Because of the complexity of Braille software, most this document will focus on explaining Braill and basic Unicode implementation primarily for sighted users. Visually impaired users should consult their translation software for information on Unicode implementation.
Braille was developed by Louis Braille in the early 19th century based "night writing" developed for the French army. Like Braille, night writing uses a series of raised or embossed dots in different patterns so that readers can quickly decipher the text by touch alone. You can test for yourself on public signs with both Braille and embossed letters.
Although embossed or raised letters have also been used, most visually impaired users report that the system of embossed dots is easier to interpret by touch then written letter forms. A literate Braille reader may be able to scan 100-200 words per minute by touch.
Braille for English and other languages is organized in cells of six dots (two columns, three dots high). By tradition, each dot position is assigned a number. Number 1 is the upper-left dot, number 2 is below 1 in the left column and number 3 finishes the left column. Dot number 4 is top dot in the right column followed by dot 5 and dot six in the lower right. See diagram below.The dot order in the left column is 1-2-3 and the dot order in the right column is 4-5-6. The six dots, plus the option of dot vs. no dot leads to 26 or 64 possible configurations.
Braille cell of six dots
Braille characters are described in terms of dot postitions. For instance English letter A is just dot 1 in Braille; letter B is dots 1 and 2; letter C is dots 1 and 4; and letter D is dots 1,4,5. See Braille Alphabet Chart for details.
Grade 1 Braille is simply English text transcribed letter for letter into Braille. However, many users prefer Grade 2 Braille which contains a number of contractions for diagraphs like "th" and "ch" as well as common words. Therefore, translation of English text to Braille is generally not a simple 1-to-1 substitution.
In addition to English, Braille alphabets have been developed for other languages which include symbols for non-English sounds. A list of tables is available from Scripts for the Blind. A set of Braille characters, for math, music and phonetic symbols is also available.
Notice that almost all Braille scripts, even ones for Japanese, are phonetically based. For instance, Japanese Braille requires that text be converted to kana only (no kanji ideographs).
In addition, most Braille alphabets try to use the same patterns for the same sounds. For instance pattern 1-4-5 is "D" in English, delta (Δ) in Greek, dalet (ד) in Hebrew, dal (د) in Arabic and letter (Д)in Russian.
Visually impaired with a computer can use Braille in several ways. First, they can print documents on a Braille printer which uses a thicker paper which supports embossing. A Braille user can also use Braille displays, although many visually impaired users also use a screen reader, such as Jaws, which reads text aloud in a synthesized voice.
Note that most Braille users with a computer require Braille translater software which can read, produce and print Braille. A common example is from Duxbury, but other systems are also available. Because of the complexity of Braille software, the rest of this document will focus primarily on sighted users.
Visually impaired users should consult their translation software for information on Unicode implementation.
The Unicode block "Braille Patterns" contains all cell configurations, but for cells of 8 dots, not just 6. Note that an 8 dot cell allows for an extra 128 characters in a Braille alphabet/orthography. Dots 1 through 6 are organized just as they were before, but the bottom row now contains dots 7 then 8. Thus the dot order in the left column is 1-2-3-7 and the dot order in the right column is 4-5-6-8.
Unicode Braille Patterns cell with eight dots
Characters in the Unicode Braille Patterns block are named after the dot patterns, not any other meaning. Thus the character corresponding to English Braille letter B is "Braille Pattern Dots-12".
See the Braille Unicode table for details.
Like other scripts, Braille was encoded in older 8-bit systems such as North American ASCII Braille. Generally spealing, the traditional characters were replaced by Braille equivalents.
Many older files may be encoded in these older systems.
These recommendations are aimed for sighted users interested in replicating Braille on visual systems with Unicode.
The following browsers have the most consistent results
Internet Explorer 6 for Windows may not display Braille symbols by default. Users who prefer Internet Explorer for Windows should set the Latin font to a font which includes Braille. This problem was resolved in Internet Explorer 7.
Note some Braille fonts may only contain 6-dot cells. These contain 8-dot cells.
www.alanwood.net/unicode/braille_patterns.html (Braille Patterns Test Page)
libbraille.org/translator.php?table=american (Select American Computer)
This information only applies to sighted users who do not have access to a Braille translater.
Use the Windows Character Map to insert Braille symbols.
Use the Macintosh Character Palette to mathematical symbols
No matter which method you use to create the document, the HTML page must be declared as a Unicode page with the following meta-tag.
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
...
</head>
Although it is generally not recommended to specify fonts for a language, it may be necessary in order to force Internet Explorer 5-6 to use the correct display fonts.This can be done either via CSS styles or via the FONT tag.
Below is the CSS style class listing some suggested font specifications.
.braille {font-family: "Apple Symbols", "Braille", "DejaVu Sans","ghUBraille","Code 2000" }
Code 2000 is a popular shareware font with a large number of symbols including Braille.
If you can use either the Windows Character Map or the Macintosh Character Palette with Dreamweaver, Notepad (Win), Unipad (Win) or BBEdit (Mac) to insert Braille symbols directly into the HTML.
However, you may get uneven results between browsers.
You can you can substitute numeric entity codes for the mathematical symbosl. For example
