CutScore
Join the waitlist
AUDIO LOUDNESS BLOG / 8 MIN READ

What LUFS should a TikTok video be?

Short answer: around −14 LUFS integrated, with a true peak under −1 dBTP. The longer answer explains why TikTok turns your audio up or down on its own, and how to mix so it lands right.

Check your loudness automatically Jump to the target numbers
−14 LUFSintegrated target
−1 dBTPtrue peak ceiling
autoTikTok normalises playback
9:16vertical, sound on by feel

By Thomas, founder of CutScore · Updated June 2026

LOUDNESS CHECK · tiktok_cut.mp4
Two people recording a podcast with microphones in a soft-lit room, the kind of talking-to-camera content where TikTok loudness and the voice-to-music balance decide whether the audio feels right.
CRAFT SCORE
FIXES ADVISED
loudness pass before publish
Integrated loudness low · −19.4 LUFSwhole clip
True peak in range · −1.6 dBTP✓
Music over voice · pull music −4 dB00:11
The 30-second answer A TikTok video should sit at an integrated loudness of about −14 LUFS, with a true peak at or below −1 dBTP. TikTok normalises loudness during playback, so it quietly turns very loud videos down and leaves very quiet ones feeling weak in the feed. Mixing to −14 LUFS gives you a friendly home base: loud enough to feel present, with enough headroom that nothing crackles after TikTok re-encodes your file. The one rule that beats every number: keep the voice clearly on top of the music. If measuring this by hand sounds tedious, that is exactly the loudness check CutScore runs in one pass.
WHY THIS QUESTION IS CONFUSING

LUFS sounds like one of those acronyms invented to make audio feel harder than it is. It stands for Loudness Units relative to Full Scale, and all it really measures is how loud something feels to a human over time, not how tall the loudest spike is. That distinction is the whole game. Two videos can have identical peaks and feel a world apart in volume, and TikTok cares about the feel, not the spike.

Here is where people get stuck. The old instinct is to make your file as loud as physically possible, the so-called loudness war. TikTok, like Spotify and YouTube before it, ended that war by normalising playback. It listens to how loud your video is overall and nudges it toward a target so the feed does not lurch between a whisper and a shout. Push your mix way past the target and the platform just turns it back down, often after the limiting has already squashed the life out of it.

So the real question is not "how loud can I get." It is "what level does TikTok actually want." I have shipped clips that were technically loud and still felt thin in the feed, because I was chasing peaks instead of perceived loudness. Get the LUFS right and the problem disappears.

THE TARGET NUMBERS

What LUFS should a TikTok video be, in numbers.

Three values cover almost every TikTok you will ever post. Hit these and your audio will sit comfortably in the feed, on phone speakers and in earbuds alike.

What to measureTarget to hitWhat goes wrong if you miss it
Integrated loudness≈ −14 LUFSFar below and it feels timid next to the next clip; far above and TikTok turns it down anyway.
True peak≤ −1 dBTPPeaks above the ceiling crackle and distort once TikTok re-encodes your audio.
Voice vs musicvoice on topMusic drowning the speech is the loudest amateur tell there is, no pun intended.
Is −14 LUFS a hard rule?No. It is a safe centre, not a law. A music-led clip with no talking can sit a little hotter, and a quiet, intimate piece to camera can sit a little softer. The −14 LUFS figure matches what most platforms normalise toward, so it travels well when you cross-post the same cut to Reels or Shorts.
SKIP THE METER MATH

Reading integrated LUFS and true peak on every export gets old fast. CutScore measures both on your file and tells you the exact gain change to make, so you can mix instead of squint at meters.

Join the waitlist
HOW TIKTOK ACTUALLY TREATS YOUR AUDIO

Why mixing louder than the target backfires.

TikTok normalises loudness, so the feed stays even

When someone scrolls, TikTok wants each video to land at roughly the same volume so nobody scrambles for the side button. To do that it measures your video's loudness and nudges it toward a target. If your clip is far above that target, it gets turned down. If it is far below, it more or less stays where you left it, which means a quiet mix arrives quiet. This is the same logic LUFS on YouTube works under, and it is why exporting at −9 LUFS to "be loud" does nothing but cost you headroom.

An audio engineer adjusting faders and watching a level meter on a console, the kind of careful loudness pass that gets a TikTok video to its -14 LUFS target without the audio distorting.
Loudness is a perceived value over time, not the height of the loudest spike. Photo: Los Muertos Crew / Pexels.

True peak is the limit that protects you on re-encode

Loudness is the average feel; true peak is the absolute ceiling. TikTok re-encodes every file you upload, and that compression can push samples slightly higher than they were in your export. Leave yourself a margin of about a decibel by keeping the true peak at or below −1 dBTP, and the re-encode has room to breathe. Skip that margin and the loudest consonants and drum hits arrive cracked. A brick-wall limiter set to −1 dBTP on your master bus is the cheapest insurance in editing.

