LUFS Loudness Meter: Free LUFS Checker for Audio Files
Overview
LUFS Loudness Meter. You feed the LUFS Loudness Meter an audio file — upload it and click Analyze Loudness — and it hands back your track's integrated, short-term and momentary loudness in LUFS, calculated with real K-weighting and gating per the ITU-R BS.1770 spec. Your reading gets checked against Spotify, YouTube, podcast and broadcast loudness targets, so you'll know before you upload whether your mix is about to get turned down. The online tone generator runs on the Web Audio API — nothing is recorded or uploaded.
Your mix sounds loud in the studio, but will Spotify turn it down? Drop a track into this LUFS loudness meter and you get its perceived loudness in seconds, measured the same way streaming services measure it. It works as an online LUFS meter that runs in your browser, so your audio never leaves your computer.
How the LUFS Loudness Meter Reads Your Audio File
You can drag and drop a WAV, MP3, FLAC or AAC file onto the meter. The analysis follows ITU-R BS.1770, the standard behind every major streaming service, so this free LUFS checker agrees with any correctly built meter to within a tenth of a unit. Engineers who want readings while they mix use a plugin instead: a free loudness meter in VST, AU or AAX format sits on the master bus of a DAW and measures playback in real time. The meter only reads the signal. It never changes your audio, so any fix happens in the mix or master. The subwoofer crossover calculator tool breaks the math down so the answer is never a black box.
What Loudness Units Full Scale Actually Measure
LUFS stands for loudness units relative to full scale. US television standards call the same unit LKFS, and one LU is one decibel of level change. RMS measures raw signal level and treats every frequency the same, but human hearing does not: we hear 3kHz far more easily than low frequencies around 60 Hz. K-weighting filters the signal to match that, and channel weighting adds surround channels at a slightly higher gain. Readings are negative because they sit below 0 dBFS, the digital ceiling, so -18 LUFS is quieter than -12 LUFS. Use the free room noise floor measurement to get an instant reading without any extra hardware.
Momentary Loudness and Short-Term Loudness
Momentary loudness uses a 400 ms window, fast enough to catch a snare hit or a shouted word. Short-term loudness averages a 3-second window, which shows how loud a whole phrase feels, so you can compare the verse with the chorus and spot a drop that jumps out. Neither is the figure platforms normalize to, but both show you where the loudest moments sit.
Integrated Loudness and Gating
The integrated loudness figure averages the whole file from first sample to last. Two gates stop silence from dragging it down: absolute gating ignores every 400 ms block below -70 LUFS, and a relative gate then drops blocks more than 10 units under the running average, so long fades and quiet sections don't count. This is the one number Spotify, Apple Music and YouTube compare against their target.
Loudness Range and Dynamic Range
Loudness range (LRA) describes dynamic variation: it compares the softer and louder short-term readings across the track, ignoring the quietest 10% and the loudest 5%. Limited pop often reads 4 to 6 LU, while orchestral recordings with real loud sections can pass 15. Plugin meters set that LRA reading beside PSR (peak to short-term ratio), PLR (peak to loudness ratio) and crest factor, often on a histogram or a scrolling loudness graph.
True Peak and Sample Peak
An ordinary peak meter only reads the stored samples. A true peak meter oversamples the signal to find each inter-sample peak, a point where the rebuilt waveform rises between two samples after digital-to-analog conversion or encoding to lossy codecs. A file whose peak level reads -0.1 can still overshoot once converted, and that clipping is audible distortion. Most streaming platforms ask for a true peak ceiling of -1 dBTP, and a dense master deserves a little extra margin so the encoder has room.
LUFS Meter Targets for Streaming Platforms and Broadcast
Each service publishes its own loudness target, and the meter's integrated figure is what you hold against it. Music services cluster near -14 LUFS, while broadcast rules sit far lower because TV and film mixes need headroom for dialogue and effects.
| Platform or standard | Integrated target |
|---|---|
| Spotify, YouTube, Tidal, Amazon Music | -14 LUFS |
| Deezer | -15 LUFS |
| Apple Music, Apple podcast delivery | -16 LUFS |
| SoundCloud | No normalization |
| EBU R128 (European television) | -23 LUFS |
| ATSC A/85 (US television) | -24 LKFS |
| Netflix | -27, measured on dialogue |
Genre Loudness Levels for Mastered Music
Normalization doesn't erase stylistic habits. Commercial releases in dense styles such as trap and metal are usually mastered well above the streaming target, while jazz and classical keep more headroom and range. Rather than chasing one number, load a few reference tracks from your own genre into the meter and compare their integrated readings with yours.
