Judy Dixon
Part 1 of this article discussed the history of braille codes in the United
States. Part 2 will look at the braille codes used by screen readers on
computers and mobile devices and how those codes are being used to allow braille
readers to access different types of print representations.
Screen Readers and Braille
When using a braille display directly with a computer or mobile device, it is
the screen reader that determines the braille codes available on the device, not
the braille display itself.
Over time, in an effort to represent all the languages of the world and all the
alphabets that had ever been in use, a system called Unicode was created to
provide a way to represent hundreds of thousands of characters. The Unicode
standard now forms the basis for virtually all the characters on modern
computers and mobile devices.
The initial Unicode standard was published in 1991 by the Unicode Consortium.
Its goal was to create a variable bit standard to represent all the characters
in modern text including all the Latin, Asian, Greek, Arabic, Cyrillic, and
Hebrew alphabets, music, the International Phonetic alphabet, emojis, and even
braille. All 256 eight-dot braille characters have a Unicode value in the range
from U+2800 to U+28ff.
The screen reader is responsible for rendering these characters in braille. For
the most part, this is a fairly smooth process but there are still hiccups from
time to time.
A good example of this is the apostrophe character. In Unicode, the character
U+2019 is often used to represent both the apostrophe and the single closing
quotation mark. Print readers don’t even notice the difference because the two
characters are visually the same. For braille readers, this is a problem because
the apostrophe and the closing single quote are very different characters in
braille. The apostrophe is dot 3 and the closing single quote is a two-cell
symbol, dot 6, dots 3-5-6. Braille readers often wonder why braille translated
by a screen reader often contains the symbol for the closing single quote where
the apostrophe should be. Changes to UEB rules have been made to attempt to
alleviate this situation but the problem has not been entirely solved.
There are many other examples where the print text contains a character that
looks correct on the screen, but is, in fact, not the correct character so the
braille translation is not correct. This happens frequently with symbols such as
the degree sign or the multiplication sign where the author simply uses a
lowercase letter instead of the correct symbol.
Many braille readers prefer to read contracted braille. As screen readers have
evolved, a contracted braille option for electronically displayed braille has
become the norm. The screen reader accesses a braille table that provides
information on how the text should be contracted. These tables contain
algorithms which transform character sequences in the source text and apply
these transformations based on rules which consider the surrounding text. The
available tables vary somewhat depending on the operating system and/or the
screen reader being used.
Braille on iOS and iPadOS Devices
In Settings, Accessibility, VoiceOver, Braille, there are three settings that
relate directly to the braille code that is being used on an iPhone or
iPad. The options available in these settings depend on the default language for
the device. It is possible to choose a language for VoiceOver, separate from the
default language for the device. In VoiceOver Speech settings, the user can
choose to have VoiceOver detect language, and choose which language(s) appear on
the rotor. The current VoiceOver language is used with a braille display.
The first option in braille settings is Output. It determines what code of
braille you will see on your display. The options are:
Six-dot: (Uncontracted Six-dot braille),
Eight-dot (Uncontracted Eight-dot braille), or
Contracted (contracted braille).
Uncontracted six-dot braille does not use any contractions as the name suggests
but it does use dot 6 to indicate capital letters, dots 3-4-5-6 as a number sign
to indicate numbers, and it uses braille punctuation. Uncontracted eight-dot
braille also does not have any contractions but it uses dot 7 to indicate
capital letters, numbers are written in the lower portion of the cell, and
punctuation characters are from the original North American ASCII Braille code
which were also used in the Computer Braille Code.
You can change the braille output code by tapping on the Output button and
selecting from the available options or you can change it on the fly at any time
by typing a g-chord (dots 1-2-4-5-space) on the braille display.
The second option in braille settings is Input. The available options here are
exactly the same as those for Output. You can change the braille input code by
tapping on the Input button and selecting from the available options or you can
change it on the fly at any time by typing an 8-chord (dots 2-3-6-space) on the
braille display.
The third option is in braille settings Braille Tables. By default, in the
United States, one table is loaded—English Unified. This is what Apple calls
Unified English Braille (UEB). If you tap the Braille Tables button, an Add
Braille Table option will appear. This list is organized by language with
English at the top. Under English, there are nine tables available, listed in
three groups:
RTFC
Liblouis
System
English (UK)
English (US)
English (Unified)
The table referred to as English (US) is the code called English Braille,
American Edition that was used in the United States prior to the adoption of
Unified English Braille in 2012.
Braille on Android Devices
In August 2022, with the release of Android 13.0, braille features were
integrated into Google’s screen reader, TalkBack. Prior to version 13, Android
users who wanted to use a braille display with TalkBack had to install a
separate application to support braille. Google offered a limited application
called BrailleBack, but there were other options as well.
