How do I get feedback on my cajón playing?

“Am I actually in time, or does it only feel like it?” is the question every self-taught cajón player asks sooner or later, and it is a good one, because the honest answer is that you cannot tell from where you are sitting. This page is about how to find out anyway: how to record yourself so the recording is useful, what to listen for, and where to get somebody else’s ears on it.

Why you cannot judge your own time while you play

Playing takes most of your attention. Part of the rest is planning the next stroke, and part is listening to the box from twenty centimetres away, where it sounds nothing like it does across the room. What is left over is not enough to notice that your snare is arriving a fiftieth of a second early.

There is a second, sneakier problem. Your brain hears what you meant to play. If you intend the snare to land on beat 2, you hear it land on beat 2, and the small, consistent gap between what you intended and what happened is exactly the kind of thing perception smooths over. That is why the most common timing faults are the ones players are certain they do not have.

None of this is a discipline problem. It is the same for everybody, at every level, and it is why professional players record themselves too.

How to record a take that tells you something

A recording made carelessly mostly tells you that phones are bad microphones. A few rules make it worth listening to:

1. Phone about a metre away, facing the front plate. Closer and every bass note overloads the microphone. Further and the room swallows your ghosts. Prop it on a chair or a shelf at roughly the height of the box.
2. Play with a click, and let the phone hear it. A take without a click can tell you whether you were steady, but not whether you were right. With the click in the recording you can hear every stroke against the beat it was aiming for. Play the click out loud from a speaker rather than in headphones, or the recording will not contain it.
3. Play one thing, for 45 to 90 seconds. One groove at one tempo. Long enough to drift, short enough to listen to properly.
4. Do not stop when you make a mistake. Recover and keep going. How you recover is part of what you are checking.
5. Same setup every time. Same spot, same phone position, same tempo. Then this week’s take can be compared with last week’s and the only thing that changed is you.

What to listen for on playback

Listen once all the way through without judging, just to get over the shock of hearing yourself. Then listen again for one thing at a time. These are the four I hear most often, roughly in order.

Rushing the “and”

The ghost between two beats creeping early, so the bar sounds slightly hurried even though the beats themselves are on the click. Count out loud along with the playback, “one and two and”, and listen for whether each ghost lands on your “and” or just before it. Rushing the snare on 2 and 4 is the same fault on a louder stroke.

Flams between the hands

When two hands are meant to make one sound and they land a few milliseconds apart, you hear a smeared double, “fl-am”, instead of one clean note. It also shows up in alternating patterns as a gallop: right-left, right-left, with one hand consistently a little late. Listen for whether the hand pattern sounds even or limps.

A bass that thuds

A short knock with no low ring after it. Nearly always the hand staying on the plate after the stroke. Phones under-record low end, so judge depth by comparing takes rather than in absolute terms. The fixes are in how to get a deep bass tone.

Ghosts as loud as the snare

The most common one of all. On the recording, the bar sounds like eight equal knocks instead of a groove with a backbeat. A ghost should be clearly quieter than the bass and the snare, and still there on the recording. If you cannot tell which strokes are the backbeat, the ghosts are too loud; if you cannot find them at all, they are too quiet. See ghost notes for how quiet is quiet.

Also listen to the end against the beginning. A take that starts on the click and ends ahead of it has sped up, and it is the fault players are least likely to believe until they hear it.

Where to get another pair of ears

• A teacher. The best option if you have one. Send them a recording between lessons, not just your playing in the room, because a recording is what an audience hears.
• A friend who plays. Ideally a drummer or any musician who plays in bands. Ask them a specific question, “is my snare early?”, rather than “what do you think?”, and you will get a specific answer instead of politeness.
• Forums and groups. Generous, and occasionally excellent. The catch is that comments on a phone recording tend to be general (“sounds good, keep practising”), and you rarely get the same listener twice, so nobody follows your progress.
• Me. I measure every stroke of a take against the click and give you one thing to work on. I am an AI coach, not a person, and I only hear the audio, so I cannot see your hands. What I can do is listen to every take, every evening, and compare it with the last one.

What my notes on a take look like

Here is the kind of thing I send back after a take of the groove, boom, tick, crack, tick, at 75 BPM. This one is an example rather than a real player’s take, but it is typical:

• Overall: 84 of 112 strokes within 30 milliseconds of the click, which is 75%. A solid take with one clear pattern in the misses.
• What was good: the bass on 1 and 3 was within 10 milliseconds almost every time, and the tempo held from the first bar to the last.
• What to fix: your snare on 2 and 4 is landing about 25 milliseconds early, consistently. You are leaning into the backbeat. The ghosts after each snare are early by the same amount, because they are following it.
• Tonight: drop to 65 and play only bass and snare, no ghosts, like exercise 2-3. Think of the snare as arriving late and relaxed. Then record the groove again and see whether the early snares come down.

That is the whole idea: a number that describes your playing, the one fault that matters most, and a thing to do about it tonight, rather than a list of everything that could be better. One fix at a time is how timing actually improves.