Delivering a Podcast Episode at -16 with a LUFS Checker
I'm finishing episode 47 of a two-host interview show, 52 minutes 18 seconds long, and the question is simple: can I turn it up to Apple's recommended -16 LUFS (with a tolerance of about 1 LU either side) without breaking the -1 dBTP ceiling? The answer is no, not with gain alone: the episode needs about 1.7 dB of limiting.
I export a 24-bit WAV of the final mix and drop it into the meter. The reading comes back at -19.4 LUFS integrated, a true peak of -2.7 and a loudness range of 7.8 LU. That integrated loudness figure is the one to work from, so the gain I need is \(-16 - (-19.4) = +3.4\) dB. Adding 3.4 dB lifts the true peak to +0.7, which is 1.7 dB over the ceiling. Before reaching for a limiter, I solve for the largest clean gain: \(-1 - (-2.7) = +1.7\) dB, which only gets the episode to -17.7, outside the tolerance.
| Gain applied | Integrated | True peak | Limiting needed |
|---|---|---|---|
| +1.7 dB | -17.7 LUFS | -1.0 | None, but too quiet |
| +2.4 dB | -17.0 LUFS | -0.3 | 0.7 dB |
| +3.4 dB | -16.0 LUFS | +0.7 | 1.7 dB |
So I apply +3.4 dB into a limiter with its ceiling at -1 dBTP and measure again, because limiting strips a little energy and the integrated reading usually slips slightly below -16. If it lands lower than -16.5, I add a few tenths of gain and rerun the meter rather than trusting the arithmetic.
The number I'd watch here is the 7.8 LU range, not the integrated value. One guest speaks noticeably softer, and integrated loudness averages both voices, so the episode can hit -16 while that guest still sits several units under the host. Gating hides none of this, because quiet speech sits well above the -70 LUFS absolute gate. The fix is leveling the softer voice before the limiter, then checking the short-term reading during their answers. One caveat: if the same file goes to an audiobook platform that specifies RMS and peak limits instead of LUFS, this reading alone doesn't prove compliance; measure RMS separately.
Loudness Normalization and Your Audio Loudness
Loudness normalization is the playback half: a service measures each upload with the same integrated figure this meter shows, then applies one fixed gain change so every song plays at a similar playback volume. Anything louder than the target gets turned down. Audio normalization adds no compression or limiting of its own; it applies a single volume reduction across the whole track, so a louder master buys nothing except thinner dynamics.
Worked Example: Calculating the Loudness Penalty
The loudness penalty is the gain a platform removes, found from the meter's integrated reading:
$$\text{Penalty} = \text{Target} - \text{Integrated} \quad (\text{only when Integrated} > \text{Target})$$
Say your master reads -9.2 LUFS integrated, with a true peak of -0.2 and an LRA of 5.3. Against Spotify's -14 LUFS, the penalty is \(-14 - (-9.2) = -4.8\) dB, so the track plays that much quieter and its true peak lands near -5.0. The same file loses more wherever the target is lower:
| Platform target | Penalty (dB) | True peak after turn-down |
|---|---|---|
| -14 LUFS | -4.8 | -5.0 |
| -15 LUFS | -5.8 | -6.0 |
| -16 LUFS | -6.8 | -7.0 |
| -23 LUFS | -13.8 | -14.0 |
A loudness penalty is not a target to hit. A -4.8 result only means the platform does the turning down for you; if the quieter playback sounds weaker than your references, the cause is lost transients, not level.
Using an Online LUFS Analyzer for Mixing and Mastering
During mixing, keep loudness metering on the master bus so you know where you stand before any limiter goes on. While mastering, compare against reference material level-matched to the same integrated value, so you judge tone and punch rather than level. A free LUFS meter in a browser tab suits a final check, and a LUFS meter online needs no install, unlike a plugin.
- Push level with limiting in small steps, and set the limiter to catch inter-sample peaks.
- If the integrated reading is still under target after the limiter, add saturation and re-measure instead of more gain reduction.
- If the short-term reading barely moves between sections, the master is probably overcompressed.
- Measure stereo, surround and 5.1 versions separately when the deliverable has more than two channels.
Checking an Album or EP for Even Loudness
For a full record or EP, measure every track and compare each integrated figure with the record's average. A song a couple of units louder or quieter than its neighbours jumps out when the record plays through, even with normalization on, because album mode on most services keeps the level gaps you mastered.