dB Level Trainer
This dB level trainer helps you recognise loudness differences in decibels. The trainer plays two short bursts — pink noise or a steady tone — that differ by an exact dB amount, in a random order. Your job is to name the gap. Start easy with larger gaps up to 12 dB, narrow the range as you improve, then chase the 1 dB just-noticeable edge on Hard. After each round you see the real-world reference for that step.
ℹ This is an uncalibrated practice aid, not a meter. The dB gap is the exact digital gain ratio applied to the second burst (ratio = 10dB/20 on amplitude) — that part is precise. But how loud that gap sounds depends on your absolute listening level, the spectrum, your headphones or speakers, and your ears. Your score is a personal practice metric, not a hearing test, certification, or measurement. Use a moderate, comfortable volume, and prefer headphones in a quiet room.
Press “New round” to begin. You will hear Burst A, then Burst B. Then pick how many dB louder the louder one was.
By how many dB did the louder burst differ from the quieter one?
Quick dB reference
- ~1 dB — about the smallest level change most people can just notice on steady sound (the “just-noticeable difference”).
- ~3 dB — double (or half) the power; a clear but modest step.
- ~6 dB — double (or half) the amplitude / pressure; an obvious jump.
- ~10 dB — roughly “twice as loud” in everyday perception (a common rule of thumb, level- and spectrum-dependent).
How It Works
The trainer synthesises one sound source — either pink noise or a steady 1 kHz tone — and plays it as two short bursts about a third of a second each, with a brief gap between them. One burst is at a reference level; the other is scaled by an exact decibel amount. Decibels of amplitude are converted to a linear gain multiplier with the standard formula ratio = 10(dB / 20). So a +6 dB burst is played at almost exactly twice the amplitude of the reference, +3 dB is about 1.41×, and +1 dB is about 1.12×. Each burst gets a short click-free fade in and out so you judge the level, not a pop.
Which burst is louder is randomised every round, so you cannot just learn “the second one is always louder.” This decibel ear training approach — a blind loudness difference test with instant feedback — mirrors the research-backed method used in professional listening curricula. You then choose the gap from a set of buttons sized to your difficulty: Easy offers a coarse spread of larger gaps (up to 12 dB), Medium a slightly tighter range, and Hard the toughest near-threshold choices of 1–4 dB. The Mixed mode draws across the range. After you answer, the trainer reveals the true gap, whether the louder burst was A or B, and the real-world meaning of that step. Your running score, accuracy, current streak and best accuracy update each round.
Two honesty points worth repeating. First, the dB difference itself is exact in the digital domain — it is a precise gain ratio applied to the sample values. Second, what you actually hear is not exact: perceived loudness depends on the absolute playback level, the frequency content (pink noise and a 1 kHz tone feel different), your headphones or speakers and their volume curve, room noise, and your own hearing. Because nothing here is calibrated to a real sound-pressure level, treat your results as ear-training practice, never as a measurement or a hearing assessment.
This kind of level discrimination training is the core of audio ear training for audio engineering and music production. Sound engineers and mixing professionals who can reliably detect small gain changes — even the subtle 1–2 dB tweaks involved in gain matching practice or mix automation — make faster, more confident decisions at the console. Level matching between tracks and gain staging both depend on the same ability to hear a precise dB difference, and that skill improves with deliberate practice. If you want to understand how these dB ratios translate into linear numbers, the dB to amplitude converter shows the exact gain multiplier for any decibel value.
Frequently Asked Questions
Is the dB difference accurate?
The gain ratio is exact: the louder burst is scaled by 10^(dB/20) in amplitude, computed in the browser, so a +6 dB burst really is about twice the amplitude of the quieter one. What is not exact is how loud that gap sounds to you — that depends on your volume setting, the sound’s spectrum, your headphones or speakers, and your ears. This trains the ear; it does not measure anything.
What do 1, 3, 6 and 10 dB actually mean?
About 1 dB is roughly the smallest level change most people can just notice on steady sound (the just-noticeable difference). 3 dB is double or half the power. 6 dB is double or half the amplitude (sound pressure). 10 dB is the common rule of thumb for “about twice as loud” in everyday perception — a generalisation that itself varies with level and frequency content.
Why does the same dB gap feel easier on tone than on pink noise (or vice versa)?
Loudness perception is not flat across the spectrum. A steady 1 kHz tone and broadband pink noise stimulate your hearing differently, and your gear may reproduce them at different effective levels. Many people find a level change easier to hear on one than the other. That is expected, and it is part of why the score is a personal practice metric rather than a calibrated result — try both and compare.
Is my score saved, and is anything uploaded?
Your session score and best accuracy are stored only in your own browser using localStorage, so the Audio Skills Progress Tracker can summarise your practice. Nothing is recorded, and no audio or score data is ever uploaded — the sound is generated locally and played straight to your output. Using private/incognito mode or clearing your browser data erases the saved results.
Can this damage my hearing or replace a hearing test?
No, it cannot replace a hearing test — it is an uncalibrated game, not a diagnostic tool. Keep the volume moderate and comfortable; the trainer caps its own output and fades each burst, but you control your device volume. If you have hearing concerns, see an audiologist. Start at a low level, especially with headphones, and never crank it to hear the 1 dB step.
How many rounds should I practise per session for decibel ear training?
Perceptual training research broadly supports short, focused sessions over a few weeks rather than marathon single sessions. Around 20–40 rounds per sitting — enough to build a meaningful accuracy score without listening fatigue clouding your judgement — tends to work well. Start on Easy until you hit roughly 80 % accuracy, then step down to Medium. Hard mode (1–4 dB gaps) is genuinely near the psychoacoustic just-noticeable-difference threshold and may take weeks of consistent practice to improve reliably.
What is the “just-noticeable difference” (JND) for loudness?
The just-noticeable difference (JND) for loudness is the smallest level change that a listener can detect on a given sound about 75 % of the time under ideal conditions. For mid-level broadband sounds it is roughly 0.5–1 dB, but it degrades at very low and very high levels and depends on the sound’s duration and spectral content. That is why the Hard mode uses 1 dB as its tightest step — anything closer is at or below the psychoacoustic threshold for most people and practically impossible to resolve with an uncalibrated playback system.
Can I use this for gain-staging or mix-level practice?
Yes, and it is one of the most practical applications. Gain staging in a DAW involves setting each channel so the mix bus has healthy headroom, which requires making confident decisions about whether a fader is 2 dB or 4 dB louder than a reference. Training your ear here — especially on the 1–3 dB choices — directly sharpens that skill. For the maths side of gain adjustments, the signal-to-noise ratio calculator and the dB-to-amplitude converter complement this trainer well.
Why use pink noise instead of white noise or music for level training?
Pink noise is spectrally shaped so it has equal energy per octave, closely matching the long-term average spectrum of music and speech. That makes it a widely used reference in acoustics and audio engineering — it exercises your full hearing range more evenly than white noise (which has too much high-frequency energy) and more consistently than music (which fluctuates in content). Practising level discrimination on pink noise is therefore a better proxy for real-world mixing and mastering judgements than training on white noise would be.