Converting English into Morse code is a lookup, not a calculation. Every letter,
digit and punctuation mark has exactly one pattern. Replace each character with
its pattern, put a single space between letters and a slash between words, and
you are finished. SEND HELP comes out as ... . -. -.. / .... . .-.. .--..
Three things about written English do not survive the trip. Capital letters go, because Morse has one form of each letter and no way to mark case. Accents go, because the international code covers the 26 plain Latin letters and stops there. And anything with no entry in the table at all, such as an emoji or a percent sign, has no correct output to give.
Write the code with a full stop and a hyphen from your own keyboard. A middot or a typographic dash looks tidier on the page and fails the moment somebody pastes it into a tool, which is the first thing anyone does with a converted message.
Want to hear the message as well as read it? MorseLab plays any English text back as a tone, a flashlight blink or a vibration, free and offline.
The four steps
Every conversion, from one word to a full paragraph, is these four steps in order.
- Write the message in capitals. Not strictly needed, since the code has no case, but it stops you hunting for a capital pattern that does not exist.
- Replace each character with its pattern. One lookup per character. Letters, digits and punctuation marks are all treated the same way.
- Put one space between letters.
- Put a slash between words, with a space on each side.
Worked through on SEND HELP:
S ...
E .
N -.
D -..
/
H ....
E .
L .-..
P .--.
Run together, that is:
... . -. -.. / .... . .-.. .--.
One thing to fix in your head before the tables. A punctuation mark is a character, not a separator. It sits inside the word it belongs to and gets the same single space either side that any letter gets. Only the slash means a new word.
The table you convert from
This is International Morse code, defined by ITU-R Recommendation M.1677-1. It is the only version in use, and a converter holds nothing more than this.
Letters
| Letter | Code | Letter | Code |
|---|---|---|---|
| A | .- | N | -. |
| B | -... | O | --- |
| C | -.-. | P | .--. |
| D | -.. | Q | --.- |
| E | . | R | .-. |
| F | ..-. | S | ... |
| G | --. | T | - |
| H | .... | U | ..- |
| I | .. | V | ...- |
| J | .--- | W | .-- |
| K | -.- | X | -..- |
| L | .-.. | Y | -.-- |
| M | -- | Z | --.. |
No letter runs longer than four signals. E is a single dot and T a single dash, which is why ordinary English text converts to something shorter than you would guess.
Digits
Every digit is exactly five signals, and they follow a rule you can rebuild from memory: 1 to 5 add dots from the left, one at a time, until 5 is all dots, and 6 to 0 mirror that, adding dashes from the left until 0 is all dashes.
| Digit | Code | Digit | Code |
|---|---|---|---|
| 1 | .---- | 6 | -.... |
| 2 | ..--- | 7 | --... |
| 3 | ...-- | 8 | ---.. |
| 4 | ....- | 9 | ----. |
| 5 | ..... | 0 | ----- |
The punctuation English writing uses
Punctuation marks have patterns too, and they run longer than letters, five to seven signals each. These eight cover almost everything you will type:
| Mark | Code | Mark | Code |
|---|---|---|---|
Full stop . | .-.-.- | Exclamation ! | -.-.-- |
Comma , | --..-- | Slash / | -..-. |
Question ? | ..--.. | At @ | .--.-. |
Apostrophe ' | .----. | Hyphen - | -....- |
The rest of the standard's marks, which turn up far less often:
| Mark | Code | Mark | Code |
|---|---|---|---|
Quotation " | .-..-. | Colon : | ---... |
Open bracket ( | -.--. | Semicolon ; | -.-.-. |
Close bracket ) | -.--.- | Equals = | -...- |
Ampersand & | .-... | Plus + | .-.-. |
Underscore _ | ..--.- | Dollar $ | ...-..- |
Those two tables and the two above them are the whole of what a converter holds. For the same chart as a block you can copy into a note and keep open while you work, see the Morse code alphabet chart.
Four messages, converted
Each one adds a complication that English throws at you.
One word
ENGLISH, letter by letter, joined with single spaces:
E .
N -.
G --.
L .-..
I ..
S ...
H ....
ENGLISH = . -. --. .-.. .. ... ....
A short sentence
Three words, so two slashes.
WHERE ARE YOU
.-- .... . .-. . / .- .-. . / -.-- --- ..-
A sentence with punctuation
The comma belongs to the 6 and the full stop belongs to WATER, so each one takes a letter gap, not a slash.
MEET AT 6, BRING WATER.
-- . . - / .- - / -.... --..-- / -... .-. .. -. --. / .-- .- - . .-. .-.-.-
An email address
An address has no spaces in it, so the entire thing is one Morse word with no slash anywhere. The at sign and the full stop are just two more characters.
BOB@MAIL.COM
-... --- -... .--.-. -- .- .. .-.. .-.-.- -.-. --- --
If you send an address by ear, expect to repeat it. Nothing in the code marks where the domain starts, and a listener has only the letter gaps to work from.
