True Peak Meter Online: How to Check a Song for Free
Check the true peak of a song in your browser: what dBTP means, why a file can clip after encoding, what ceiling to aim for, and how to read the result and fix it.
To check the true peak of a song online, drop the file into the free LUFS checker. It measures the whole track and shows the true peak in dBTP next to the loudness. If the number is above −1 dBTP, the checker marks it, because that is the ceiling streaming services recommend. Nothing needs to be installed and no account is needed.
This guide explains what true peak is, why the number on your editing software can be wrong about clipping, what ceiling to use, and what to do when the reading is too high. The true peak explained page covers the theory. This article is the practical side: how to check, how to read it, how to fix it.
True peak in plain words
A digital file stores a sound as a list of numbers, called samples. A sample peak meter reads that list and reports the largest number. That is the sample peak, shown in dBFS.
But a playback device does not play the list as it stands. It rebuilds a smooth curve through the samples, and between two samples the curve can rise higher than either of them. The highest point of that rebuilt curve is the true peak, shown in dBTP. It is the level the sound actually reaches, and it can be above the sample peak. The extra bumps are called inter-sample peaks.
To find them, a true peak meter oversamples the signal, which means it computes extra points between the stored samples, and reads the highest one. The method is defined in ITU-R BS.1770, which calls for oversampling by at least four times. A meter that skips this step can under-read on dense material.
A test you can repeat
We made a test tone to show the effect. It is a sine wave at a quarter of the sample rate, with a phase offset chosen so that the stored samples never reach the top of the wave. Every sample in the file sits at −3.01 dBFS, so a sample peak meter says the file is comfortably below the ceiling. The actual wave between the samples reaches 0 dB.
We ran the file through the LUFS checker and read the true peak, then lowered the file by 1 dB and 3 dB and measured again:
| File | Sample peak | True peak read by the checker |
|---|---|---|
| Test tone, original | −3.01 dBFS | +0.12 dBTP |
| Lowered by 1 dB | −4.01 dBFS | −0.88 dBTP |
| Lowered by 3 dB | −6.01 dBFS | −2.88 dBTP |
Two things show up. The true peak is about 3 dB higher than the sample peak, which is the whole point of the meter. And the true peak moves one-to-one with the gain, so lowering the level by 3 dB lowers the reading by 3 dB. The tone is an extreme case built to show the effect, and real music will not usually show a gap this large. The reading of +0.12 for a wave that should reach 0 also tells you how precise the meter is: it is within about a tenth of a decibel here.
Why it matters on streaming
Most listeners never hear your WAV. Streaming services send a lossy version, such as AAC, Ogg Vorbis or Opus, and encoding changes the waveform slightly. A peak that was just under 0 in your file can land above 0 after the file is decoded, and then it clips. Clipping at that point is a short crackle on a transient, more noticeable on bright, dense mixes.
That is why the published guidance leaves headroom:
| Who | Ceiling | Source |
|---|---|---|
| Spotify | At or below −1 dBTP | Spotify for Artists, "Loudness normalization" (checked 4 November 2026) |
| Spotify, if the master is louder than −14 LUFS | Below −2 dBTP | The same Spotify page |
| Apple Digital Masters | No higher than −1 dBTP at the encoder input | Apple, as listed on our true peak page (checked 29 September 2026) |
| EBU R 128 (broadcast) | −1 dBTP | EBU R 128, as listed on the same page |
Spotify states −1 dBTP as best for lossy formats such as Ogg Vorbis and AAC. These guidelines can change, so read the platform's own page before a release.
Step by step: check your track
- Export the file you plan to release. Check the final WAV or MP3, not the project. A lossy file made from a WAV can have a higher true peak than the WAV itself, so measure the one you will upload.
- Open the LUFS checker. Drop the file in the box. WAV, MP3, FLAC, AAC and AIFF files up to 50 MB are accepted. The file is measured and not kept. If you have no file at hand, the page has a 20-second sample you can try.
- Read the true peak. It appears with the integrated loudness, the loudness range and the highest short-term loudness. The value is shown in dBTP, and it is marked when it is above −1.
