Spell text out in the radiotelephony alphabet of ICAO Annex 10 Volume II, with the Alfa and Juliett spellings the Recommendation actually gives, and read a spelled message back
Initializing in your browser…
Encode and decode text, files and JSON as base64, base64url or hex, checked by decoding the result back
Percent-encode and decode against four character sets, build query strings, and take a URL apart
Generate QR codes and check each one scans by decoding it back before you download it
An aircraft registration and a serial number have to be read over a bad line, and the three letters that trip up every spelling list are all in them.
Text
N123AB, then AJX
Spelled out, with the pronunciations
N123AB N November no VEM ber 1 One WUN 2 Two TOO 3 Three TREE A Alfa AL fah B Bravo BRAH voh 6 code words, about 6 seconds to read aloud AJX A Alfa AL fah J Juliett JEW lee ETT X X-ray ECKS ray Reading "Alpha Juliet X-ray" back: AJX "Alpha" was read as Alfa. Annex 10 spells it Alfa, and the difference is deliberate. "Juliet" was read as Juliett. Annex 10 spells it Juliett, and the difference is deliberate.
Two things here are worth the trouble. Three is written Three and said TREE, because the th sound does not survive a noisy channel and is not in every speaker's language; five is FIFE and nine is NIN-er for the same kind of reason, and four is given two syllables so it cannot be heard as "for". And A is Alfa, not Alpha, and J is Juliett, not Juliet: Annex 10 spells them that way because "ph" is not read as an f in most languages that use this alphabet and a single final t is silent in French. An alphabet meant to be read aloud by people who do not share a first language is exactly where those two spellings matter, and they are the two that lists on the internet reliably get wrong. Both wrong spellings are still accepted when reading a message back, because someone will have read one to you, and the page says which one it was.
Spell a word, a registration or a confirmation code out in the radiotelephony alphabet, and read a spelled message back. The table comes from ICAO Annex 10 Volume II, Table 5-1, which means A is Alfa and J is Juliett, and every entry says where it comes from.
This tool has a small accuracy surface and it is worth saying so plainly rather than dressing it up: the alphabet is 26 letters and 10 figures, there is no arithmetic in it, and the only things a tool can get right or wrong are the entries themselves and what it does either side of them. So the whole of the work here is the table and the two conversions.
Two entries are usually wrong, and they were wrong here before this pass. **A is Alfa, not Alpha, and J is Juliett, not Juliet.** Both spellings are deliberate and both are in Table 5-1. Alfa is spelled with an f because "ph" is not read as an f in most languages that use the Latin alphabet, so a French or an Italian speaker reading "Alpha" would say it wrong. Juliett has two t's because a single final t is silent in French. An alphabet designed to be read aloud by people who do not share a first language is exactly the place where those two spellings matter, and they are the two that spelling lists on the internet reliably get wrong.
The figures are the other half, and they were already right here. Three is TREE, four is FOW-er, five is FIFE and nine is NIN-er. Those four are not stylistic: the th in "three" does not survive a noisy channel and is not in every speaker's language, "five" is too close to "fire", "nine" is too close to the German "nein" and to "no", and "four" is given two syllables so it cannot be heard as "for". A tool that spells the numbers normally is wrong for the exact purpose the alphabet exists for.
Reading a message back is tolerant on purpose and honest about it. Any separator works, a pronunciation in brackets is ignored, and the spellings Annex 10 does not use are accepted too, because someone read a code to you from a tool that had them wrong and refusing it helps nobody. When one of those is used, the page says which spelling was given and which one Annex 10 gives. What it will not do is guess: a word that matches nothing in the table is named rather than dropped, because a message with one word misheard coming back a character short is exactly the failure the alphabet exists to prevent.
The decimal point is in Table 5-1 and is spoken as "decimal", not "point". Spoken names for the other punctuation a serial number contains, the hyphen, the slant, the at sign, are common practice rather than Annex 10, and every one of them is marked as such in the chart and flagged when it is used. Anything the table does not cover at all is named rather than silently dropped.
The project's tests write Table 5-1 out again from the Recommendation and compare it entry by entry: 203 module assertions and 151 browser assertions.
N123AB becomes November One Two Three Alfa Bravo: six code words, about six seconds to read aloud at a deliberate pace.
AJX becomes Alfa Juliett X-ray, which is the entry every list on the internet gets wrong twice.
A confirmation code or a serial number spelled out, with an estimate of how long it will take, which is usually longer than people expect.
The table is the one Annex 10 gives, including the figure pronunciations, which is what an examiner will expect.
Paste what was read to you and get the characters back, with anything you may have misheard named rather than quietly lost.
The chart says which spelling is in the Recommendation and why, with both accepted on the way back in.
Alfa. ICAO Annex 10 Volume II, Table 5-1 spells it with an f, because "ph" is not read as an f in most languages that use the Latin alphabet and a speaker of one of them would say "Alpha" wrong. Alpha is accepted when reading a message back, and the page points out that it is not the Recommendation spelling.
Juliett, with two t's, because a single final t is silent in French. The same reasoning as Alfa: the alphabet is meant to be read aloud by people who do not share a first language.
Because the th sound does not survive a noisy channel and is absent from many speakers' languages, and because "five" and "fire" are too easy to confuse. Nine is NIN-er because it is too close to the German "nein" and to "no", and four is FOW-er so it cannot be heard as "for". Four of the ten figures are read differently from the ordinary English word, and those four are the reason the figure table exists.
The page gives spoken names for them, and marks them as common practice rather than Annex 10, which covers only the 26 letters, the ten figures and the decimal point. The decimal point is spoken as "decimal", not "point".
It is the same table. NATO adopted the ICAO alphabet in 1956, and civil aviation, maritime radio, the ITU and NATO have used it since. The 1941 Joint Army/Navy alphabet, the one with Able, Baker and Charlie, is a different and older thing.
Because a word it did not recognise was turned into an empty string and dropped. Here an unrecognised word is named, which is the only way to notice that something was misheard.
Conversions run on your device in JavaScript. The values you enter are never sent over the network.