Vocal Mixing Ear Test
Vocal Mixing Ear Test: 7 A/B Rounds on Real Vocals
7 quick A/B rounds on real dry vocals. Hear what a de-esser, 3 kHz, 150 Hz, compression, reverb, delay and 6 dB actually sound like. No certificate, no sign-up — just your ears. Every version is rendered in your browser from the same four short takes, so the only thing that changes between A and B is the one move you are listening for.
How it works. Each round plays the same vocal phrase twice: one version is dry, the other has a single textbook-sized move applied — a de-esser-style cut, a 3 kHz boost, a 150 Hz high-pass, compression, reverb, a slap delay, or just more level. You pick which one is processed. Listen to A and B as many times as you like before answering; a one-line listening hint appears after each answer.
Put on headphones if you can. Laptop speakers physically cannot reproduce two of the rounds. The four clips were recorded by IvanFromRnD (about 1.5 s each, dry, no processing).
Nothing loads or plays until you tap Start.
Nothing leaves the browser. Decoding, processing and playback use the Web Audio API on your device; there is no server, no microphone access and no tracking. The only network request this page makes is fetching the four demo clips when you tap Start. Nothing autoplays.
| # | Question | You said | Result | Replay |
|---|
Laptop speakers hide rounds 1, 3 and 5 — use headphones. The processing here is a textbook amount so the difference is learnable; real mixes are subtler, often half of these moves or less. A bad result says nothing about talent, it says "train this": the rounds you missed are the ones to loop on your own takes this week.
- How to de-ess vocals without making them lisp
- Where to set the high-pass filter on a vocal
- Vocal compressor attack and release, explained by ear
- Vocal reverb settings that stay out of the way
- Slap delay on vocals: the 80–120 ms trick
- Air band EQ vs presence: what 3 kHz and 12 kHz each do
You now know which moves your ears already catch and which ones they do not. The Vocal Chain Bible's 88 real chains from hit records show each of these moves — de-esser, EQ, compression, reverb, delay — set to a purpose by the engineer who mixed the record, with mic, preamp, plugins and settings, every chain also in a free-plugin version. Hearing the move is step one; copying how a pro sized it is step two.
See the 88 vocal chains →Try one free chain firstWhat each round trains
| Round | Processing used | What it stands for in a real mix |
|---|---|---|
| 1 · De-essed | Static −8 dB bell at 7.5 kHz, Q 2 | A de-esser cutting the S band. Real de-essers only cut while the S is loud, so the effect is smaller and vowels stay bright. |
| 2 · +4 dB at 3 kHz | +4 dB bell at 3 kHz, Q 1.2 | The presence boost that makes a vocal "cut through" — and the same band that turns harsh when the beat already lives there. |
| 3 · High-passed | High-pass at 150 Hz, Q 0.7 | The first EQ move on almost every vocal chain: removing rumble, chest and mic proximity so the bass and kick own the low end. |
| 4 · Compressed | −30 dB threshold, 6:1, 3 ms / 150 ms, knee 6, RMS-matched | A hard vocal compressor with makeup gain. Because the level is matched, you have to hear density and breath lift, not loudness. |
| 5 · Reverb | 1.2 s synthetic decay at −12 dB | A short-to-medium vocal reverb on a send. The tail on word endings is the tell. |
| 6 · Slap delay | One 95 ms repeat at −9 dB | Classic slapback: too fast to hear as an echo, it reads as thickness on consonants. |
| 7 · 6 dB louder | Gain ×2, nothing else | The sanity round. Louder almost always sounds "better" — which is why every other round is level-aware. |
Numbers are the exact settings rendered by this page. They are deliberately bigger than most mix moves so the difference is learnable; once you catch them, try half.
More free vocal tools
Ear training for mixing vocals: why blind A/B beats theory
Most people learn vocal mixing backwards. They read what a de-esser does, watch someone set a compressor to 4:1, memorise "high-pass at 100 Hz", and then sit in front of their own take unable to tell whether the move they just made helped. Ear training for mixing vocals fixes the missing half: you need to recognise the sound of each move before the settings mean anything. That is what this vocal mixing ear test does. Seven rounds, each isolating one change on the same real, dry vocal phrase, with the processed version randomly placed in A or B so you cannot cheat by habit. You either hear it or you do not, and either answer is useful information.
The clips are deliberately short and dry. A one-and-a-half-second phrase keeps your attention on the consonants and word endings where most of these moves show up, and a dry source means nothing else is masking the change. Everything is rendered inside your browser from the same four takes, so A and B differ only in the one move the round is about.
What each of the seven rounds teaches
De-essing is about consonants, not brightness. The processed clip in round one has a static cut at 7.5 kHz, standing in for a de-esser; if you listen to the vowels you will hear almost nothing, if you listen to the S and T sounds you will hear them step back behind the voice. 3 kHz is the presence band, where words become intelligible and where vocals turn harsh when the boost is too big or the beat already lives there. Four decibels is a lot; a working mix usually does one or two. The 150 Hz high-pass teaches the difference between thin and clean. The processed version loses chest and proximity, which sounds wrong in solo and right against a kick and bass.
