How Braille Works, and How to Turn Any Text Into Braille Dots

I spent an afternoon trying to read a braille label on a medicine box by sight, chart in one hand, phone in the other, and got it wrong twice before I noticed I had swapped two dots. That small failure taught me more about the system than any diagram had: braille is not a font, it is an encoding, and once you see the rule behind the dots you can read and write it on paper without guessing. This guide walks through that rule and shows how to convert ordinary text into braille dots without owning a single embossing tool.

One cell, six dots

The whole alphabet rides on a tiny grid. Each braille character, called a cell, is a 2x3 arrangement of raised dots: two columns of three positions, numbered 1, 2, 3 down the left column and 4, 5, 6 down the right. A cell can show any combination of those six positions raised or flat, which gives 64 possible patterns including the empty cell used for a space. That is the entire vocabulary of shapes. Every letter, digit, and punctuation mark is just one of those 64 dot patterns, or a short sequence of them.

Because the positions are numbered, braille readers and transcribers describe a cell by listing its raised dots rather than drawing it. The letter a is "dot 1." The letter b is "dots 1 and 2." Once you talk about cells this way, checking your work becomes arithmetic instead of art, and a converter can show you the exact dot list for any character you type.

Grade 1: one cell per letter

Grade 1, also called uncontracted braille, is the spell-it-out form. Every letter maps to exactly one cell, with no shortcuts, which makes it the right place to start learning and the safest mode for predictable, reversible conversion.

The first ten letters, a through j, use only the top four dots (positions 1, 2, 4, 5). The next ten, k through t, repeat those same ten shapes with dot 3 added underneath. The last stretch, u through z, adds both dot 3 and dot 6. So the alphabet is really one pattern of ten shapes, shifted down the cell twice. Learn the first row and you have most of the rest.

A couple of small signals handle the things letters alone cannot. A capital letter is flagged by placing the capital sign (dot 6, written as ⠠) immediately before the cell, since braille has no separate uppercase shapes. Digits reuse the cells for a through j: the number sign (dots 3, 4, 5, 6, written as ⠼) goes in front to announce that the cells that follow are figures rather than letters. That is why 1 and a share a shape and only the leading ⠼ tells them apart.

A worked example

Take the word cat. In Grade 1 it is three cells, one per letter:

c = dots 1 and 4 = ⠉
a = dot 1 = ⠁
t = dots 2, 3, 4, 5 = ⠞

So cat becomes ⠉⠁⠞. Now add a capital and a number. Typing Hi gives ⠠⠓⠊: the ⠠ marks the next letter as uppercase, ⠓ is h, and ⠊ is i. Typing the number 25 gives ⠼⠃⠑: the ⠼ announces digits, ⠃ is 2 (the same shape as b), and ⠑ is 5. Read those rules backward and the dots resolve cleanly into text again, which is exactly what a two-way converter does.

You can run any of these through the Braille Translator. Type text and it shows the matching Unicode braille glyphs from the U+2800 block, with an optional dot breakdown that names which of the six positions each cell uses. Paste braille the other direction and it recovers the letters, reading the number sign and capital sign so ⠠⠓⠊ comes back as Hi rather than a jumble. Everything runs in the browser, so nothing you type leaves the page.

Grade 2: contractions for speed

Grade 1 is faithful but slow to read by touch, and a page of fully spelled-out words eats a lot of paper. Grade 2, the contracted form used in most published books, fixes that. It assigns single cells or short clusters to common words and letter groups: the lone cell ⠯ stands for the whole word and, ⠮ for the, and dozens of letter combinations like ch, ing, and ed collapse into one cell each. A fluent reader covers far more text per line, and a braille book stays a manageable thickness.

The catch is that Grade 2 is not a simple lookup. Whether a contraction applies depends on context, word boundaries, and a large rule set, so a one-to-one machine mapping would mis-transcribe constantly. That is the honest reason most quick converters, including the one above, stay in Grade 1: a predictable letter-for-letter result you can verify against a chart beats a contracted output you cannot trust. Treat Grade 1 as your learning and checking tool, and reach for a full braille transcription program when you need publication-ready Grade 2.

What it is good for, and what it is not

Knowing the dot rule unlocks a few everyday uses. A teacher can put a student's name on the board as cells and turn on the dot breakdown so a class sees that b is dots 1 and 2 while c is dots 1 and 4, not a vague blur. A learner can paste in braille they wrote by hand and read back what it actually spells, catching a dropped dot before it gets baked into a whole page. Someone writing about accessibility can drop real braille glyphs into a caption, where they stay crisp and selectable because they are genuine Unicode code points, not an image.

One boundary matters: what a converter produces is visual braille. The dots are drawn as on-screen glyphs in the U+2800 block, so they look right but print flat on an ordinary printer. Raised, touchable braille needs an embosser or swell paper and a font sized to standard dot spacing. Use the on-screen version to learn the cells and check transcriptions, and send the text to an embosser when you need a page someone can read by touch.

If encodings like this interest you, the same "rule behind the symbols" thinking applies elsewhere. Seeing how characters map to raw bits in the text to binary tool makes the braille mapping feel less like a special case and more like one more way humans have agreed to write the same letters down. Master the six-dot cell, and braille stops being a wall of bumps and turns into something you can simply read.

Made by Toolora · Updated 2026-06-13