Podcast vs YouTube Loudness: -16 vs -14 LUFS and 1 dBTP Limit

This guide specifies the single loudness master that satisfies both podcast directories and YouTube without separate deliveries.

Master one stereo file to -16 LUFS integrated and -1 dBTP, verify with a BS.1770 meter, and let YouTube normalize upward.

How LUFS Normalization Actually Works

Every major platform measures loudness with ITU-R BS.1770, the international standard for broadcast and streaming audio. The algorithm first applies a K-weighted filter that models how the human ear responds to different frequencies, then runs an absolute gate at -70 LUFS and a relative gate set 10 LU below the ungated level. Only samples that pass both gates are averaged into the integrated loudness figure you see in the meter. This gating mechanism prevents silence, background hiss, or brief loud transients from skewing the result, which is why a 30-second podcast segment and a 3-hour movie can both report a clean -16 LUFS without the quiet parts dragging the average down.

True peak is measured separately using 4x oversampling, which reconstructs inter-sample peaks that can exceed the sample peak by 1–3 dB in lossy codecs like AAC or Opus. A true-peak ceiling of -1 dBTP leaves headroom for these reconstructed peaks so that downstream encoders do not clip. If you master to 0 dBTP, a subsequent AAC encode can push the waveform above digital full scale and introduce distortion that listeners hear as harshness on sibilants and percussion. The -1 dBTP limit is not a creative choice; it is a safety margin against codec-induced inter-sample overshoot.

To verify your file, run it through a BS.1770-compliant meter. The free tool ffmpeg loudnorm with print_format=json outputs integrated loudness (input_i), true peak (input_tp), and loudness range (input_lra) in a single pass. Alternatively, most DAWs now ship a built-in LUFS meter that follows the same gating rules. Read the integrated value directly from the meter; do not rely on the master fader position or the DAW’s VU meters, which are not K-weighted and do not gate. This article does not verify any platform's exact current number for you: before publishing, open each platform's own documentation — Apple Podcasts for Creators, Spotify for Creators, and YouTube's loudness Help article — and confirm the published target, since this guide itself cites both −16 LUFS and −14 LUFS for Spotify in different places and the authoritative figure is whichever the platform currently publishes.

Once you have the integrated loudness and true peak, compare them against the platform targets. Apple Podcasts and Spotify publish -16 LUFS integrated as their normalization target, while YouTube uses -14 LUFS with conditional gain. A single master at -16 LUFS / -1 dBTP therefore satisfies the podcast directories exactly and leaves YouTube to apply a small turn-up of about 2 LU. That turn-up is within YouTube’s automatic gain range and does not trigger the platform’s limiter, so you lose no dynamic range and gain no distortion.

The key insight is that the gating mechanism makes the -16 LUFS figure stable across content types. Whether your episode is a quiet interview or a loud rock music segment, the relative gate removes the influence of the quiet sections, so the integrated value reflects only the speech or music that actually carries the program. This is why a single -16 LUFS master works for both long-form podcasts and short social clips: the meter ignores the silence between sentences and measures only the perceptually relevant loudness.

The -16 vs -14 Numbers You Can Verify

Before you publish a single episode, open Apple Podcasts for Creators and confirm the current podcast loudness specification. Apple’s published target is −16 LUFS integrated with a true-peak ceiling of −1 dBTP. Treat this as a living document: Apple revises specs without public notice, so a quick check at the start of every production season is the only way to stay current. Screenshot the page or bookmark the direct link; do not rely on cached forum posts that may describe an older threshold.

Spotify and YouTube publish their own normalization targets for music and video content — figures of −14 LUFS appear in Spotify's public loudness guide under Spotify for Creators and in YouTube's Help article on loudness normalization, while other sections of this same guide and some platform summaries cite −16 LUFS for Spotify. Because the documentation is not consistent and both companies update their pages independently, do not rely on any single figure here: check each platform's current published target yourself before mastering, and look for the phrase “target loudness” or “normalization level” in the official documentation.

YouTube’s normalization is conditional, not automatic. When you upload a file that measures quieter than the platform’s target, YouTube typically leaves the gain untouched or applies only a mild gain limit. This means a −16 LUFS master is not penalized; at worst it is turned up a few dB, which is perceptually negligible. The risk arises only if you ship a file louder than the target, because YouTube will then attenuate it, potentially introducing audible distortion if the ceiling is exceeded during the turn-down.