English is full of double letters, and they are where conversions break
Doubled letters are everywhere in English: HELLO, MEET, DINNER, SORRY, LETTER, COFFEE. The gap between two letters is part of the message. Close it up and you have not written the same word sloppily, you have written a different word.
| English | Correct | Run together | Reads as |
|---|---|---|---|
| MEET | -- . . - | -- .. - | MIT |
| DINNER | -.. .. -. -. . .-. | -.. .. -.-. . .-. | DICER |
| BETTER | -... . - - . .-. | -... . -- . .-. | BEMER |
| SEE | ... . . | ... .. | SI |
| SORRY | ... --- .-. .-. -.-- | ... --- .-..-. -.-- | S, O, a quotation mark, Y |
| HELLO | .... . .-.. .-.. --- | .... . .-...-.. --- | H, E, a run that is no character, O |
The pairs that hurt are the ones that merge into something the reader can still
decode. Two Es run together are .., which is I. Two Is are ...., which is H.
Two Ts are --, which is M. Two Ns are -.-., which is C. Two Ks are -.--.-,
which is a closing bracket. Two Rs are .-..-., which is a quotation mark. Two
Hs are the dangerous pair: ........, eight dots, is the error signal, so a
merged WITHHOLD tells your reader to throw away the word before it. Every other
pair collapses into a run that is not a character at all, and a reader simply
stops and asks you to repeat.
The same rule explains the one string everybody has seen. SOS is a prosign, sent
unbroken as ...---... with no letter gaps inside it. Written as three separate
letters it is a longer, different signal, and it never stood for "Save Our
Souls". That was invented afterwards to explain a pattern chosen because it is
unmistakable. There is more in
SOS in Morse code.
What English carries that Morse cannot
| In your English text | In the standard | What to do |
|---|---|---|
| Capitals | Nothing marks case | Ignore it. Help and HELP convert the same |
café, naïve, résumé | No accented letters | Strip the accent: CAFE is -.-. .- ..-. . |
| Percent, asterisk, hash | No code | Write the word, or leave it out |
| Pound and euro signs | Only the dollar $ | Spell the currency: .--. --- ..- -. -.. ... for POUNDS |
| Emoji | No code | Drop it |
| Line break, tab | No character | Use the prosign BT, -...-, for a new paragraph |
| Double spaces | One word gap only | Collapse them to a single slash |
Anything outside the Latin alphabet is the same problem written larger. Arabic, Japanese, Devanagari and Chinese have no patterns in the international code, so romanise the text yourself first. That way you decide what gets sent rather than leaving it to whatever the tool happens to do with a character it cannot find.
Numbers: digits or words?
Both convert cleanly, so this is a choice about length. Every digit is five signals, which sounds long until you count what the spelled-out word costs. Length here is measured in units, where one unit is the length of a dot.
| Number | As digits | Units | As a word | Units |
|---|---|---|---|---|
| 3 | ...-- | 13 | - .... .-. . . | 31 |
| 7 | --... | 13 | ... . ...- . -. | 33 |
| 8 | ---.. | 15 | . .. --. .... - | 35 |
| 20 | ..--- ----- | 37 | - .-- . -. - -.-- | 49 |
| 10 | .---- ----- | 39 | - . -. | 15 |
For every single digit the figure wins, usually by a wide margin, and it is easier to catch by ear because the five-signal shape announces that a number is coming. The last row is the exception worth knowing. A zero costs 19 units on its own, so 10 runs to 39, while TEN is only 15, because T, E and N are three of the cheapest letters in the code. The same flip happens at 100, which is 61 units against 59 for HUNDRED.
So: digits by default, and spell out a round number that is mostly zeros.
How long the converted message takes to send
Morse timing is arithmetic. Everything is measured in units, and the ratio never changes.
| Element | Units |
|---|---|
| Dot | 1 |
| Dash | 3 |
| Gap between signals inside one character | 1 |
| Gap between characters | 3 |
| Gap between words | 7 |
A dash is not "a bit longer" than a dot. It is exactly three times as long. Operators say the two marks aloud as dit and dah, which is worth knowing the first time you hear a converted message rather than read one.
To price a message, add up the units, then use the one formula that matters:
Dot length in seconds = 1.2 ÷ WPM.
| WPM | Dot | Dash | Gap between letters | Gap between words |
|---|---|---|---|---|
| 5 | 240 ms | 720 ms | 720 ms | 1,680 ms |
| 12 | 100 ms | 300 ms | 300 ms | 700 ms |
| 20 | 60 ms | 180 ms | 180 ms | 420 ms |
| 25 | 48 ms | 144 ms | 144 ms | 336 ms |
SEND HELP works out at 71 units: 27 for SEND, 7 for the word gap, 37 for HELP. Multiply by the dot length and you have the sending time.
| Message | Units | At 5 WPM | At 12 WPM | At 20 WPM |
|---|---|---|---|---|
| HELLO | 49 | 11.8 s | 4.9 s | 2.9 s |
| SEND HELP | 71 | 17.0 s | 7.1 s | 4.3 s |
| ARE YOU OK? | 111 | 26.6 s | 11.1 s | 6.7 s |
| WHAT IS YOUR NAME | 139 | 33.4 s | 13.9 s | 8.3 s |
| MEET AT 6, BRING WATER. | 195 | 46.8 s | 19.5 s | 11.7 s |
That last line is the practical lesson in converting English. A five-word sentence takes the better part of a minute at a beginner's speed. Short words and dropped punctuation are not laziness, they are how the medium works. The full arithmetic, including why the reference word PARIS is exactly 50 units, is in Morse code timing and WPM.
