Course materials are accessible if students with vision and hearing differences can make effective use of the same materials used by other students. While providing alternative materials for some students may be necessary at times, the objective of accessibility is to design a single set of materials that can be used successfully by all students. This requires planning at the time course materials are selected or created and developed, and thoughtful evaluation and retrofitting of materials already in use.
One way to think about accessibility is to consider that a given item within the materials of a course might communicate through three distinct channels:
Images, such as charts, graphs, photographs, and video,
Text, a particular type of visual information made up of letters, numbers, and punctuation, and
Audio, including spoken dialog, music, and other sounds.
Each of these channels, in various formats, may present different kinds and degrees of challenges to different users. More accessible materials communicate the SAME information through two or three of these channels. Less accessible materials communicate through only one channel, or communicate DIFFERENT information through different channels, increasing the likelihood that an individual user might miss something important.
Depending on the types of materials and how they are used, making course materials more accessible may include activities such as:
For a broader discussion of what accessibility is, start with:
https://en.wikipedia.org/wiki/Accessibility
Detailed characteristics of accessible materials are described at:
https://www.w3.org/WAI/WCAG20/quickref/
Not completely.
Only one aspect of accessibility works reasonably well in a fully automated mode, and that is reading simple text out loud. Screen readers, PDF readers, and many operating systems have built-in abilities to convert written text to audio, at least for major world languages, and these work reasonably well, although not perfectly. In practice, however, many documents are not very understandable when processed by text-to-voice readers. This happens because the layout of many documents relies too heavily on visual cues other than text, or is simply too complex. To operate effectively with screen readers, documents must be designed from the beginning with accessibility in mind (see "Online Text," below).
Practical problems also often prevent computerized translation of voice to text from working reliably. Computerized voice-to-text transcription works best when a single speaker has time to train a particular program to their individual speech patterns. Live interactions, unusual vocabulary, poor audio quality, multiple speakers, and varying dialects or accents will quickly reduce the output text to gibberish.
Descriptive audio, verbally describing the contents of an image or the nature of video action, remains far beyond the reach of existing pattern recognition and language processing software.
Making course content accessible can be assisted and supported by a variety of technologies, but relies heavily on human awareness and planning.
Generally speaking, it is the responsibility of each instructor to ensure that course materials are accessible. The Library, the Online Learning Team, and the Office of Accessible Education can provide guidance and can assist with some of the technical steps to support the efforts undertaken by instructors and departments. Depending on the details of materials under development or already in use, instructors and departments might need to plan on tasks such as the following:
Students with vision differences may use individual programs (such as a PDF reader) as well as a generic "screen reader" program that can read displayed text out loud. How this works depends on the reader software to which the student is accustomed, the layout of the text being read, and the program or file format in which the text is recorded (web page, word processing document, presentation slides, etc.)
In general, complex text layouts that rely on visual queues such as lines, borders, white space, and color to organize materials cannot be processed well by screen reader software. A simpler layout with a top-down organization is generally more accessible. A few tips are:
More thorough guidelines and suggestions for making web pages accessible can be found at:
https://www.w3.org/standards/webdesign/accessibility
Guided examples, with explanations, are available at:
http://accessiblewebsitedesignexamples.com/
A selection of before and after changes, with analysis, can be found at:
https://www.w3.org/WAI/demos/bad/
Course materials often originate as common office-type documents, such as word processing files, spreadsheets, and slide presentations. Many different programs are available, and different programs may provide different tools to help with designing more accessible documents. In general, the advice described above for "Online Text" should be applied.
For Microsoft Office programs, the following tools and tips may be of use:
Accessibility checker:
Creating Accessible Excel Workbooks:
Creating Accessible Power Point Presentations:
https://support.office.com/en-us/article/Create-accessible-PDFs-064625e0-56ea-4e16-ad71-3aa33bb4b7ed
Creating Accessible Word Documents:
Many of these tips can be applied to similar programs from other vendors as well.
Slide presentations produced using Microsoft PowerPoint, Open Office Impress, Apple Keynote, or other programs, can usually be read out loud by text-to-voice programs. Many of the programs used to view such presentations (including Adobe Acrobat PDF reader and similar programs) have their own, built in ability to read the text out loud. For best results, the guidelines described above should be followed.
Slide presentations can also be narrated, meaning that the creator can script and record audio that is linked to each slide. To replace the use of a text-to-voice screen reader, though, the narration must be thorough, including descriptions of the text and images on each slide. Less thorough narration may require the visually challenged learner to listen to the audio narration AND use a screen reader to understand the text and images on the screen.
Some slide editing programs, such as MS PowerPoint, provide screens for the editing of presenter's notes. These can act as a script for narration, and with a compatible slider viewing program, can be accessible to the end-user, thus providing both text and audio of the same information. See PDF Files Created From Office Documents, below, for some considerations on saving enhanced slide presentations in as PDF documents.
Once a slide presentation is fully narrated, it is often a comparatively small step to convert the presentation into a video. This is sometimes appropriate or even necessary, but can also be a step away from broad accessibility. Be aware that slides converted to video can not be read by text-to-voice screen readers.
