Binary Converter

How does text become binary?

A computer stores text as numbers. Each character maps to a code point, and that code point is written in base 2, using only 0s and 1s. This binary translator uses UTF-8, the encoding behind most of the web, where every ASCII character takes exactly one byte of 8 bits and accented or non-Latin characters take 2 to 4 bytes.

Take the word Hi. The letter H has code 72, which is 01001000 in binary, and the letter i has code 105, which is 01101001. So Hi in binary reads 01001000 01101001. To decode it, split the string into 8-bit bytes, read each byte as a number, then look up the character.

How do you read a byte?

A byte holds 8 bits, and each position carries a weight: 128, 64, 32, 16, 8, 4, 2 and 1, from left to right. Add the weights of the bits set to 1 to get the value. For 01001000, the second bit (64) and the fifth bit (8) are on, and 64 + 8 = 72, the code for H. The tool draws this table for the first byte of your conversion so you can check the math yourself.

Numbers or text: what is the difference?

Converting the number 42 and converting the text 42 are two different jobs. As a number, 42 in base 2 is 101010 (32 + 8 + 2). As text, 42 is two characters: the digit 4 (code 52, 00110100) and the digit 2 (code 50, 00110010). Auto-detect reads a digits-only input as a number, so pick the text to binary mode when you want the character bytes instead.

What does auto-detect do?

The rules stay simple. An input of only 0s, 1s and spaces whose bit count is a multiple of 8 is decoded as binary text. An input of only digits is converted as a decimal number. Anything else is encoded as UTF-8 text.

What do common characters look like in binary?

Notice the pattern: the uppercase and lowercase versions of a letter differ by exactly one bit, the 32 weight, which is why case conversion costs a processor almost nothing. For every code from 0 to 127, see the ASCII table reference page.

Why will some binary not decode?

Not every sequence of bytes is valid UTF-8. The byte 11111111 (255), for one, never appears in UTF-8 text, so the tool flags it instead of printing garbage. If your binary came from a number rather than text, decode it with the binary to number mode, where 11111111 is simply 255.