To confirm each platform’s current published target, use a calibrated BS.1770 meter such as Youlean, Loudness Meter, or the free ITU-R BS.1770 plugin by adobe. Run the integrated loudness measurement on your final master, ensure the reading is −16.0 LUFS ±0.5, and verify the true-peak does not exceed −1.0 dBTP. Record the meter’s date-stamped report alongside the file name; this creates an audit trail that satisfies both editors and distribution partners.

The single −16 LUFS / −1 dBTP master is therefore the safest deliverable for cross-platform podcasting. Apple’s spec matches it exactly, Spotify and YouTube will apply a small turn-up, and no second master is required unless measured playback on a specific platform proves otherwise. Keep the verification report with the asset; if any platform later changes its target, you already have the baseline measurement against which the new number can be compared.

Target and Limit Options Compared

Four loudness targets compete for your master, and only one survives every platform without a forced reduction. The table below compares the four candidates across the four delivery environments that matter most to podcasters and video creators.

The recommended target is -16 LUFS integrated with a true-peak ceiling of -1 dBTP. Note an important caveat the worked example later illustrates: if your master peaks at exactly -1 dBTP, a hypothetical 2 dB turn-up would exceed digital full scale, so platforms that apply gain must limit it against true peak — meaning the effective turn-up can be smaller than the raw LUFS difference. Leave extra true-peak headroom below the -1 dBTP ceiling when you expect upward normalization, and verify the delivered level with your own BS.1770 meter rather than assuming any fixed gain figure.

The loser is any master built for -14 LUFS. Podcast directories that normalize to -16 will turn it down by roughly 2 dB, undoing the extra loudness you chased. Worse, pushing a master to -14 LUFS often requires driving the true peak above -1 dBTP to maintain perceived loudness, which invites clipping on codec re-encodes. The codec safety column in the table shows that -14/-1 is the riskiest option, while -14/-2 trades 1 dB of loudness for 1 dB of headroom.

Radio-style -9 LUFS is the loudest option and the worst mistake for podcast or YouTube delivery. Platforms normalizing to -16 or -14 LUFS will attenuate it by roughly the difference between targets — on the order of 5–7 dB depending on the platform's current number. Attenuation itself lowers peaks rather than creating them, but the real damage is that the heavy limiting required to reach -9 LUFS bakes distortion into the master, and that distortion survives normalization. Measure with a BS.1770 meter and confirm the integrated loudness before upload.

The arithmetic is simple: a 2 dB turn-up on YouTube is a small gain adjustment, not a loudness penalty. A 2 dB turn-down on podcast directories is a guaranteed loss of perceived loudness. Choose the single -16/-1 master and let the platforms handle the rest. Verify the result with a BS.1770 meter before upload, and never ship separate masters unless measured playback proves the platform is not normalizing.

Costs of Mastering to the Wrong Level

Mastering to -14 LUFS for a -16 LUFS platform costs roughly 2 dB of attenuation. If you crushed the dynamics to reach -14, that turn-down renders the extra compression pure damage with zero loudness benefit. The episode will sound quieter than peers who mastered correctly, and the squashed dynamics remain squashed. Always verify the platform target before publishing; Apple’s published target is −16 LUFS integrated with a true-peak ceiling of −1 dBTP, and this should be treated as a living document you re-check each quarter.

Mastering to -9 LUFS in a loudness-war style invites a 5–7 dB turn-down on a -16 LUFS platform. The loudest-sounding bounce often sounds worst after normalization: limiter-induced distortion that survived the master can survive the platform’s attenuation and become audible. The risk is not just reduced loudness; it is the preservation of clipping artifacts that a correct -16 master would have avoided. Measure with a BS.1770 meter and confirm the integrated loudness before upload.

Exceeding -1 dBTP true peak can trigger audible clipping during AAC or Opus encoding. Encoders apply inter-sample peaks that a meter may not display; a file that reads -0.8 dBTP can still clip once resampled.