Scanning a paper document produces a digital copy of the document and stores that digital copy in a file that conforms to a standard known as PDF (Portable Document Format). Documents stored in PDF files can be displayed on almost any computer using Adobe Acrobat reader and many other freely available programs.
Some strengths of PDF files, for purposes of increasing accessibility, are:
Some weaknesses of PDF files, for the purposes of increasing accessibility, are:
A PDF file may be nothing more than a set of raster images -- digital photographs -- of the original documents that were scanned. PDF display programs can NOT read out loud from a PDF file that contains only images. The PDF file must contain text --individual letters, numbers, and punctuation-- for the read-out-loud function to work. For PDF files created by scanning, this requires a process known as Optical Character Recognition (OCR), in which computer software attempts to identify which parts of the digital photograph of an original page contain text, to identify individual letters, numbers, and marks, and then to identify words.
When a PDF file is displayed on a computer, it is not always possible to tell just by looking whether the PDF file contains only images or a combination of images and text. What appears to be text on the screen might be nothing more than a digital photograph of the original text. If the PDF display program allows the user to select and copy text from the display and accurately paste that text into another program (such as a word processor), then the PDF file contains text. If it is not possible to select individual words or lines from within a displayed PDF document, then the PDF contains only images (at least with regard to the section being displayed at that moment).
Before creating a PDF file by scanning a paper document, check the settings of the scanning software:
The quality of both the images and the recognized text of a scanned document are strongly affected by the original document. Where feasible, choose an original that is in good physical condition, with good contrast between text and paper, with few stains or tears, and little pencil or ink marking. A modern document printed with a commonly used typeface will support OCR much better than a document using obsolete, inconsistent, or unusual letter shapes. Text in multiple columns may require special handling, especially if boundaries between columns are not always clear. If you require assistance with re-scanning materials, please contact Jessica Perry at perry@loyno.edu.
An existing PDF file that contains no recognized text can often be reprocessed by PDF scanning programs to complete the OCR steps and produce a new PDF file containing text data. However, the OCR process will be limited by the quality of the original scan. If the original document was of poor quality, or if the quality (image density) settings used during the scan were low, the OCR reprocessing will not produce reliable results. Once a scan is complete, there is no way to improve the quality (raise the data density and file size) of the PDF file: the original must be re-scanned with settings for higher quality.
Here's a quick video on how to retroactively apply OCR to your pdf and do some edits that will make it more accessible. It is also possible to edit OCR, if it does not apply perfectly. Jessica Perry, in the Monroe Library, is well versed with editing OCR and tagging illustrations.
If the student requires headers or anchors to move around in a document in class, we can also provide assistance.
For more information on Optical Character Recognition (OCR), start with:
https://en.wikipedia.org/wiki/Optical_character_recognition
Word processing documents, slide presentations, and spreadsheets can usually be saved as PDF files. The main advantage to doing so is that the PDF file can be displayed on almost any computer, with a high level of certainty that the document will look the same on all of them, all using free software. When a PDF file is created directly from one of these office documents, there is no need for a step of Optical Character Recognition (OCR). Everything that is text in the original is passed through to the PDF file as text, rather than graphics, with nearly perfect accuracy.
The primary determiner of the accessibility of the PDF is the layout of the original document. As described above, a document with a top-down design, simplified tables, reduced visual clutter, and proper use of headings and subheadings is likely to become an accessible PDF file.
When saving a document as a PDF file, different programs may provide varying options that can impact the quality of the output file. Generally, select settings for higher quality (larger files) and retention of "tags" (keeping heading structures, etc.), if the program provides such settings.
One potential limitation, depending on software and settings used, is that alternative text (text that can be turned on or off by the user and that explains a chart or image) might not be transferred to the output PDF file, or might not be implemented in a way that is practical for a particular environment. It might be necessary to extract alternative text separately and append that to the PDF file.
The PDF file format can support highly accessible documents in other ways as well. For example, a slide presentation saved to PDF with the right settings will contain the primary slides, the audio narration, and the presenter's notes (often the narration script). However, the ability of PDF viewers and readers to support all these options may vary. The free Adobe Acrobat Reader and Adobe Acrobat Reader DC for Mac and Windows support all these possibilities. Many other PDF viewers may not.
Increasingly, commercially available DVDs are packaged with auxiliary media to enhance accessibility. These can include:
Where practical, preference should be given to materials that include subtitles and descriptive audio. For DVDs already owned by Loyola's library, the library catalog will usually state whether a DVD includes these auxiliary media. This can be found by displaying the full catalog record and checking the entry labeled "Language."
For video recordings produced by a department or instructor, media for the enhancement of accessibility would have to be developed by the instructor and/or department. This could include:
Each of these is a technical task requiring significant person-hours as well as proper hardware and software. Transcription and captioning services can be purchased from third-party service providers.
