About

Base46 encoding uses a reduced 46-character alphabet (0 - 9, A - Z, a - j) to represent arbitrary binary data as printable text. It produces longer output than Base64 due to the smaller radix but avoids characters like +, /, and = that cause issues in URLs, filenames, and certain legacy systems. Misidentifying which encoding a payload uses will corrupt data silently. A single wrong character in the charset mapping destroys the entire decoded output with no partial recovery. This tool validates input against the exact charset before conversion and reports precise error positions.

Base64 uses 64 characters and is standardized in RFC 4648. Base46 is a niche encoding found in specific embedded systems, QR code optimizations, and proprietary protocols where the character set must be restricted. This converter handles both directions with full UTF-8 support. Note: this tool assumes the most common Base46 alphabet variant. If your system uses a custom alphabet, results will differ.

Formulas

Base46 encoding treats the input byte array as a big-endian unsigned integer and performs repeated division by 46. Each remainder maps to a character in the Base46 alphabet.

Encoding: input → N = len−1∑i=0 byte_i ⋅ 256^len−1−i

Then repeatedly: r = N mod 46, N = N − r46

Output character: c = alphabet[r]

Decoding reverses the process:

N = len−1∑i=0 value(c_i) ⋅ 46^len−1−i

Where alphabet = 0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghij (indices 0 - 45). The expansion ratio is log(256)log(46) ≈ 1.447, meaning every input byte produces approximately 1.45 Base46 characters. For Base64, the ratio is 43 ≈ 1.333 with padding.

Reference Data

Encoding	Alphabet Size	Characters Used	Bits per Char	Expansion Ratio	Padding	URL Safe	Common Use
Base16 (Hex)	16	0 - 9, A - F	4.00	2.00×	None	Yes	Hashes, MAC addresses
Base32	32	A - Z, 2 - 7	5.00	1.60×	=	Yes	TOTP tokens, Crockford IDs
Base36	36	0 - 9, A - Z	5.17	1.55×	None	Yes	Short URLs, case-insensitive IDs
Base45	45	0 - 9, A - Z, special	5.49	1.46×	None	Partial	EU COVID certificates (QR)
Base46	46	0 - 9, A - Z, a - j	5.52	1.45×	None	Yes	Embedded systems, proprietary protocols
Base58	58	Alphanumeric minus 0OIl	5.86	1.37×	None	Yes	Bitcoin addresses, IPFS
Base62	62	0 - 9, A - Z, a - z	5.95	1.34×	None	Yes	URL shorteners, tokens
Base64	64	A - Z, a - z, 0 - 9, +/	6.00	1.33×	=	No	Email (MIME), data URIs
Base64url	64	Same but -_ instead of +/	6.00	1.33×	Optional	Yes	JWT, URL parameters
Base85 (Ascii85)	85	ASCII 33 - 117	6.41	1.25×	None	No	PDF, PostScript
Base91	91	Printable ASCII minus -'\	6.51	1.23×	None	Partial	Compact binary-to-text
Base128	128	Full 7-bit ASCII	7.00	1.14×	None	No	Experimental, protobuf varint

Frequently Asked Questions

Base46 has a smaller alphabet (46 vs 64 characters), so each character carries fewer bits of information: log₂(46) ≈ 5.52 bits vs log₂(64) = 6 bits. More characters are required to represent the same data. The expansion ratio for Base46 is approximately 1.45× vs 1.33× for Base64.

This implementation uses the standard Base46 alphabet: digits 0 - 9 (indices 0 - 9), uppercase A - Z (indices 10 - 35), and lowercase a - j (indices 36 - 45). Any character outside this set will trigger a validation error during decoding.

Yes. The encoder first converts the input string to its UTF-8 byte representation using the TextEncoder API. Multi-byte characters (e.g., emoji, CJK) are properly handled. The resulting byte array is then treated as a big-endian integer for Base46 encoding. The decoder reverses the process, reconstructing the byte array and decoding it back via TextDecoder.

Leading zero bytes are preserved using a prefix counting mechanism. Each leading zero byte in the input maps to the first character of the alphabet (0). During decoding, these prefix characters are converted back to zero bytes before the big-integer reconstruction. Without this step, data beginning with null bytes would lose information.

No. Unlike Base64 (RFC 4648) or Base32 (RFC 4648), Base46 has no formal standard. It exists as a convention in specific embedded systems and proprietary protocols. The alphabet order may vary between implementations. Always verify the exact charset mapping when interoperating with external systems.

The tool validates input against the Base46 charset before attempting conversion. Base64 uses characters k - z, +, /, and =, which are outside the Base46 alphabet. The tool will report the exact invalid characters and their positions, preventing silent data corruption.