Ear Gym: Free Ear Training for Music Producers

Most mixing problems are listening problems: you cannot fix a muddy low-mid or an over-squashed drum bus until you can hear it. Ear Gym is a free ear training tool for music producers that runs in the browser. Each round plays an original and a processed version of the same audio, and your job is to name the move: which EQ, which compressor setting, which filter, how much saturation, wider or narrower.

Open Ear Gym

How a round works

Pick a workout from the tabs at the top: EQ, Compression, Filter, Saturation or Stereo Width. Choose a track from the pack selector, then a source: Full Mix, Drums or Instruments. Press play and flip between A (Original) and B (Processed) as often as you need.

Below the player sit 4 answer cards, each with a small curve and the exact setting, such as “-3 dB · 150 Hz, low shelf cut”. Select one and press Lock In Answer. The feedback panel then tells you whether you were right and shows the move that was actually applied, so a miss still teaches you something. Show Hint reveals a one-line listening cue, and Skip moves on without counting the round.

The keyboard is faster than the mouse once you settle in:

• A / B: original / processed
• Space: play / pause
• 1-4: select an answer
• Enter: lock in
• S: skip

EQ: frequency and direction

The EQ pool covers the moves you make every session: high-shelf air boosts between 6 and 12 kHz, high-shelf cuts that tame the top, low-shelf boosts and trims between 80 and 180 Hz, mud cuts around 220 to 420 Hz, presence boosts between 1.8 and 4.5 kHz, and narrow notch cuts. Gains stay small, mostly 2 to 5 dB, which is the range where real decisions happen.

The wrong answers are not random. They differ in shape, direction, or by roughly an octave or 3 dB, so the first thing to settle is direction: did the processed version gain or lose energy? Then ask where. A common beginner mistake is hearing “brighter” and jumping to an air boost when the low end was actually trimmed. Try this: play the Full Mix, decide boost or cut before you look at the cards, and only then read the frequencies.

Compression: transients, not loudness

Compression rounds use four families: gentle glue at 2:1 or 3:1, slow attack at 3:1 or 4:1 that lets the front of the hit through, fast attack at 4:1 or 6:1 that clamps the transient, and hard limiting at 8:1 or 10:1.

Make-up gain is applied, so the processed version is not simply quieter. The usual trap is picking whichever one sounds louder. Listen to the first few milliseconds of each drum hit instead, then to the tail. The Drums source is the clearest place to start: the difference between a kept snare crack and a flattened one is much easier to hear without pads and bass on top.

Filter: what disappears

Filter rounds are a good warm-up. The options are a high-pass between 90 and 220 Hz, a low-pass between 7 and 13 kHz, and a band-pass centred between 600 Hz and 1.4 kHz. The skill is noticing what is missing rather than what changed. A high-pass takes kick body and sub weight; a low-pass softens hi-hats and air; a band-pass narrows everything into a telephone-like middle. Once you can name the type quickly, the harder question is the cutoff, so compare a 90 Hz high-pass with a 220 Hz one on the Drums source and listen for where the kick’s body starts to thin.

Saturation: harmonics, not volume

Saturation has five steps, from soft saturation with light drive to hard saturation with heavy drive. The app pulls the output down as the drive goes up, so the more distorted version is not automatically the louder one. Listen for density in the midrange and for edge on transients. The beginner mistake is judging only the extremes: anyone can hear heavy drive, but telling soft-light from soft-medium is where the training value is. The Instruments source tends to show harmonic thickening more clearly than a full mix.

Stereo width: centre versus sides

Width rounds either widen the image (subtle, medium, heavy) or narrow it (subtle, medium, heavy). None of the narrowing options collapses to mono, so you are judging degrees, not on and off. Use headphones here; laptop speakers make most of these moves inaudible. Listen to what happens at the edges of the image and to whether the centre gets stronger or weaker. A useful exercise is to stay on one source and run only width rounds until you can say “wider or narrower” before you play B a second time.

Real tracks, three sources

The audio in Ear Gym comes from tracks built from Ghost Syndicate sample packs, and new packs are being added over time. Each pack gives you a full mix plus separate drum and instrument material, which matters for practice: the same compression move reads very differently on a drum loop than on a pad. If a track becomes too familiar, switch packs rather than memorising one loop. When a pack catches your ear, its page is one click away through the “Buy” link next to the artwork.

A practice routine that sticks

Ear Gym sets a daily goal of 20 rounds and a goal of 10 rounds per category. Each correct answer earns 40 XP, and the sidebar tracks accuracy, current and best streak, average answer time, and accuracy by category. Progress is not saved between visits, so treat it as a session tool and keep the tab open while you train.

A workable routine takes 10 to 15 minutes:

1. Warm up with 10 Filter rounds.
2. Spend 10 rounds on your weakest category, using the hint for the first few.
3. Repeat that category with hints off and watch the average time drop.
4. Once accuracy holds, switch from Full Mix to Drums or Instruments, or to a different pack.

The accuracy-by-category panel shows where to spend the next session. Then take one move you keep missing and recreate it in your own project with a stock EQ or compressor. That last step is what connects the exercise to your mixes.

How it compares with other ear trainers

Dedicated ear trainers often use pink noise or test tones and ask you to identify exact frequencies. That approach goes deeper on pure frequency recognition, and many of the more complete options are paid. Ear Gym is narrower and deliberately practical: multiple choice, a fixed set of common moves, and real bass-music material instead of noise. It needs no install and no account. The limits are the ones mentioned above: answers come from a fixed pool rather than free entry, and stats reset when you reload the page. As a companion to a test-tone trainer, it covers the part those tools tend to skip: hearing a move inside music.