Estimated reading time: 2 min (426 words)
A searchable document contains text that can be read by text-to-speech programs (e.g., Read&Write and Natural Reader) or by screen readers (e.g., JAWS and NVDA) for blind/visually impaired readers. Create documents with searchable text.
A scan is an image of a page therefore scanned PDFs are the most commonly inaccessible files. What a human reads as "text" are patterns of light and dark pixels. For a scanned PDF to be accessible, it must be put through an Optical Character Recognition (OCR) process, which analyzes those patterns, identifies letters and words, and converts them into actual text -- that is, into strings of characters, spaces, and punctuation -- which text-to-speech and screen-reader software can read aloud to the user.
Creating OCR Scans
To test whether a scan is accessible, try selecting/highlighting text in it using your cursor. If you are unable to select text, it is an image file and is not accessible.
Bryn Mawr Multifunction Copiers
The multifunction copier scanners automatically OCR scanned text, so articles and book chapters you scan now should be accessible. However, old scans of articles and book chapters are often non-OCRed, image PDFs.
If you have an inaccessible scan and it is of decent quality, you can convert it in one of two ways:
- Using Adobe Acrobat
- Upload it to the Document Converter and convert it to a "tagged PDF" that is OCRed and accessible.
If your scan is of poor quality, the Document Converter will be unable to convert it successfully. In this case, it is best to make a new scan from a clean original, or try converting it in Adobe Acrobat.
Note: The Document Converter can usually handle scans in which two facing pages appear on the same scanned page (i.e. a "two-up" scan), as long as each page has only one column of text. For multi-column documents (and best results generally), scan pages one a a time.
A decent-quality scan is one with:
- Very little or no underlining, highlighting, or margin notes (these confuse the OCR software)
- No text is cut off or obscured by shadows, page distortions, fingers, etc.
- All pages facing the same direction
What is a screen reader?
A screen reader is an app that literally reads ALL content on the screen—menus, images, hyperlinks, dialogue boxes, folders, etc. JAWS, NVDA, and Apple Voice Over are prominent screen readers. Individuals who are blind or visually impaired rely on screen readers to access content via keyboard shortcuts. If page information is not appropriately labeled (i.e., tagged), the screen reader cannot identify it. That information will not be conveyed to the user. Consider how frustrating it would be if you were navigating a document in which the only feedback was “unknown image.”