SensusAccess is an automated text conversion tool that converts files, text, and webpages into a variety of alternate media formats, including e-books, audiobooks, and digital Braille. In the process, it also makes files more accessible and easier to work with, such as turning a scanned image of a page into accessible, digital text with structured heading styles that can be navigated, searched, and read aloud by a screen reader or text-to-speech program.
A wide range of input file types are supported, including PDFs, most image files, Microsoft Word documents with and without math equations, plain text files, PowerPoint presentations, HTML documents, RTF files, LaTeX documents, and e-books in EPUB and MOBI (Kindle) formats. SensusAccess also supports a variety of languages, including all major European languages, US English, Latin American Spanish, Arabic, Cantonese, Japanese, Korean, Mandarin, Russian, Taiwanese, and more. Documents can also be multilingual. For a list of places to find digital texts for use with SensusAccess, see Sources for E-Text.
SensusAccess is free to use to anyone with an @illinois.edu email address, including students, faculty, staff, and alumni.
Jump to…
Using SensusAccess
1. Choose a file source
This can be a file you upload, the URL of a website, or plain text pasted into a textbox.
2. Choose a service type
SensusAccess offers four service categories:
- Audio: MP3 and DAISY Talking Books
- Braille: Grades 1 and 2
- E-book: EPUB, EPUB3, and MOBI
- Accessibility: This option generally has the most choices for file output types, including CSV, DOC/DOCX, HTML, PDF, RTF, TXT, and XLS/XLSX, depending on the original file type.
3. Choose conversion options
Depending on the service type you selected, you’ll be given a list of options, such as the output file type, font size, or reading speed.
4. Enter your email address and submit the request
The converted file will be delivered to your @illinois.edu address. Smaller documents are processed within minutes, while large files, like entire textbooks, may take several hours to be delivered.
What can I use it for?
Scanned pages and image-only readings
Inaccessible files, such as image-only PDFs of journal articles or textbook pages you scanned yourself, can be made into accessible text files that can be searched, annotated, navigated using heading structures, and read aloud by assistive technology programs.
Providing multiple formats
Lecture notes, handouts, and other instructional materials can easily be provided to students in multiple file formats, such as PDF, EPUB, and Word files, allowing students to choose the format that works best with their devices, apps, and study habits.
Making lecture slides accessible
PowerPoint presentations, which can be challenging to use for students with screen readers, can quickly be turned into accessible tagged PDFs.
What should I NOT use it for?
Files that must be fully accessible
While SensusAccess can make files more accessible, no automated tool can make them fully accessible. For instance, SensusAccess cannot add alt text to images, and it may make errors in converting documents (See “What might it struggle with?” below). If you have a student who needs fully accessible files, please contact DRES’ Accessible Media Services at dres-accessible-media@illinois.edu.
Confidential, sensitive, or personally identifiable information
Although SensusAccess deletes submitted files as soon as they are converted, there is always a risk when sending files over the internet that user-provided information could be intercepted by third parties. Do not upload any documents to SensusAccess that contain private or personally identifiable information, such as social security numbers or student grades.
Copyright-restricted material
You must comply with all applicable copyright laws when using SensusAccess. You also agree that you are legally accessing and duplicating materials with appropriate permission or qualification as fair use under the law. For more information about copyright and fair use, please see the University’s webpage on copyright resources.
What might it struggle with?
No automated tool can make a file fully accessible. You should always check for conversion errors and ensure the converted file complies with accessibility best practices for that file type, like adding alt text to graphics and checking the heading structure of text documents.
Below is a list of common errors to check for:
Low-quality source files
Blurry pages, poor color contrast, and warped text lines can all result in lower-quality converted files. Try to clean up your files before submitting them to SensusAccess. If you’re scanning or photographing a document yourself, make sure there is good lighting, the page is pressed flat, and the image resolution is high enough to be legible. If you’re uploading an existing file, check to make sure the pages are oriented the right way up before uploading, particularly whole-page tables that may be rotated from the rest of the document.
Similar characters
Characters that look alike, such as O and 0 or I and l, may be substituted for one another. This can be particularly problematic for MP3s and other files that will be read aloud, since trying to figure out and mentally substitute in a corrected word can easily distract a reader from the rest of the sentence that’s being read. Consider converting your files to Word initially before converting to an audio file, so you can fix any frequent or egregious character substitutions before trying to listen to the finished product.
Complex page layouts
Pages with lots of different elements — multiple columns of text, graphics, textboxes, pull quotes, etc. — may result in incorrect page layouts in the finished file. Check that the reading order makes sense (EG, lines of text aren’t merged across columns, graphics aren’t inserted mid-sentence, captions aren’t included as part of body paragraphs, etc.) and that no elements overlap one another (Such as captions being positioned on top of their graphics).
Alt text
Converted files will not have alt text added to graphics, and depending on the file types involved, alt text that was present in the original file may not make it to the converted file. Be sure to go into your finished files and add descriptions to any non-decorative images.
Purely decorative images, such as watermarks and visual embellishments that don’t convey semantic information, should be marked as decorative if the application allows for it, or have a null set, a single space surrounded by quotation marks (” “) entered as their alt text. This lets a screen reader user know that the image isn’t necessary to their understanding of the surrounding text, rather than being a relevant image that’s missing proper alt text.
Graphics and text
Pictures that include text — such as labels on a diagram — may convert strangely. The image may have chunks missing where the text was cropped out, or the text may be in a textbox over the picture, but not visible due to font color or textbox size issues. Alternatively, the text may be left as part of the image and not transcribed into accessible text at all.
Parts of an image can also be misinterpreted as text, resulting in an image with those elements edited out.
Tables
Check that any tables were transcribed sensibly, especially complex ones that include merged cells or inconsistent column structure. Make sure they haven’t been converted into graphics or text, that columns or rows haven’t merged, and that column headers are displayed above the correct columns.
Be especially conscious of tables with numbers; decimals and minus signs may be missing, and numbers may have become letters, like “IBOS” instead of “1805.”
Small details
Small details may be missing or incorrectly transcribed, such as missing footnote numbers or words that were originally split across lines still appearing hyphenated mid-line. Boundaries between things — the end of a line of text, page breaks, a body paragraph followed by a textbox — are the most likely locations for these errors to occur.
Audio files
SensusAccess will always try to reproduce the original file as closely as possible, which may not be exactly what you want, particularly with MP3s and audiobooks. For example, page headers and footers will be read out for every page, which can be very distracting to readers. Consider converting a file to Word initially so you can remove extraneous text, and check that sentences aren’t interrupted by other page elements, like graphics and text boxes appearing mid-sentence. Once your file has been edited, it can then be re-converted to an audio file for distribution.