search

w3c / webvtt

WebVTT Standard
https://w3c.github.io/webvtt/
Other
103 stars 40 forks source link

Users of magnification may have overlap issues (From Bugzilla) #426

Open johnfoliot opened 6 years ago

johnfoliot commented 6 years ago

MIGRATED FROM BUGZILLA: https://www.w3.org/Bugs/Public/show_bug.cgi?id=28509

Concern: the concern is likely that if text is explicitly positioned on certain lines, there is potential that enlarging the text may lead to text growing outside the rendering area (e.g. text positioned in the top-most line), or successive lines may grow into each other.

Reply: First we have to understand that when content is enlarged, the video is typically enlarged together with the text. Therefore, the occurrance of this problem is rare. Secondly, there is a conflict between authoring requirements and rendering limitations. The browser rendering algorithm will [adjust?] as well as it can, but once text breaks out of the video rendering boundaries, there is not much it can do.

APA Response: APA recognizes the technical constraints noted here with regard to rendering limitations. Authoring Guidance recommendations should nonetheless indicate the potential of this problem, and urge content authors to strive to have captions (etc.) be no greater than 50% of the default width of the viewport (which would allow for a text increase of roughly 200% without clipping). APA notes that for low-vision users, even at full-screen, those users may still need to enlarge the caption text to meet their reading needs.

Concern: the concern is likely that the display area of captions is limited to the background area of the video element it is rendered onto and that with magnification the captions may go outside this rendering area.

Reply: The area outside the video element is no usable to render captions onto (think about full-screen mode, or if the video is on a Web page there is other content around the video element). Therefore, after having done all it can to try and retain visibility of all caption text, the browser will hit the limit of what it can do.

APA Response: APA again recognizes the technical constraints noted here with regard to rendering limitations. We once again recommend good authoring guidance to ensure that content authors are aware of the potential issue raised, so that authoring decisions regarding line-lengths and amount of caption text rendered on screen at any single instance can be made with this knowledge.

Concern: the concern is likely about what happens when the text is zoomed, but the video isn't.

Reply: If there are tools that do this, then you will hit the issues of overlapping text and disappearing text when hitting the boundaries of the rendering area faster than normal. However, it is unlikely that a tool would exist that zooms just the text and not the video element on screen. Normally, all content on a Web page is zoomed when magnification or zooming tools are in use.

APA Response: It was APA's understanding that one of the benefits of WebVTT was that it could be further styled by the content author using CSS. User stylesheets today provide the ability for users to modify and enlarge onscreen text, and tools and browser extensions exist today to accomplish this task.

The presumption that video content would be zoomed to the same level of caption text is, from APA's perspective, unfounded and incorrect, and the emergent WCAG 2.1 specifically will have a new Success Criteria (Success Criterion 1.4.12 Text spacing - http://rawgit.com/w3c/wcag21/master/guidelines/index.html#text-spacing) which currently notes that caption files (when supplied as stand-alone time-stamped documents) are covered by this SC. Please also see: https://rawgit.com/w3c/wcag21/text-spacing/understanding/21/text-spacing.html

Normally, all content on a Web page is zoomed when magnification or zooming tools are in use.

Respectfully, this is factually incorrect. Browser-based zoom traditionally operates like this, however other Assistive Technology tools may only zoom a part of the whole web page, or may only apply zoom to text (and/but not imagery). Some user-agents and platforms also allow for end-user font magnification (for example, on the Android platform, individual users can choose from different default font sizes, that are applied to all on-screen content.

APA again recognizes the technical limitations noted here with regard to rendering limitations, and strongly recommends that appropriate authoring guidance to address all 3 related issues noted here be included (directly) in the Recommendation.

silviapfeiffer commented 6 years ago

Thanks for going through those old issues. If I understand correctly, you'd like me to add authoring guidance around magnification and its potential impact.

Could it be done with a single paragraph such as: "Authors of WebVTT caption and subtitle files are encouraged to test the rendering of their files using Assistive Technology. Specifically, the use of magnification tools by low-vision users may require editing their files to allow for sufficient space within cue lines and cue line heights. It is recommended to allow for a 200% magnification without clipping."