Fixed Layout Accessibility
Fixed-layout EPUB accessibility is one of the challenges the EPUB WG intends to take on during it’s charter period. Unlike much of the other work of the WG, this particular effort is mainly about producing guidance for content creators and reading systems on how to produce accessible documents using the tools provided in the existing specifications. There are no plans to change the EPUB specification to add new requirements for fixed-layout documents.
This wiki contains the main proposals gathering during the latter half of 2020 through discussions at TPAC and with experts in the field of content production. These proposals should be considered a starting point to thinking about the challenges of FXL accessibility, and encourage questions, experiments, and discussion.
If you are interested in this topic, or in any of the proposals, please feel free to log issues in the repository, or contact the working group via email.
The EPUB specification already encourages and supports the use of SVG for images and diagrams, as SVGs are more accessible and produce higher-quality images at various screen sizes. SVG is not just an image format, however, and has many functional uses for text layout.
One of the biggest reasons to use fixed-layout as a format is to recreate “pixel-perfect” layouts for complicated content. SVG allows for this precision placement of text, while also exposing the text to AT, as long as it’s marked up correctly.
Mixed modalities is a clumsy name for mixing fixed-layout and reflowable sections within the same EPUB file. The EPUB 3 specification allows for mixed modalities in the spine of an EPUB. It is an under-implemented feature, both in reading systems and production pipelines. We believe mixed modalities deserves a second look as a good way to provide an accessible experience for content that has a lot of text content interspersed with images, diagrams, or other fixed content. A good example for this method might be a cookbook, where a fixed page can be used for a full-bleed image of the dish, and reflowable text for the recipe.
This third proposed method is a bit more technically complicated, and does not rely on existing EPUB features. It was proposed when looking at some of the most complicated content to make accessible, image-reliant formats like manga, comics, or graphic novels. Live text in these formats is tricky, as the “font” of a comic might be the artist’s writing, or part of the image itself. In this method, an accessibility layer would be added, creating a method for “captioning” (displaying live text that is visible to AT and the user directly), or simply providing a textual source for AT that’s not visible on the page (where the text would be displayed in its original form).
Sections to be written:
- legibility (Wendy)
- media overlays (Ken)
- considerations for different content types (i.e. text books, cookbooks, children's books, etc.)
Sample Content to be created:
- comic w/ image map
- textbook
- cookbook
- kids book
- print replica novel/coffee table book
- poetry
- tour book
- magazines