Braille Translator

Type it and the braille appears as you go, one cell per letter. The letters themselves are the easy part; what this also does is add the indicators braille needs and print does not — a capital sign before each capital letter, a doubled capital sign before a word written entirely in capitals, and a number sign to start a number.

Grade 1, which spells every word out letter by letter. Grade 2 contracts around 180 common words and letter groups — "the" is a single cell, "ing" is a single cell — with rules about where each contraction may be used, and transcribing it properly is a much larger job than a character table. Grade 1 is what braille labels and signage use, and what somebody learning braille reads.

With the first ten letters, after a number sign. The digits 1 to 9 are a to i, and 0 is j — so 123 is the number sign followed by the cells for a, b and c. The catch is what comes next: a letter from a to j straight after a number would read as another digit, so braille puts a letter sign in between. "1a" is four cells rather than two, and this tool adds that for you.

Not directly, and it is worth being clear about why. What you get here is the Unicode braille characters, which display the dot patterns on screen — useful for reading, checking and sending — but an embosser needs the right paper, spacing and line length for its own format. Copy the cells into the software that drives the embosser rather than printing this page.

Paste it into the same box. It recognises braille characters and turns them back into print, undoing the capital and number indicators as it goes. One ambiguity survives, because it is in the code rather than in this tool: the same cell is both a question mark and an opening quotation mark, so it is read as a quote when a letter follows it directly and as a question mark otherwise.

The letters, digits and punctuation here are English Grade 1. Most languages use the same twenty-six cells for the Latin alphabet, so a Spanish or German word written without accents transcribes correctly — but each country's braille code assigns its own cells to the accented letters, and those are not included. An á, an ö or a ł comes back unchanged and is named under the result, so you can see exactly which characters a national code would need to supply.

For a label, a short sign or a piece of practice text, it does the job. For a book, an exam paper or anything with tables, maths or layout in it, no — a real transcription makes decisions about structure that no character mapping can, and professional transcribers exist for that reason. Everything here runs in your browser, so nothing you paste is uploaded either way.