Represent binary data as Braille
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.


Convert binary data to Braille and back. The idea is that the Braille text visually resembles the original binary. For example, the binary sequence 0b11110001 0b10100101 becomes "⣇⢕". Each column represents each nybble, with the most significant bit at the top.


This is a hijacking/repurposing of Braille in the same way that Base64 repurposes alphanumeric ASCII — it is most likely of no use to Braille users. Or to anybody, for that matter. For an actual Braille module, try braille.


npm install braille-encode


var brailleEncode = require("braille-encode");

var buf = new Buffer("d41d8cd98f00b204e9800998ecf8427e", "hex");

var str = brailleEncode.encode(buf); 
console.log(str); // "⡓⣘⠙⣋⢹⠀⡥⠐⢏⠁⢈⡉⠟⡏⠢⡾"

var buf2 = brailleEncode.decode(str);
console.log(buf.equals(buf2)); // true


Given 1MB of input, braille-encode returns 3.00MB of UTF-8, 2.00MB of UTF-16 or 4.00MB of UTF-32.

Compare Base64, which returns 1.33MB of UTF-8, 2.67MB of UTF-16 or 5.33MB of UTF-32.

Notes on dot numbering and significance

I numbered the eight Braille dots as follows:

8 4
7 3
6 2
5 1

Each dot, if filled, has the following significance:

128  8
 64  4
 32  2
 16  1

Note that this is different from how Braille conventionally numbers the dots. Braille has:

1 4
2 5
3 6
7 8

Which would suggest that the significance of each dot is:

 1    8
 2   16
 4   32
64  128

For example, the byte 0b11110000 would be represented as "⣰". However, this would be relatively difficult to understand. Since the Unicode chart uses this ordering, this means encoding/decoding isn't a matter of simply taking the hex and adding/subtracting 0x2800.