About

Baudot code (ITA2) is a 5-bit character encoding used in teleprinter communications since the early 20th century. Each character is represented by 5 binary digits, yielding 32 possible combinations. Because 32 slots cannot cover both letters and figures, the ITA2 standard uses two shift states: LTRS (code 11111) switches to the alphabetic plane, and FIGS (code 11011) switches to the numeric/punctuation plane. A lost shift character during transmission causes all subsequent characters to decode incorrectly - a phenomenon called "mutilation" that historically corrupted entire telegrams. This converter implements the full ITA2 standard with stateful shift tracking. It assumes LTRS shift as the default initial state, consistent with standard teleprinter protocol. Note: regional ITA2 variants exist (US TTY vs. British); this tool follows the international ITA2 mapping.

Formulas

Baudot ITA2 decoding is a stateful finite automaton. The decoder maintains a shift register S that is either LTRS or FIGS. For each 5-bit input code c, the output character is determined by:

decode(c, S) =

{

S ← LTRS, output nothing, if c = 11111S ← FIGS, output nothing, if c = 11011T_S[c], otherwise

Where T_LTRS and T_FIGS are the two lookup tables of 32 entries each. The initial state is S₀ = LTRS. For encoding (text → Baudot), the algorithm iterates each character, finds which table(s) contain it, inserts a shift code when the required table differs from the current shift state, then emits the 5-bit code. Characters present in both tables (SP, CR, LF, NUL) do not require a shift change.

Reference Data

Binary	Decimal	LTRS	FIGS	Notes
00000	0	NUL	NUL	Null / Blank
00001	1	E	3
00010	2	LF	LF	Line Feed
00011	3	A	−	Hyphen in FIGS
00100	4	SP	SP	Space
00101	5	S	BEL	Bell signal in FIGS
00110	6	I	8
00111	7	U	7
01000	8	CR	CR	Carriage Return
01001	9	D	$	Dollar / ENQ variant
01010	10	R	4
01011	11	J	'	Apostrophe in FIGS
01100	12	N	,	Comma in FIGS
01101	13	F	!	Exclamation in FIGS
01110	14	C	:	Colon in FIGS
01111	15	K	(	Open paren in FIGS
10000	16	T	5
10001	17	Z	"	Double quote in FIGS
10010	18	L	)	Close paren in FIGS
10011	19	W	2
10100	20	H	#	Hash / Pound in FIGS
10101	21	Y	6
10110	22	P	0
10111	23	Q	1
11000	24	O	9
11001	25	B	?	Question mark in FIGS
11010	26	G	&	Ampersand in FIGS
11011	27	FIGS	FIGS	Shift to Figures
11100	28	M	.	Period in FIGS
11101	29	X	/	Slash in FIGS
11110	30	V	;	Semicolon in FIGS
11111	31	LTRS	LTRS	Shift to Letters

Frequently Asked Questions

All subsequent characters decode against the wrong lookup table until the next explicit shift code appears. For example, if a FIGS shift is dropped, numeric data renders as gibberish letters. Historically, operators would periodically re-send shift codes to limit the blast radius of such errors - a technique called "unshift on space" (USOS) where a space character automatically reverts to LTRS mode.

Émile Baudot's original 1870 code (ITA1) used a different bit-to-character mapping and was designed for a 5-key keyboard played like a piano chord. ITA2, standardized by the CCITT in 1930, rearranged the mappings for mechanical efficiency in teleprinters. The two are not interchangeable. This tool implements ITA2, which is the version commonly meant when people say "Baudot code" in modern contexts.

No. The ITA2 standard encodes only uppercase Latin letters A through Z. Teleprinters of the era had no concept of case sensitivity. If you input lowercase text for encoding, this tool automatically converts it to uppercase before generating the Baudot sequence.

Codes 00000 (NUL), 00010 (LF), 00100 (SP), and 01000 (CR) are "unshifted" - they produce the same output regardless of the current shift state. This was a deliberate design choice: formatting characters needed to work reliably even after a lost shift code. It also means these codes never require inserting an extra shift character during encoding.

Each Baudot character requires 5 bits versus ASCII's 7 or 8 bits, yielding roughly 37% less bandwidth per character. However, shift codes add overhead. In the worst case (alternating letters and figures every character), overhead doubles the effective bit cost to 10 bits per character. For typical English prose (mostly letters with occasional punctuation), the overhead is approximately 5 - 10%.

Characters outside the ITA2 repertoire (e.g., @, %, lowercase) are flagged and replaced with a placeholder marker. The tool reports unmappable characters in the output summary. For lowercase input, automatic uppercase conversion is applied before encoding.