- Read the notes below the numbers. If the true peak is above −1 dBTP, the checker says so in plain words and explains that lossy encoding can make the file clip. It also shows whether the loudness is ready for Spotify, Apple Music, YouTube and podcasts.
- Fix and measure again. Change the master, export a new file and drop it in. Compare the new reading with the old one.
The same measurement sits inside the Echoteka app as the LUFS analyzer tool. Echoteka uses one meter for both the checker and the before-and-after report of a master, so the number you see in the checker is the number the report would give for the same file.
How to read the result
| What you see | What it usually means | What to do |
|---|---|---|
| True peak below −1 dBTP | The file has headroom | Nothing for peaks. Check loudness next |
| True peak between −1 and 0 dBTP | Close to the ceiling | Lower the limiter ceiling a little and measure again |
| True peak above 0 dBTP | The wave goes over full scale between samples | Lower the ceiling or the level before you export |
| True peak well above the sample peak | The track is dense or heavily limited | Use a true peak limiter and push less |
A high reading is not a fault in the song. It is a sign that the final stage of mastering let some peaks through.
How to bring the true peak down
- Use a true peak limiter as the last stage. Set its ceiling to −1 dBTP. It oversamples inside, so it catches the overshoots a sample peak limiter misses.
- Push less. The louder the master, the more it overshoots. Streaming services turn loud masters down anyway, so the last decibel of loudness usually buys nothing. The loudness penalty checker shows how much a platform will turn your track down.
- Lower the gain. Taking the whole file down by the number of decibels you are over fixes the peak, at the cost of a quieter file. In the test above, 3 dB of gain moved the reading by 3 dB.
- Check the file you release. Run the final export through the checker, not an earlier version.
- Mind the lossy export. If your distributor encodes your master, leave extra room. Spotify asks for −2 dBTP when the master is louder than −14 LUFS.
If a track has already been clipped, lowering the level does not undo it. The article on audio clipping explains what can and cannot be repaired.
True peak and loudness are two checks
A song can pass one and fail the other. The true peak says how high the highest wave goes. The loudness in LUFS says how loud the track is on average. A quiet track can still have a high peak, and a loud master can have a peak well below the ceiling. Read both. The article on how to read loudness numbers shows how the readings fit together.
AI-generated songs
Songs from generators are often exported near full scale, and a file that looks fine on a sample meter can still be over the line. The Suno loudness report recommends keeping the true peak at or below −1 dBTP before release. A check takes a moment, and it tells you whether the file needs attention. Measuring peaks does not change where a song came from. If a platform or distributor asks you to disclose AI use, disclose it.
Common mistakes
- Trusting the editor's peak meter. Many show the sample peak only. A reading of −0.1 dB can hide a true peak above 0.
- Checking the project and not the export. The final file is what gets encoded.
- Setting the ceiling at 0. Leave room. The usual ceiling is −1 dBTP, and −2 dBTP for loud masters on Spotify.
- Fixing the peak by clipping. Hard clipping removes the overshoot and adds distortion. Use a true peak limiter or lower the level instead.
- Assuming one pass is enough. After each change, export and measure again.
FAQ
What is true peak? It is the highest level the sound wave reaches, including the points between stored samples. It is measured in dBTP and can be higher than the sample peak.
What true peak should my song have? At or below −1 dBTP for most streaming releases. Spotify asks for below −2 dBTP when the master is louder than −14 LUFS.
Is true peak the same as peak in my editor? Often not. Many editors show the sample peak. A true peak meter oversamples to find the peaks between samples.
Can a file clip when its meter never reaches 0? Yes. The wave between samples can rise above both neighbours, and lossy encoding can push it over. That is why the ceiling is −1 and not 0.
Does the checker keep my file? No. The file is measured and not stored. The page accepts files up to 50 MB.
How do I lower the true peak? Use a true peak limiter with a ceiling of −1 dBTP as the last stage, push the loudness less, or lower the level of the whole file, then export and measure again.
Read the true peak of your song
The LUFS checker shows true peak next to loudness. It is free, needs no account and does not keep your file.
Check true peak free