line breaks on the two understanding documents, as they had some ridiculously inconsistent and wide spacing) |
|
A long discussion ensued today on the WCAG 2.x backlog meeting, going back to one of the original points that brought about this PR. It was suggested on the call that this point may be valuable to be brought up to AGWG for further discussion /cc @bruce-usab @mbgower As said in the call, I'd suggest the survey for AGWG should be carefully worded to just focus on the core question/disagreement, which I think is on the lines of (please feel free to edit as needed):
|
|
Further, @bruce-usab raised a concern that too tight a language for 1.2.5 would risk failing cases that before were deemed ok, like talking heads videos where there's some on-screen info about the speaker, but no AD and no pauses for it to be added. @mbgower pointed out that there's a technique here that covers this exact scenario, G203 - but I'd like to see this scenario acknowledged even in the understanding itself (rather than relying on authors/auditors having to find it by looking through techniques). I will make another stab at including this aspect (which boils down to the squishiness of "what IS important information", in essence) in the understanding for 1.2.5. |
|
@mbgower @patrickhlauke would you be supportive of a new general Failure Technique (for 1.2.5) where the audio track, despite being heavy on narration (or even non-stop dialog) is insufficient for it to be characterized as audio description? Edit to mention that I think a new Failure Technique is a better approach than a carefully worded survey. I agree that Patrick outline boils down to the two possible and logical mutually exclusive options. |
|
@bruce-usab just want to check what you mean by
If the narration doesn't sufficiently cover important visual information, it can't be characterised as being audio description, as it doesn't describe the visual. or am i misunderstanding what you mean here? (because of course, if the narration already covers important visual info, then that obviates the need for AD in the first place) or do you mean by the above, in essence
? |
|
I don’t see how G203 can be used to meet 1.2.5 as it’s not synchronized. |
|
@mraccess77 I think the point that G203 tries to make (but possibly muddies the argument a bit, which should really be made clearer directly in the understanding) is that some videos do have additional visual information, but that it's not necessarily information that needs to be synchronised, as it's relatively static throughout the video. It provides a bit of leeway to say that it's equivalent in that case to have AD, or to make sure the (non-synchronised) visual info is provided in another way as an alternative. Agree though that the argument here seems to be more along the lines of providing an accessible alternative to the visual part of the video. But it comes down a bit to the idea that the "artifact" that is being evaluated usually is the page as a whole, rather than the pure video in isolation, so it's about the end result of whether or not a user with low/no vision can still get the full information from the video in combination with other content in the page itself. @bruce-usab would you agree/be happy with some attempt to include clarification in the actual understanding (for 1.2.3 and 1.2.5) that tries to clarify this aspect in some way, and would that inclusion alleviate your concerns here about not all of a sudden failing these sorts of videos? |
|
I believe the intent of the audio description requirement was when pauses were available in the video. If there are not pauses then audio description that inserts description into pauses isn't an option and a video could pass. However, the exception I believe is at the beginning and end of videos as I since those are opened ended that description audio would be needed for those situation because there is effectively a pause there. This technique seems to muddy the water by allowing for other non-synchronized page content to be used for the beginning and end description and also implying that AD was required for other portions where there was not pauses but that non-synchronized description could be used to conform. Since AD must be synchronized I don't see how we can say this would meet the requirement. I think at best we are saying AD wouldn't apply and it's advisory to use this technique to provide non-synchronized information that was missed because AD wouldn't fit. |
Yes.
I can't say until I read the wording. I agree that that an insufficient audio track is an a11y barrier. I am of the opinion that if the audio track is non-stop talking, then 1.2.5 is satisfied. Understanding might call out that a barrier remains, without going so far as to say that an SC is failed. (I believe we have had to do a similar dance in a few other Understanding articles.) |
Sure it can. Many people expect that the presence of audio description is unambiguous, that there is a human narrator with a distinct voice added to the soundtrack. Note 3 from the definition makes it explicit this is not a requirement under WCAG. I agree with you that squaring Understanding 1.2.5 with G203 is tricky. G203 also makes my idea for a 1.2.5 Failure Technique (that an inadequate audio track cannot count as Audio Description) problematic. IMHO, if we want to say that a non-stop talking head is not necessarily enough to meet 1.2.5, that detail needs to part of a Failure Technique, not as one sentence in Understanding. |
but i think this is yet again where we disagree again on the exact same point? I (and most others on the call) are of the opinion that it doesn't satisfy the SC as is - an absence of pauses for AD does not mean it passes the SC that requires AD. you even agreed in the chat during the call when I explicitly asked along the lines of "so an absence of pauses is not an excuse to automatically pass 1.2.5"? the further discussion since last week was about seeing if this can be softened to make concessions for cases where somewhere else in the page at least some basic (non-synchronised) information is present that would at least obviate the need for the AD (like not having AD to read out the speaker name would NOT fail if that non-synchronised info is present on the page itself where the video is embedded in). i thought we were finally on the same page, after our call today, but this seems to go right back to the fundamental disagreement? this seems to flip-flop between calls, unless we're talking across purposes? |
|
I walked the task force through some of the decision tree I began creating for this. It might help focus some discuss and thoughts. GIven the definition of "audio description" is:
it becomes a challenge to figure out what to do with a video where it is adequately narrated by the creator (i.e., no audio descriptions were added) in regard to whether to mark the video as N/A or "meets". As Bruce points out, the not-terrifically worded third note makes the intended outcome clear, just not how a user reports on it. It's the use of the word "additional" which I find frustrating here, given the definition!! Strip away the word "additional" here -- or even "audio", and it becomes more consistent: "...no additional description is necessary." (and in fact makes it state the same thing as Note 1 of the Understanding documents).
I've tried to resolve this in the decision tree. |
|
@mbgower but that's not the crux of what @bruce-usab seems to be saying, the "not applicable vs pass" is a separate tangential discussion. unless i'm misunderstanding. we've been talking about a video that has non-stop audio/dialogue but that also DOESN'T convey something visual in all that existing audio which is important. it's not about whether or not it's "not applicable", it's a fairly direct "this video has visual information that, if i can't see it, is not conveyed in any other way". IF the non-stop audio/dialogue IS conveying everything visual that's important already, sure, we can argue over whether that's a "pass" or "not applicable", but that is not the case we seem to be constantly deciding on, then backing away from ... a video that clearly has visual aspects that are currently not conveyed in the existing audio/dialogue already, but that also has no pauses where an author could have added extra AD ... does it get a free pass for the lack of pauses or not? to try and square that circle, we last discussed situations where it may be deemed not to fail IF that extra bit of information does not need to be synchronised, and is provided at least somewhere else in the same page/sample (so that, overall, somebody who can't see the video can still get all the information they need - not purely from the video, but from the video in the context of the page that it's in)... anyway, I'll go ahead and try to incorporate this thought into this PR for now and see how that shakes out. but frankly, i'm beginning to wonder if we're all talking about different things/problems here all of a sudden... |
|
Whether is not applicable or meets may not matter as the understanding conformance documents states "Note This means that if there is no content to which a success criterion applies, the success criterion is satisfied." The wording of the SC "Audio description is provided for all prerecorded video content in synchronized media." is problematic as it seems assume it is always needed. |
so that's a separate concern to the one I've been tackling here in the PR. we can try to bash at this further as part of this PR as well, or separate that out as a related, but distinct, concern |
<blockquote>for clarityCloses #1768
Preview | Diff