The Basics
What is ear training?
Ear training is the practice of developing your ability to identify and understand what you're hearing — frequencies, dynamics, space, and texture. Just like sight-reading music or learning an instrument, it's a skill that improves with deliberate, consistent practice.
Experienced mixing engineers don't reach for an analyzer every time they hear a problem — they hear it first. They can tell a 2kHz harshness from a 4kHz one, or notice that a compressor is pumping without looking at a meter. That intuition is built through thousands of hours of critical listening. SonicSandbox exists to accelerate that process.
The core idea is simple: expose yourself to processed audio, make a judgment call, and immediately get feedback on whether you were right. Over time, your brain builds a reference library of sonic signatures. Frequencies start to have colors. Compression starts to have a feel. Space starts to have a shape.
Why It Matters
What better ears actually get you
Faster decisions. When your ears know what to listen for, you spend less time second-guessing and more time making music. You stop asking "is this too bright?" and just know.
Better mixes. Most mixing problems aren't visible on a spectrum analyzer — they're felt. Muddiness, harshness, a kick that doesn't punch, a vocal that doesn't sit. Trained ears catch these faster and solve them with fewer moves.
Stronger communication. When you can describe what you're hearing precisely — "there's a low-mid buildup around 300Hz" or "the attack is too fast, the transients are being buried" — you communicate more effectively with collaborators, clients, and yourself.
Translation across systems. If you can hear how a mix sounds, not just how it measures, you'll catch problems before they appear on other speakers. Great ears are the ultimate reference monitor.
Audio Sources
Train with any audio — including your own music
Every SonicSandbox game supports four audio source modes. Different sources serve different learning goals — and training on music you already know is one of the most powerful things you can do.
A broadband signal with equal energy per octave. Pink noise is the cleanest training signal — nothing to distract you, just the processing itself. Great for learning a concept from scratch before bringing in music.
Short transient sounds like claps and snares. One shots train your ears to catch fast dynamic events — compression artifacts, transient shaping, and frequency coloring all become more obvious on tight percussive material. More sounds coming soon.
Real music loops across genres and arrangements. Training on loops puts your skills in a real-world musical context — you're no longer identifying EQ on a test signal, you're hearing it on a drum groove or a chord progression. This is where fundamentals become instincts.
Upload an entire folder of your own music and the games cycle through your tracks automatically. Training on professional music, loops, or any other audio samples. You can upload one audio file or a folder with many audio files and the games will cycle through the tracks. Use arrow keys to skip and seek through tracks.
The Games
Precision tools. Rapid-fire games.
SonicSandbox games fall into two categories. EQ Match, Compressor, Reverb Master, Distortion Master, and Delay Master are deep, technical tools — fully customizable, with the ability to isolate any individual parameter so you can focus on learning one thing at a time. Freq Quiz, Freq Hunter, Quick Compress, Spatial Specialist, Level Logic, Quick EQ, and Signal Chain Architect are built for speed — one click to answer, instant next round, designed for rapid reps and quick self-quizzing.
The most customizable game on SonicSandbox. A hidden EQ curve is applied to the audio and you drag your own bands to match it by ear. Every aspect of the challenge is configurable: enable one band or several, choose from 6 filter types (peaking, high shelf, low shelf, high pass, low pass, notch), set the dB range, toggle boost-only or cut-only mode, restrict the Hz range, and dial in Q width constraints. For focused study, you can isolate a single parameter — match only the frequency while gain is revealed, or match only the gain while frequency is locked — so you can drill exactly where your ears need work.
What it trains: Frequency identification across the full spectrum, understanding how gain and Q shape a sound, and recognizing how each filter type colors audio differently. The most technically demanding game for anyone serious about EQ literacy.
A target compressor with hidden settings is applied to the audio. You dial in your own compressor — threshold, ratio, attack, release, and knee — to match what you hear, then toggle between your attempt and the target to compare. The game is highly customizable: you can isolate any single parameter, so if you want to drill just attack recognition while the other knobs are revealed, you can. This makes it equally useful for beginners learning what compression does and experienced engineers sharpening their precision.
What it trains: A deep understanding of how each compressor parameter shapes sound — how threshold affects how much gets compressed, how ratio controls the squash, how attack and release time the dynamic response, and how knee smooths the transition into gain reduction.
Two audio clips play in quick succession — one compressed, one dry. Click the one you think is compressed and the next round begins immediately. No setup, no dialing — just listen and decide. It's the fastest way to build instinctive compression recognition through pure repetition.
What it trains: Rapid, instinctive recognition of compression's sonic signature — the reduced dynamic range, softened transients, and slight density that compression adds. Designed for high-rep speed training rather than deep technical analysis.
