Master a Song for TikTok, Reels & Shorts: Loudness Guide

Master a song for TikTok, Reels and Shorts to roughly -14 LUFS integrated with a -1.0 dBTP true-peak ceiling. Short-form feeds apply loudness normalization, so a hotter master (say -8 LUFS) gets turned down on playback and can end up sounding quieter and less punchy than a clean -14 LUFS file. One well-made master survives all three.

What LUFS should a song be for TikTok?

Aim for about -14 LUFS integrated with a true-peak ceiling of -1.0 dBTP. That is the same target used across streaming and it is the safest single number for short-form video, where your track competes with dialogue, other creators' clips and platform sound effects.

Be honest about the uncertainty here: unlike Spotify, which publishes a -14 LUFS normalization target with a -1 dBTP headroom recommendation in its own artist documentation (Spotify Loudness Normalization, as of July 2026), TikTok, Instagram and YouTube Shorts do not publish an exact playback loudness number for music. What creators and engineers measure in practice clusters around the same -14 LUFS convention, which is why it is a sensible target rather than a guarantee. Treat it as a well-supported estimate, not a spec sheet. For the full breakdown of streaming targets, see our LUFS streaming targets guide.

Do Reels and Shorts normalize loudness like Spotify?

Short-form platforms do apply loudness normalization on playback, though how aggressively varies and the exact behavior is not all officially documented. YouTube is the clearest case: it normalizes toward roughly -14 LUFS and only turns loud uploads down, and it exposes a measured "content loudness" figure to creators. Instagram Reels and TikTok behave similarly enough in listening tests that mastering hot buys you little.

There are real differences worth knowing. Normalization is usually applied per-clip, not album-referenced, so there is no cross-track consistency to lean on the way there is on Spotify. And some in-feed contexts historically did less normalization than others, which is exactly why chasing a moving target with a hot master is risky: you optimize for one surface and get penalized on another. Mastering to a clean, moderate level is the one approach that holds up whether a given feed normalizes hard, lightly, or not at all.

Does mastering louder help or hurt on short-form?

On a normalized feed, pushing your master to -8 or -7 LUFS does not make you louder. The platform simply attenuates it back down toward its playback target, and you are left with a turned-down, more squashed version of your song. The loudness-war reflex is genuinely backwards here.

There are two ways a hot master actively hurts you:

• Lost dynamics. To reach -8 LUFS you crush transients with heavy limiting. When the platform turns the whole thing down, those transients do not come back. A -14 LUFS master that kept its punch will often sound louder and more alive than a -8 master normalized down to the same level.
• Introduced distortion. If your true peaks exceed the ceiling, downstream limiting and codec conversion can add harshness that was not in your original. That is why headroom matters as much as the loudness number.

The honest counter-message: loud is not a strategy on normalized platforms. Clean and controlled is. If your goal is competitive loudness without the damage, read how to make a song louder without clipping.

Can one master cover TikTok, Reels and Shorts?

Yes. Because all three converge on roughly the same normalization convention, a single master at about -14 LUFS integrated, -1.0 dBTP true peak is the pragmatic universal file. You do not need a separate export per platform for music, and you generally do not need a louder "TikTok version."

Verified July 2026 — prices/specs change; re-check the source.

If you also post longer background-music videos, the same logic applies; our YouTube background-music loudness guide covers the longer-form nuances.

How to survive platform transcoding (true-peak headroom)

Every upload gets re-encoded. TikTok, Reels and Shorts convert your audio to a lossy codec (AAC-family), and lossy encoding can push reconstructed peaks above your original sample peaks. If you mastered to 0.0 dBFS, transcoding can shove you into clipping that was not there in your export.

That is the entire reason for the -1.0 dBTP ceiling: it is headroom against the codec, not a loudness setting. Measured on a true-peak meter (ITU-R BS.1770-4), a -1.0 dBTP master gives the encoder room to breathe without producing crispy, distorted transients after upload. For the full explanation of why sample peaks and true peaks differ, read what true peak (dBTP) means.

Practical checklist for a transcode-safe short-form master:

1. Integrated loudness around -14 LUFS (measured, not guessed).
2. True peak at or below -1.0 dBTP — use a true-peak limiter, not just a sample-peak one.
3. Keep meaningful transients intact; do not limit into a brick wall to chase loudness you will lose anyway.
4. Export at the highest quality you can (WAV for upload where the app allows it) so the platform's encoder is the only lossy stage.

Check your loudness free before you post

Guessing is the mistake. Before you commit a song to a Reel or a Short, you want real numbers: integrated LUFS, true peak in dBTP, and loudness range. TrackGleam masters your track in the browser and then measures the finished file (ITU-R BS.1770-4, gated) so you can confirm you actually landed near -14 LUFS with a -1.0 dBTP ceiling — not just hope you did.

It runs 100% in your browser via WebAssembly: nothing uploads, no account, no login, and the free master is unlimited. You can preview the AI-tuned GleamAI result in full before deciding whether to pay for it, and the standard free master already hits the short-form-safe target described above. If you make AI music, the same workflow handles Suno and Udio tracks. Measure once, export one file, post it everywhere.

Master a track free — no signup, nothing uploads