Most PDF documents have an embedded "text layer" that tells Perusall where the text is on the page and what its contents are. This text layer enables students to highlight text directly on the page, and it also enables Perusall's built-in read-aloud feature to work. Having a text layer is essential for accessibility, as it also enables screen readers to access the text within the document.
If you upload a PDF document to your Perusall Library that does not contain any embedded text layer, you will see an accessibility warning when the document properties are visible in the Library.
When you receive this message, we recommend removing the PDF from your library, running the PDF through an OCR (Optical Character Recognition) application to add a text layer, and then re-uploading the PDF to Perusall. You can also optionally enable Perusall's built-in automatic rotation for all uploaded documents within Settings > Advanced (note this will make PDF-based uploads slower as the OCR process is completed for each page).
Because of the way PDFs are rendered in Perusall, images in PDFs won't have "alt text" associated with them. To provide context for students who may need to use a screen reader, you can use Perusall's figure annotation tool to drop a pin somewhere on the image, with a comment that describes the contents of the image. This also makes it easier for such student using a screen reader to effectively annotate the image, as they can simply respond to the comment to engage in a conversation about that image.
Related to