Check the conversion before you send it
Three tests catch nearly every error, and they take under a minute.
Count the characters. Count the characters in your English message, spaces left out, then count the groups in your Morse, treating each run between spaces or slashes as one group. The two numbers must match. SEND HELP has eight characters and eight groups.
Read it back cold. Decode your own string from the table as though you had never seen the English, and see whether you get the same message. This is the test that catches a merged double letter, because the read-back gives you DICER instead of DINNER.
Check every gap you typed. One space between letters, / between words,
nothing at all between the signals inside a character. .- is A. . - is E
then T. They are different messages and they look almost identical.
I built MorseLab partly to make that read-back instant: type or paste English in one box and the Morse appears in the other, then play it as a tone, a screen flash, a flashlight blink or a vibration at whatever speed you choose. It converts the other way too, decoding live Morse audio through the microphone and reading printed text through the camera, and it carries the chart, the prosigns and the Q-codes for when you want to check a character. No account, no signal needed, 16 languages.
Common mistakes
Merging a double letter. The commonest error there is, because English is full of doubles. MEET arrives as MIT, DINNER as DICER.
Prettified dots and dashes. Typographic bullets and en dashes look right and decode nowhere. Retype them as a full stop and a hyphen.
Sending the slash as a character. The slash is a writing convention for a
word gap. The punctuation mark / is -..-., and sending that between words
puts a stroke into your message.
Treating punctuation as a word break. A comma or a full stop takes a normal
letter gap. MEET AT 6, keeps the comma attached to the 6.
Adding a full stop that was not there. .-.-.- at the end of a name or a
single word will be written down by whoever reads it.
Assuming an accented letter has a code. It does not, in the international standard. Strip the accent and send the plain letter.
Expecting the conversion to teach you the code. Converting is a lookup and you can do it today. Reading Morse by ear is a different skill that takes weeks of short daily sessions. Anyone promising fluency in a day is selling something. How to learn Morse code sets out what a real plan looks like.
Questions people ask
How do I translate English into Morse code?
Look each character up in the table, put one space between letters and a slash
between words. HELLO WORLD is .... . .-.. .-.. --- / .-- --- .-. .-.. -...
There is exactly one correct answer per character, so a tool and a careful hand
produce the same string.
Does capitalisation matter in Morse code?
No. There is one pattern per letter and nothing to mark case, so Help, HELP
and help all convert to .... . .-.. .--.. Case is lost and cannot be
recovered by anyone decoding it, which is why charts are printed in capitals.
How do I write a space between words in Morse code?
With a slash, written as a space, a forward slash, then a space. In sound it is
a gap of seven units, against three units between letters. If you drop the word
gaps, .- - -- .- could be AT MA or ATMA or several other readings.
What about apostrophes in words like don't?
The apostrophe has a code, .----., and it is sent as a character with normal
letter gaps: DON'T is -.. --- -. .----. -. Most operators drop it and send
DONT, -.. --- -. -, because the meaning survives and the message is 22 units
shorter.
Can Morse code handle accented English words like café?
Not the accent. The international standard has patterns for the 26 plain Latin
letters only, so send CAFE, -.-. .- ..-. ., and let context carry the rest.
The same goes for names and loanwords: strip every accent before you convert.
Should I write numbers as digits or spell them out?
Use digits. Each one is five signals, which is always shorter than the word: 8 is 15 units against 35 for EIGHT. Digits are also easier to catch by ear, because the five-signal shape tells a listener a number is coming.
Is there a Morse code for the percent or euro sign?
No. The standard's punctuation set is short, and the only currency symbol in it
is the dollar $, which is ...-..-. For anything else, spell the word: send
PERCENT or EUROS as letters.
Why does my converted message decode as something else?
Almost always spacing. Either a double letter has been merged, or the gaps have collapsed somewhere in copying, or a character has picked up an extra space and split in two. Read your own string back from the table and the fault shows up at the exact character where the two readings diverge.
Can I convert English into Morse code without the internet?
Yes. A converter needs the lookup table and nothing else, so there is no reason for one to require a connection. An app holding the table locally will convert, play the tone and blink the light with no signal at all, which matters if the reason you are sending Morse is that nothing else is working.
Download it: MorseLab for iPhone and Android. Free, and the link picks the right store for you.
Keep reading
- Morse code to English — the same table read backwards, and what to do when the spacing is missing.
- The Morse code alphabet chart — every letter, number and mark in one reference.
- Morse code numbers and punctuation — the five-signal digit rule and every mark in the standard.
- Morse code generator — turning the converted string into tone, light and vibration.
- How a Morse code translator works — the mechanism behind both directions.