The voice-to-music balance beats every number

You can nail −14 LUFS to the decimal and still ship audio that feels off, because LUFS is one number for the whole mix. If your trending track is sitting at the same level as your voice, the platform reads it as "correctly loud" while a human reads it as "I cannot understand a word." Pull the music down four or five decibels under the voice and check it on a phone speaker, the worst speaker most of your audience owns. If the words survive there, you are fine. If the music is winning, it is too loud, no meter required.

RATHER SEE IT THAN READ IT?

Here is a real CutScore report on an everyday vertical clip: the loudness, the true peak, the voice-to-music balance, scored, with timestamps and the exact fixes.

See a sample report
HIT THE TARGET

Three moves to get to −14 LUFS.

You do not need a mastering studio. You need a meter, one limiter, and a minute of attention before you export. Do these in order.

1
2-MIN FIXAUDIO
Normalise the whole mix toward −14 LUFS integrated
Put a loudness meter on your master output and play the entire video so it reads integrated, not momentary, loudness. If it lands at −19 LUFS, raise the master gain by about five decibels and re-read. Most editors have a built-in EBU R128 meter, and free standalone ones exist if yours does not.
How Read the integrated value over the full clip, or let CutScore measure it and name the exact gain change.
2
AUDIOPEAKS
Cap the true peak at −1 dBTP
Once you are near the loudness target, put a brick-wall limiter on the master bus and set its ceiling to −1 dBTP. This stops any spike from going over the line that TikTok's re-encode would turn into distortion. It is one plugin, set once, and you can leave it on for every export forever.
How Limiter ceiling at −1 dBTP. Confirm with a true-peak meter, not a sample-peak one. See true peak.
3
EAR CHECKBALANCE
Drop the music under the voice and listen on a phone
The meter cannot tell you if the music is burying you, only your ears can. Lower the music four or five decibels below the voice, then play the export through a phone speaker, not your good headphones. If every word survives that test, the balance is right. If you reach for the lyrics, the music is still too loud.
How Phone speaker, arm's length, in a slightly noisy room. That is how most people will actually hear it.
CROSS-POSTING

Does the same LUFS work for Reels and Shorts?

Mostly, yes. The big short-form platforms normalise toward similar levels, which is why one well-mixed cut travels without a re-export.

TIKTOK

Around −14 LUFS

Mix to about −14 LUFS integrated with peaks under −1 dBTP and the feed will leave it alone. The thing TikTok punishes is not "slightly under target," it is wild swings and crushed, over-limited audio. Get the voice clear and you are most of the way there.

REELS

The same ballpark

Instagram normalises in the same neighbourhood, so a −14 LUFS TikTok cut drops onto Reels without trouble. If you want the deeper version, I wrote a whole piece on loudness for Instagram Reels.

SHORTS

YouTube's −14 LUFS too

YouTube normalises toward −14 LUFS, and Shorts ride the same system, so your TikTok master is already at home there. The wider question of how loud to go on YouTube has its own answer in how loud a YouTube video should be.

How CutScore checks your TikTok loudness CutScore is an AI video quality coach for pre-publish QC. It measures your loudness with an EBU R128 meter, reads the true peak, and listens to how the voice sits against the music, then tells you whether you are on the −14 LUFS target and exactly how much gain to add or cut. You also get the rest of the craft scored: exposure, pacing, the hook, captions, export settings. One score, the evidence behind it, and a prioritised fix list, before anyone in the feed hears it. It judges the craft of the video itself, so it sits next to a growth tool rather than competing with one. More on the method and the standards.
Get your loudness checked automatically See everything we check
KEEP READING
LUFS for YouTube True peak Loudness for Reels Check audio levels before upload A full sample report
QUESTIONS

Frequently asked.

Aim for an integrated loudness around −14 LUFS with a true peak at or below −1 dBTP. TikTok normalises playback toward a consistent level, so mixing far below the target makes your audio feel weak, and mixing far above it gets turned down and can distort. The −14 LUFS figure is a safe, platform-friendly home base for talking videos with music.
Yes. TikTok applies loudness normalisation during playback so that videos in the feed sit at a roughly even level. That means a quiet mix is not rescued by cranking it on export, and a very loud mix gets pulled back down. Mix to a sensible target instead of trying to be the loudest file in the feed.
No, −14 LUFS is a sound default and it is not quiet once normalisation does its job. People confuse loud peaks with loud perceived volume. What matters is how loud the whole thing feels, which is what LUFS measures. Hit around −14 LUFS integrated, keep peaks under −1 dBTP, and make sure the voice sits clearly on top of any music.
Run your final export through a loudness meter that reports integrated LUFS and true peak. Most editors include an EBU R128 or ITU-R BS.1770 meter, and free standalone meters exist too. Read the integrated value over the whole video, not a single loud moment. Or hand the file to CutScore and it reports the loudness, the true peak and the exact gain change to make.
EARLY ACCESS

Stop guessing at your loudness.

CutScore measures your TikTok loudness, true peak and voice-to-music balance, then tells you exactly what to change, with the evidence to back it up. Join the waitlist for early access.

Join the waitlist