In TalkBack 13.0 Settings, under Braille Display, Languages, two
English-language braille codes were offered—Unified English Braille Contracted
and Unified English Braille Uncontracted. There was no option for Computer
Braille, ASCII Braille, or any code that would offer a one-to-one
representation. This meant that it was not possible to read a translated braille
file such as a book on BARD Mobile on an Android device with TalkBack.
In January 2023, Google released TalkBack version 13.1 with an expanded list of
available braille tables. Under English, the available tables now are:
Unified English Braille contracted
Unified English Braille uncontracted
English 6-dot Computer Braille
North American Braille Computer Code
English U.S. Braille (EBAE)
English Braille (U.K.)
English Braille (India)
Braille on a Windows PC
At the moment, the most popular screen readers for a Windows computer are JAWS
and NVDA. Both screen readers offer a selection of braille settings.
In JAWS, under Settings, Braille, General, Translation, JAWS offers braille
translation for dozens of languages. When set to English, Output has five
options:
Input options vary depending on what has been selected for Output.
JAWS also has an option to display text in 8-dot braille mode. This setting is
very helpful when reading documents that have already been translated into
braille. With Output set to Computer Braille and eight-dot braille turned off,
material already translated into braille, including contracted braille, music,
and so on can easily be read with a braille display.
In NVDA, available Input and Output tables for English in the U.S. are:
English U.S. 6-dot computer braille
English U.S. 8-dot computer braille
English U.S. Grade 1
English U.S. Grade 2
English North American Braille Computer Code
Unified English Braille Code Grade 1
Unified English Braille Code Grade 2
The fact that different braille codes have similar names and the same braille
codes have different names from screen reader to screen reader can be very
confusing to braille readers.
Braille on a Mac
On the Mac, braille settings are in VoiceOver Utility under the Braille
category. The Braille category starts with a tab control with 4 tabs:
Translation, Layout, Status, and Displays. Translation is selected by default.
The first option on the Translation tab is Output. The options are the same as
with iOS except that the Uncontracted six-dot braille option is just called “Six
dot.” The next option, Input, has the same options as Output.
Further along in VoiceOver Utility is the Braille Translations table option. The
table lists tables that have been installed, and there are Add Braille Table and
Remove Braille Table buttons. If Add Braille Table is chosen, a screen with two
tables opens. The first table is for selecting a language. English is at the
top, followed by an alphabetical list of languages. The next table shows
whatever tables are available for the language selected in the first table. The
options for English are:
Liblouis:
RTFC:
System:
English (UK)
English (US)
Choosing a Braille Table
Most screen readers sold in the United States have a default braille table,
usually Unified English Braille. This table will display contracted braille.
Some screen readers have a setting or command allowing the user to untranslate
the word at the cursor. This can be helpful when one is not sure of the braille
translation being used, the braille translation is incorrect, or there are
special characters that are not understood by the braille table being used.
For those who prefer not to see contracted braille, an uncontracted braille
option may be a better choice. It will show only six-dot braille with braille
capital signs and number signs, but all the words will be fully spelled.
A “computer braille” option would be the best choice for those who want to see
exactly what is on the computer screen. In general, each character is presented
in a single braille cell. There are some characters, such as various bullets
that are shown as two- or three-cell characters even in computer braille.
Reading Transcribed and Formatted Braille
Documents that contain transcribed and formatted braille are available for
download from several sources, most notably as books and magazines from BARD and
as books from Bookshare. These documents are often referred to as BRF files
because they usually have a .brf file extension. They are plain-text files that
are best read with one of the computer braille tables. In iOS, BARD Mobile, the
app from the National Library Service presents a six-dot braille view of the
.brf files. iOS does not have a six-dot computer braille option so reading .brf
files with other applications can be a bit difficult.
Reading Languages other than English
Many languages use accented letters. Often, the braille characters used for
those letters in the native braille code of the language are different from the
braille code used for the accented letters in English braille.
If you select an uncontracted version of a braille code for a language with a
Latin alphabet, it should be very similar to uncontracted Unified English
Braille. Spanish, for example, has the same dot configurations for all letters
of the alphabet as well as most punctuation. Spanish does have seven symbols for
accented letters (á, é, í, ó, ú, ü and ñ) that are not used in English. there
are some differences in the way punctuation is used as well, with question marks
and exclamation marks both before and after text.
The contracted braille codes of English and Spanish do have significant
differences.
Conclusion
Modern screen readers and braille-aware applications offer many options for
braille readers. While reading braille, it can be very helpful for a braille
reader to be aware of the current braille code being read, and to understand the
available options. In this way, braille readers can benefit from the flexibility
of changing braille codes to meet their personal reading preferences, as well as
to adjust the braille code to the material being read.
In the future, the structure of braille files will very likely be in Unicode.
Screen readers will need to display them correctly to the user and embossers
will need to just emboss the dots they are being sent. The advantage of this is
that the creator of the file will be able to specify exactly what dots are
conveyed to the user.