The cleaner the original, the better the conversion. SensusAccess reads the text in a scanned page much the way your eyes would, so anything that makes a page hard to read will also make it hard to convert.
Start with the best copy you can find
- A digital original, like a Word file or a text based PDF, almost always converts better than a scan.
- If a reading is hard to convert, look for a cleaner version. Students can ask their instructor, and instructors can check with the library.
If you are scanning, aim for pages that are
- Straight, not tilted or curved near the spine
- Sharp and in focus, not blurry or grainy
- Evenly lit, with no shadows
- Free of handwriting, highlighting, underlining, and margin notes
- One page per image when you can, rather than two facing pages
If the first result is not great
- Clean up the original and try again.
- For image based files, convert to a tagged PDF first, then convert that tagged PDF to your final format. This extra step can sometimes improve the result.
After you convert, check for
- Words that came out garbled or misspelled
- Columns that got mixed together
- Tables, charts, and math that did not come through clearly
- Headers, footers, or page numbers showing up in the middle of the text
SensusAccess is a great first step, but it cannot describe a chart or fully fix a complicated table on its own. Documents like that may need more hands on accessibility work.