A reverb with hidden parameters is applied to the audio. You adjust room size, decay time, pre-delay, high-frequency damping, and wet/dry mix to match what you hear, then toggle between your attempt and the target to compare. Like the Compressor, you can isolate any single parameter — practice just room size while the rest are shown, or just pre-delay while everything else is locked. Matching reverb by ear is one of the harder skills in mixing — Reverb Master lets you build it one parameter at a time.
What it trains: Spatial perception and reverb literacy — hearing the difference between a small room and a large hall, recognizing how pre-delay creates separation between a source and its tail, and understanding how damping shapes the character of a space.
A hidden distortion is applied to the audio. Match the drive amount, type (soft clip, hard clip, or tape saturation), mix level, and tone. As with the other deep tools, you can isolate any single parameter — focus purely on identifying distortion type while drive and tone are revealed, or match only the drive amount while the rest are shown. Each distortion type has a distinct harmonic character, and being able to identify them by ear changes how you reach for saturation in a session.
What it trains: Harmonic awareness — distinguishing the smooth warmth of tape saturation from the aggressive edge of hard clipping, judging how much drive is being applied, and hearing how tone shapes the high-frequency character of a distorted signal.
A target delay with hidden settings is applied to the audio. Adjust delay time, feedback amount, and wet/dry mix to match what you hear, then toggle between your attempt and the target to compare. Like the other deep tools, you can lock any parameter to focus on one at a time — isolate delay time while feedback and wet/dry are shown, or guess all three simultaneously for the full challenge. A visual echo diagram reveals the delay structure once you submit.
What it trains: Timing perception and delay literacy — hearing the rhythmic relationship between a delay echo and the source signal, recognizing how feedback multiplies into long trails or tight slaps, and understanding how wet/dry balance places delay in the foreground or background of a mix.
A frequency band is boosted or cut in the audio. Pick the correct frequency from 9 labeled options and the next round begins immediately — no setup, just listen and click. Rounds fire fast, making this the best game for building raw frequency memory through volume. Customizable options let you narrow the frequency range, adjust the boost/cut amount, or toggle between boost-only and cut-only modes.
What it trains: A mental map of the frequency spectrum — learning which frequencies sound muddy, harsh, warm, or airy. This is the fastest path to the core vocabulary of EQ, and the foundation everything else is built on.
A bandpass-filtered sound plays at a hidden position in the stereo field. Click the 2D grid to mark both the frequency (Y axis) and the stereo pan position (X axis) — one click, and the next round fires immediately. The dual-axis format makes this the most challenging of the quick games, but the rapid-fire pacing keeps the reps coming fast.
What it trains: Simultaneously processing two independent sonic dimensions — frequency content and stereo placement. In a real mix these decisions happen together, and Freq Hunter trains you to hear them that way, quickly.
Two games in one, both built for speed. In panning mode, a sound plays at a hidden stereo position — click to place it and the next round begins. In width mode, a sound plays with a hidden stereo width from mono to wide — click your estimate on the slider and move on. One click per round, fast feedback, high volume.
What it trains: Stereo perception — hearing where a sound sits in a mix and how wide or narrow its image is. Critical for panning decisions, mid/side processing, and catching phase issues that collapse the stereo field.
A reference signal plays, then a level-adjusted version follows. Identify whether the gain was boosted or cut, and by how much — select from a set of labeled dB increments. One answer per round, immediate feedback, next round begins instantly. Simple format, high reps, fast results.
What it trains: Volume sensitivity and gain perception — developing an ear for how specific dB amounts actually feel. This skill translates directly to level-matching, gain staging, and trust your instincts when A/B referencing in a session.
Two EQ curves are shown — one applied to the audio, one not. A/B listen between the processed signal and the dry, then click the panel that matches what you hear. No dialing in, no sliders — just listening and deciding. Adjust difficulty, number of bands, and whether the curves are shown or hidden to control the challenge.
What it trains: Recognizing the tonal character of different EQ shapes by ear — learning to hear which frequencies are boosted or cut and whether the effect is subtle or dramatic, without having to reproduce the settings yourself.
A signal chain of two to four effects — EQ, compression, reverb, and saturation — is applied to the audio in a specific order. Listen to the result and drag the effect blocks into slots to reconstruct the order. This isn't about matching parameters; it's about recognizing how the sequence of effects changes the sound.
What it trains: Signal flow intuition — hearing how the order of processing shapes the character of a sound. Compressing before reverb sounds different from compressing after it. EQ before saturation behaves differently than EQ after. This is a fundamental mixing concept that most ear training ignores entirely.
Reference
Key terms, explained
New to audio processing? Here are the core concepts that underpin every SonicSandbox game.
Best Practices
How to get the most out of ear training
The difference between people who develop great ears and people who plateau is almost always about how they practice, not how much.
Pick a game and run a few rounds. Your ears will tell you exactly where to focus next.
Open the games →