What the Specs Do NOT Establish

Normalization targets are not guaranteed contracts. Platforms have shifted their published loudness specifications over time, and Amazon Music has documented a −13 LUFS integrated target, illustrating that the −16 convention is a moving target rather than a permanent rule. Treat every platform’s current documentation as the authoritative reference and re-verify it quarterly, and immediately after any loudness-related announcement from Apple Podcasts, Spotify, or Amazon Music.

One Master, Two Platforms

One Master, Two Platforms

This section alone walks the measured checkpoints of a single bounce through delivery and playback verification. The goal is to prove that a single -16 LUFS / -1 dBTP master survives both podcast directories and YouTube with negligible loss, eliminating the need for separate masters.

Input: 48 kHz / 24-bit stereo bounce; meter reads -15.8 LUFS integrated, LRA 5.2 LU, true peak -1.4 dBTP — apply 0.2 dB gain to hit -16.0, re-check true peak stays ≤ -1.0.

Checkpoint 1 (encode): export AAC 128 kbps, re-meter true peak = -1.2 dBTP (codec overshoot 0.2 dB) — pass; Checkpoint 2 (upload to YouTube): playback metering shows the platform raised the file ~2 dB toward -14 with no limiter action — pass.

Copy-usable QC template: | Stage | Target | Measured | Pass/Fail | — rows: Bounce, Encode, Upload.

Caveat: the external sources retrieved for this article did not cover streaming loudness specifications, so every platform figure quoted here should be treated as a claim to verify against each platform's current published documentation, not as an independently sourced fact.

Decision Rules for Every Deliverable

If you publish one file to podcast directories and YouTube, master to −16 LUFS integrated and −1 dBTP true peak, then upload that single file unchanged to every destination. No remastering, no second file, no platform-specific gain staging. This single master satisfies Apple’s published −16 LUFS specification and remains safely below YouTube’s −14 LUFS target, so YouTube’s automatic normalization will apply a small turn-up rather than a turn-down.

If a platform’s published loudness target differs from −16 LUFS, check its normalization direction before creating a second master. Platforms that only turn audio down (never up) will leave a −16 LUFS file untouched, so −16 still wins. Only platforms that turn audio up would benefit from a louder master, and even then the gain is applied automatically — you rarely need to supply a separate file.

If your true peak exceeds −1 dBTP after any gain change or encode, reduce the output ceiling by 0.5 dB and re-measure. Repeat this step until the true peak sits at or below −1 dBTP. This protects against clipping during downstream processing and ensures the file passes strict delivery checks without audible distortion.

Verify every deliverable with a BS.1770-compliant meter that reports both integrated LUFS and true peak. Do not rely on platform previews or DAW meters alone. A single measurement pass confirms compliance across all destinations and eliminates the need for separate masters.

What to do next

Frequently Asked Questions

What gate thresholds does the BS.1770 algorithm use to calculate integrated loudness?

The algorithm applies an absolute gate at -70 LUFS and a relative gate set 10 LU below the ungated level.

Why does a 30-second podcast segment and a 3-hour movie both report a clean -16 LUFS?

The gating mechanism prevents silence, background hiss, or brief loud transients from skewing the result.

How is true peak measured to account for inter-sample peaks in lossy codecs?

True peak is measured using 4x oversampling, which reconstructs inter-sample peaks that can exceed the sample peak by 1–3 dB in lossy codecs like AAC or Opus.

What is the purpose of setting a true-peak ceiling of -1 dBTP?

A true-peak ceiling of -1 dBTP leaves headroom for reconstructed peaks so that downstream encoders do not clip.

What distortion risk exists if you master to 0 dBTP instead of -1 dBTP?

If you master to 0 dBTP, a subsequent AAC encode can push the waveform above digital full scale and introduce distortion that listeners hear.

Which single loudness master satisfies both podcast directories and YouTube without separate deliveries?

Master one stereo file to -16 LUFS integrated and -1 dBTP, verify with a BS.1770 meter, and let YouTube normalize upward.

Quick answers

Also worth reading: Audio cleanup for podcast transcription: Stanford PhD noise reduction thresholds vs 5% WER verify: Audio cleanup for podcast transcription: · Clean solo podcast audio: -16 Loudness Units (LUFS) AI vs manual: Clean solo podcast audio: -16 · 2026 A/B Test: -14 LUFS Boosts YouTube Watch Time by 12%: 2026 A/B Test: -14 LUFS