Compression is the round most people fail, and it is the fair one, because the compressed clip is matched to the same RMS level as the dry clip. Without the loudness clue you have to hear what compression actually does: quiet syllables and breaths come up, the front of each word is slightly softened by the 3 ms attack, the phrase feels denser and more "finished". Reverb at −12 dB is quieter than many beginners run it and still clearly audible on word endings once you know to listen there. Slap delay at 95 ms is too short to register as an echo; it reads as weight and doubling on consonants, which is exactly why it is on so many rap and rock vocals. Six decibels of level is the sanity check: if that one gets by you, something is wrong with your playback, not your ears.
How to train for five minutes a day with your own takes
The test shows you which moves you can already hear. The practice that moves the needle is the same A/B idea on your own vocals, in your own DAW, with the bypass button. Pick one move per day. Loop a four-second phrase, set the plugin to roughly the amount used here, and toggle bypass while you listen to one specific thing from the list above. Once you catch it every time, halve the amount and do it again. Within a couple of weeks the moves that were invisible become obvious, and more importantly, you start noticing when a real mix has too much of one.
- Level-match everything. Use makeup gain or output trim so the processed and bypassed versions are equally loud. Louder always wins blind, and that teaches you nothing.
- Use the same two or three takes. Familiarity is the point; you are training recognition, not judgement of new material.
- Headphones for the top and bottom rounds. De-essing, high-pass and reverb tails live in ranges most small speakers cannot reproduce.
- Retake this test monthly. Retry reshuffles which clip and which letter is processed, so you cannot memorise the answers.
Reading your score honestly
Three or fewer is normal if you have never trained for this; it is where almost everyone starts, and it is exactly the gap that studying real chains closes, because each chain is a worked example of what a pro decided a de-esser, an EQ or a compressor should do on a specific voice. Four or five means you catch the big moves and miss the subtle ones; those are the rounds to loop. Six or seven means the textbook amounts are easy for you, so the next level is hearing half of them in a full mix rather than on a dry solo. None of these numbers measures taste, talent or how your mixes sound. It measures whether a specific move is audible to you today.
FAQ
Is this vocal mixing ear test free?
Yes. No sign-up, no email, no certificate. Tap Start, listen, answer seven rounds and get a score with a listening hint for each move. Retry as often as you like.
Do I need headphones?
Strongly recommended. Rounds 1 (de-esser, 7.5 kHz), 3 (150 Hz high-pass) and 5 (reverb tail) depend on frequency ranges and quiet detail that laptop and phone speakers do not reproduce well. Any closed headphones or earbuds are enough.
Why is the compressed version not louder?
Because that would make the round about loudness, not compression. The compressed clip is rendered, then scaled to the same RMS level as the dry clip, so you have to hear density, raised breaths and softened consonant fronts. That is also how you should A/B a compressor in your DAW: with makeup gain set so bypass and active are equally loud.
Is the de-esser round a real de-esser?
No, it is a static −8 dB cut at 7.5 kHz, which is an honest stand-in: it affects the same band a de-esser targets. A real de-esser only cuts while the S is loud, so in a mix the effect is smaller and vowels stay brighter. Listening to consonants is still the skill.
What does the score mean?
0–3 is normal for a beginner, 4–5 means you hear the big moves and miss the subtle ones, 6–7 means textbook amounts are easy and you should practise at half strength. It measures whether a specific move is audible to you today, not talent or how good your mixes are.
Does the page upload anything or use my microphone?
No. The four short vocal clips are fetched from the store's CDN when you tap Start, then all processing and playback happens in your browser with the Web Audio API. There is no server, no microphone access and no tracking. Nothing autoplays.
Why are the processing amounts bigger than I would use in a mix?
So the difference is learnable. A −8 dB cut, +4 dB at 3 kHz and 6:1 compression are textbook sizes that make each move clearly audible on a dry clip. Once you catch them every time, train with half the amount on your own takes; real mixes usually live there.
Can I take the test again with different clips?
Yes. Retry reshuffles which of the four recorded phrases each round uses and whether the processed version lands on A or B, and every version is re-rendered, so you cannot memorise the answers.
Know the problem. Now copy the chain that solves it.
The Vocal Chain Bible is a 187-page PDF with 88 real vocal chains from hit records, as reported by the engineers in public interviews — mic, preamp, plugins and settings, each chain also in a free-plugin version, plus the full Vocal Chain Maker app. Every move you just tested is in there, sized by the person who mixed the record; hearing it is step one, copying how they set it is step two.
88 chains · every chain in a free-plugin version · Try one free chain firstFree download
Get the free Grammy Sauce PDF
The plugins and settings 100+ Grammy-winning engineers actually use — collected from 5 years of interviews, in one free PDF.
Want the full pro setup? Get the Vocal Chain Bible — 88 real chains →