You exported your mix, dropped it next to three commercial tracks in a playlist, and within seconds you knew: yours is the quiet one. If you keep asking why does my mix sound quiet compared to other songs, the answer is almost never "your mix is bad" — it's that you've been diagnosing loudness with the wrong tool, and the gap is measurable in numbers you haven't been looking at.
This is a diagnostic problem before it's a mixing problem. The producers who close the loudness gap fastest aren't the ones with the fanciest plugins — they measure first, locate the specific deficit, and fix exactly that. Here's how to run that diagnostic using numbers instead of guesses.
The first mistake happens before any processing: you're reading the wrong meter. Your DAW's master meter shows peak level in dBFS — the highest instantaneous sample value. Peaks tell you whether you're about to clip. They tell you nothing about how loud your mix feels over time.
Loudness is perceived average energy, measured in LUFS. A mix peaking at -0.1 dBFS can sound 8 dB quieter than a mix peaking at -6 dBFS — the first has transient spikes eating all the headroom while the second packs dense average energy underneath a lower ceiling.
Do this first: Measure your mix's integrated LUFS with a LUFS checker and write it down. Then measure your reference. The difference is your loudness gap, in dB. -19 LUFS vs. -9 LUFS = a 10 dB gap — now a target, not a vague feeling.
Crest factor is the difference between your peak and your average level — the single most diagnostic number for loudness problems, and almost nobody checks it.
A modern commercial mix has a crest factor of 8–10 dB. If yours is 14–18 dB, you have too much dynamic range — transients eat all the headroom and your average level starves. If your crest factor is already 7–8 dB and you're still quiet, the problem isn't dynamics; it's density or spectral balance.
How to measure: Subtract integrated LUFS from true-peak dBTP. -0.5 dBTP minus -18 LUFS = 17.5 dB crest factor. Compare against genre targets:
5+ dB wider than target? Your fix is compression and limiting (section 4). Already in range and still quiet? Skip to section 3 — your problem is spectral.
Here's the deficit that hides from peak and LUFS meters: subsonic and low-frequency energy eats headroom without contributing perceived loudness. Human hearing is less sensitive below 100 Hz (Fletcher-Munson), so a kick with 8 dB of energy at 35 Hz takes 8 dB of your peak budget while adding almost nothing to perceived loudness. Commercial mixes ruthlessly filter below 30 Hz.
If your LUFS reads low and crest factor reads wide but you've already compressed and limited, the culprit is unmanaged low end — energy parked in frequencies the ear can't use.
Diagnose it: Pull your mix into a spectral analyzer and look below 40 Hz. A plateau of energy across the whole track is stolen headroom. Commercial references show a steep roll-off below 30 Hz and controlled energy between 30–60 Hz.
Fixes: - High-pass every track that doesn't contribute musical low end: vocals at 80–100 Hz, guitars at 90–120 Hz, hats at 150 Hz. Sweep and listen. - 24 dB/oct high-pass at 20–25 Hz on your master bus as a baseline. - Kick or bass content below 30 Hz? Shelf or low-pass it out. Confirm with a spectral analyzer.
Re-measure LUFS after cleaning the low end. You'll often gain 1–3 dB with no audible change — you removed energy you couldn't hear.
If your mix sounds quiet compared to other songs, the fix is compression and limiting applied strategically, not smashed. The goal: tighten the range between your loudest transients and your average level so you can raise the whole mix without clipping.
Bus compression (before limiting): - Glue compressor on your mix bus: 2:1 ratio, 30 ms attack, auto or 100 ms release, 2–3 dB gain reduction on the loudest passages. Raises average density without crushing transients. - Check drum and vocal buses. Drums swinging 12 dB? Compress the drum bus 3–4 dB. Vocal riding 10 dB above the bed? Compress 4–6 dB or automate clip gain first.
Limiting (the final push): - True-peak limiter on your master bus, ceiling at -1.0 dBTP to prevent inter-sample peaks on lossy codecs. - Push input gain until 2–4 dB of gain reduction on the loudest transients. More than that during mixing and you're mastering, not mixing — you'll mask problems you need to hear. - Re-measure integrated LUFS. At -10 to -8 LUFS, you've closed the gap. Still 3+ dB below your reference after 4 dB of limiting? The problem is upstream — loop back to sections 3 and 4.
The last diagnostic trap is comparison methodology. A/B-ing your unmastered mix against a released, mastered, platform-normalized commercial track at equal fader position? You're comparing a rough cut to a finished product. It isn't fair.
Commercial tracks went through mastering that added 3–6 dB of loudness through multiband compression, limiting, and sometimes soft clipping. Streaming platforms then normalize them down to around -14 LUFS (Spotify) or -16 LUFS (YouTube). In your DAW, before normalization, the commercial track will always be louder — that's the point of mastering.
Compare correctly: - Level-match your reference down by 6–8 dB so its perceived loudness matches your mix. Now compare tonal balance, width, and density — not loudness. At equal loudness, your brain stops biasing toward the louder track. - To hear your mix at competitive loudness, use a LUFS checker to measure your current level, then push a limiter to -10 LUFS temporarily. Don't export that version — use it to hear what the mastering engineer will receive. - On Spotify, a -8 LUFS master and a -16 LUFS mix both play back at -14 LUFS. Your quiet mix won't sound quiet on Spotify — it'll sound quiet in your DAW and your car, where there's no normalization. That's the gap you're closing.
If your mix sounds quiet compared to other songs, the fix is never "turn it up." Loudness is a function of average energy, crest factor, spectral balance, and limiting — all measurable. The producers who fix loudness fastest measure first: LUFS for the gap, crest factor for the cause, spectral analysis for hidden deficits. Once you have the numbers, the fix is specific instead of guesswork.
MixDiagnose runs this diagnostic automatically. Upload your mix and get integrated LUFS, true-peak, crest factor, and a spectral breakdown — compared against genre-specific commercial targets. You see the gap in numbers, the cause, and the fix to apply.
Measure the gap. Fix the cause. Move on.