Koch vs Farnsworth — which method to learn Morse?
The two reference Morse pedagogical methods, compared in detail. When to use which, how to combine them, and why Mission Morse fuses them.
If you're looking up how to learn Morse, you'll run into two names: Koch and Farnsworth. They're the two most-cited pedagogical methods in the amateur radio world for 70 years. Are they really different? Should you pick one over the other? Spoiler: the best learners use both, and Mission Morse fuses them by default. Here's why.
The Koch method — learning at full speed
Ludwig Koch was a German neuropsychologist who studied Morse learning in the 1930s. His 1936 thesis demonstrates, with experiments, that Morse is learned best at full speed (12-15 wpm), not slowly.
To understand why, you need to know how your brain encodes a Morse signal:
- At 5 wpm (5 words per minute), a dot lasts 240 ms and a dash 720 ms. You have plenty of time to count mentally "one, two, three" to identify the length. You encode the code as a sequence of distinct symbols.
- At 15 wpm, a dot lasts 80 ms and a dash 240 ms. You no longer have time to count — you must recognise the letter as a global sound, like a musical chord.
Koch's observation: a learner moving from 5 to 15 wpm has to unlearn the mental counting and re-learn everything by ear. It's slow and demoralising. Better to suffer for three days at 15 wpm than redo everything in six months.
The original Koch protocol:
- Learn 2 letters at 18 wpm character.
- When you reach 90% accuracy on those 2, add a 3rd letter.
- Repeat until the full alphabet is in.
- At no point do you slow the character speed below 18 wpm.
It's hard. The first hours are frustrating. But by the end of the first week, you recognise E, T, A, I as global sounds — not as symbol sequences.
Farnsworth timing — real letters, fake gaps
Donald Farnsworth, an American amateur, published in 1959 in QST (the ARRL magazine) a subtle refinement of the Koch protocol. His critique: at full 18 wpm, gaps between symbols, letters and words are also at 18 wpm, which doesn't give the brain time to identify the letter before the next one arrives.
His fix:
- Keep characters fast (effective speed = 18 wpm). Each letter keeps its "real" sound.
- Lengthen gaps between letters and words to bring overall speed down to 5-8 wpm. The brain has time to identify each letter before the next arrives.
Why it's brilliant: you learn the "right" sound of each letter from the start, without falling into the mental-counting trap. As you progress, you shrink the gaps without touching character speed. You go from 8 wpm overall to 18 wpm overall while keeping the same letter sound — your brain has always learned the "right" version.
It's a teaching genius move. Farnsworth turned a hard method (pure Koch) into a method that's both effective AND practical.
The comparison table
| Criterion | Pure Koch | Farnsworth |
|---|---|---|
| Character speed | 18 wpm from day 1 | 18 wpm from day 1 |
| Overall speed | 18 wpm from day 1 | 5-8 wpm at first, 18 later |
| Inter-letter gap | Standard (3 dots) | Stretched (10-15 dots) |
| Initial difficulty | Very high | Moderate |
| Final-sound fidelity | Maximum | Maximum (gaps shrink progressively) |
| Risk of bad anchoring | Very low | Very low |
| Total estimated time | 4-5 weeks | 4-6 weeks |
Intuitive conclusion: Koch is faster in theory but emotionally harder. Farnsworth is slightly longer but much more motivating.
Why almost everyone uses Farnsworth today
Three reasons:
Drop-out is enemy number one. A method requiring 4 weeks but where 30% of learners quit after 5 days is, on average, less effective than one requiring 5 weeks where 90% finish.
The end result is identical. Whether you learned at 18 wpm overall from day 1 or you went through 8 wpm overall for two weeks, your ear hears the same sounds at the end. The difference is cosmetic.
Farnsworth is compatible with partial practice. You can do 5 minutes of Morse at 18/8 wpm every day without entering "total focus" mode the way pure Koch demands. Crucial when you learn alongside a full-time job.
The Mission Morse approach — Koch + Farnsworth + 5 styles
Mission Morse fuses both methods by default:
- Character speed: 12 wpm (between 12 and 18, configurable). Follows Koch.
- Overall speed: 8 wpm at first (gradually shrunk to 16). Follows Farnsworth.
- Learning order: 2 new letters per step, by usage frequency and code simplicity. Follows Koch (not alphabetical).
- Learning modalities: 5 styles (Visual, Audio, Mnemonic, Rhythm, Story) that adapt to your cognitive encoding mode. Mission Morse innovation.
This combination lets beginners clear the alphabet in 4-6 weeks without quitting, while landing directly on the "right" letter sounds.
When is pure Koch preferable?
Three cases where we recommend pure Koch without Farnsworth gaps:
You want speed-running. If your goal is 35-40 wpm for DX contests, never get used to long gaps. Pure Koch puts you in race condition from day 1.
You already learned Morse at 5 wpm (e.g. in scouts 20 years ago) and you struggle to speed up. Pure Koch at 20 wpm for 3 weeks is the best way to reset your ear.
You learn in a group with a coach who keeps motivation up. The solo frustration of pure Koch disappears in groups.
In all other cases (~95% of beginners), Farnsworth is the right choice.
What about the Russian (Petrenko) method?
For completeness, let's mention the Petrenko method (USSR, 1960s): very close to Koch but focused on continuous written transcription rather than letter-by-letter recognition. The learner writes letters in the margin without thinking, the brain recognises them with a 1-second delay.
It's effective to reach 25-30 wpm very quickly, but requires fast-writing training. Rarely used outside the original military context.
The PARIS standard timing
To understand wpm (words per minute) figures, an international convention called PARIS: a standard "word" is defined as the length of "PARIS" in Morse, i.e. 50 dot-units (50 dot-times). At 12 wpm, you send 12 × 50 = 600 dot-times per minute, i.e. one dot every 100 ms.
This convention lets you compare speeds across apps and operators. When a Mission Morse quiz shows "12 wpm character / 8 wpm overall", it means characters are sent at PARIS 12 wpm, but inter-letter pauses are stretched to bring the total down to 8 wpm.
How to set your speed
Mission Morse lets you configure character and overall separately in audio settings. Recommendations:
- Complete beginner (week 1): 12/8 wpm.
- Regular learner (weeks 2-4): 14/10 wpm.
- Alphabet validation (week 5): 16/14 wpm.
- Pre-QSO: 18/16 wpm.
- Comfortable QSO: 20/18 wpm.
- DX contest: 25-30/25-30 wpm.
Don't be in a rush to climb. A 2-wpm increase every 7-10 days, while staying at 80% accuracy, is faster long-term than a 5-wpm leap where you miss 50% of questions.
The trap of "Morse apps" that don't respect Farnsworth
Many free Morse apps on Android and iOS stores make this mistake: they let you drop character speed to 5 wpm. Result — the learner thinks they're progressing while memorising a false letter sound that they'll have to entirely unlearn when speeding up.
Mission Morse caps character speed at 12 wpm minimum by pedagogical design, even if it frustrates some beginners at first. A choice that follows Koch and pays long-term.
To go further
- Mission Morse method — detailed explanation of the 13 Koch steps.
- Learning Morse in 2026 — full beginner guide.
- Top 10 mnemonics — accelerate first-letter memorisation.
▶ Start with Koch + Farnsworth pre-configured — free, guest mode available.