For an overview of Close Captions and Subtitles:
https://en.wikipedia.org/wiki/Subtitle_(captioning)
For more information on descriptive audio:
http://www.acb.org/adp/ad.html
If a video was produced by a department or instructor, if a script was written in advance and closely followed, AND if the script already exists as a word processing document, then it might be cost-effective to create subtitles in house. Both online and installable software is available to help synchronize a script with a video in a way that creates text-based subtitles. The work of synchronizing the text with the video to produce valid subtitles will take several times the length of the video itself, and requires considerable attention to detail.
Without a complete and accurate script already written, the process of creating accurate subtitles becomes time-consuming and burdensome. Transcribing a video requires skill and experience, and typically takes 5 to 6 times the ordinary playing time of the video, before synchronizing text with video can even begin. Automatically created transcriptions generated by speech-recognition programs require very careful editing and are unlikely to save much time. For these reasons, it is often more cost-effective to send a video out for subtitling by an outside service provider.
The University currently contracts with VidGrid to provide captioning. It can be completed in 24 hrs. However, the video must be uploaded to VidGrid for them to provide captioning. We cannot provide captioning for material that we don't own. If you are using Collaborate to create video, you must have it set to be downloadable, or still have access to the session in order to reset the download settings. YouTube, which we have access to through the University's G-suite accounts, does provide subtitling, but they can be rudimentary or just wrong, especially with technical material. If you are having students create video for one another in the online environment, using their university-supplied YouTube accounts, with proper privacy settings, will at least provide some captioning.
Many companies will transcribe a video and convert the transcript to timed-text subtitles for a fee. When selecting a service provider, be aware that transcription and subtitling requires skill and experience. Many online advertisers claim to offer transcription service for "as low as" $1 per video minute, but this rock-bottom rate is rarely achievable in practice. Anticipate paying $1.50 to $3.00 per minute, with specialized vocabulary and non-English dialog pushing rates toward the higher end. An ongoing relationship with a reliable service provider often tends to reduce the per-minute rate over the long term. Advanced planning also helps, as the per-minute rate for 24-hour turnaround is often higher than the per-minute rate for 7-day or 14-day turnaround.
Increasingly, video transcription services are performed entirely online. After establishing an account with a payment method, all the user has to do is upload a video file to a website. The upload might also include text documents to assist the transcribers, such as lists of speakers' names or a lexicon of unusual or technical terms. The service provider will then process the video and make the subtitle file available for download from the same website.
Downloadable Subtitles and Scripts
Increasingly, DVDs are release for sale with subtitles included. However, older DVDs and DVDs produced outside of North America might not have subtitles at all. In such a situation, it might be helpful, as a stop-gap measure, to download subtitles or a full script from the internet, if such can be found.
Subtitles can be found online for a surprising variety of videos, but often only for popular, commercial productions. Subtitles are unlikely to exist online for obscure or privately produced videos. Only subtitle files in a text-based format such as SubRip (.srt) or WebVTT (.vtt) will be of use. Image-based subtitle files (such as .sub/.idx) can be converted to text, but the process is error-prone and time-consuming. However a text-based subtitle file is retrieved, it is usually necessary to verify that it is of good quality and matches the video being used.
SOME of the more reliable websites providing subtitle files are:
For a smaller number of movies, full scripts can sometimes be found online. One such source is:
Streaming Video on Canvas
Many courses link to streaming video from within Canvas. Going forward, these videos will include subtitles where feasible. These subtitles will always be displayed, which some viewers might find distracting. However, please take into consideration the following limitations:
Existing Streaming Files. Thousands of videos already exist on our streaming service, many without any kinds of subtitles or alternative audio tracks. For this reason, it is important that each instructor review videos used in current or anticipated courses and request reprocessing of videos as needed.
Original Sources. Existing streaming files can be reprocessed to include alternative audio or subtitles only if those items are available on the original DVD (for purchased videos), from the vendor (for some educational videos), or from the department or instructor (for videos produced in house). If it is necessary to create a new descriptive audio track or subtitles, this will have to be done by the instructor or academic department, or sent out to a third-party service provider.
Time. Reprocessing a video to include subtitles or descriptive audio may take time. The process will take longer if it is necessary to download, edit, or convert subtitles or to send a video out for transcription. Advanced planning, therefore, is important.
Flash Video. For technical reasons, streaming videos are currently stored and streamed in Flash video format (.flv). This format supports only one audio track per file, which means that each alternate audio track requires a complete and separate video file. This could be a consideration, for example, when supplying a video in more than one language or with descriptive audio for the visually challenged. Maintaining multiple copies of a video increases the storage space required by our streaming service, and requires separate Canvas course links for each copy.
Default Subtitle Choice. By default, videos will be processed for streaming with English audio and English subtitles. An instructor needing a different arrangement of audio and subtitles (e.g., for a language class) needs to notify the Online Learning Team as far in advance as possible.
Live Interactions
Real-time exchange of audio and visual information between two or more participants often requires live, fee-based interpretation in order to achieve accessibility. An interpreter participating in the exchange might verbally describe visual images, might translate conversation into sign language, or might transcribe a conversation into text.