Streaming loudness targets: what the platforms publish.

Yes, in Spotify’s own words: “We adjust tracks to −14 dB LUFS, according to the ITU 1770 standard.” It also publishes the three Premium playback settings, Loud at −11, Normal at −14 and Quiet at −19 dB LUFS, and asks that a master louder than −14 LUFS keeps its true peak below −2 dBTP. Spotify in detail. It turns quieter tracks up as well, as far as the true peak allows with 1 dB to spare, and it normalises a whole album with one gain when it is played as an album. Normalisation is off on its web player and on third-party devices.

SoundCloud recommends −14 integrated LUFS: “Target the loudness level of your master at −14 dB integrated LUFS and keep it below −1 dB TP (True Peak) max.” For a louder master it asks for −2 dB TP. It says it applies loudness normalisation as tracks play, but it does not state the level it normalises to.

For a stereo track, Apple publishes no loudness figure. Neither the Apple Digital Masters document nor the Sound Check article contains one; the Digital Masters document recommends leaving “at least 1 dB of headroom”. Apple’s asset guide does set limits for Dolby Atmos and 5.1/7.1 deliveries: “The integrated loudness value must not exceed -18 LKFS”, and “True-peak level must not exceed -1 dB TP”. The −16 often attributed to Apple appears in none of these; AES TD1008’s track figure is also −16. Apple Music and Sound Check in detail.

No. YouTube documents a “Stable Volume” setting with no figure, and says it is turned off for YouTube Music and official music videos. The −14 usually quoted for YouTube is not in YouTube’s help documentation. What YouTube does publish.

No. TIDAL’s help centre has no article on loudness; a search for “loudness” returns zero articles.

No. Amazon Music publishes a how-to for switching normalisation on, and no figure.

No. Deezer documents a user toggle for volume normalisation, with no number.

AES TD1008, an Audio Engineering Society technical document addressed to distributors of streams: −16 LUFS when a service normalises track by track, −14 for the loudest track of an album when it normalises by album, and a maximum true peak of −1 dBTP at the input of a lossy encoder. It is not a platform, it says it does not make recommendations for content production, and no platform has adopted its −16 in its own documentation.

Spotify normalises to −14 LUFS and says so in its own words: “We adjust tracks to −14 dB LUFS, according to the ITU 1770 standard.” That is what it plays back at, not a rule about how you must master. Spotify also asks that a master louder than −14 LUFS keeps true peak below −2 dBTP.

For a stereo track, Apple publishes no loudness figure. Neither the Apple Digital Masters document nor the Sound Check article contains one. Its asset guide sets a −18 LKFS ceiling and a −1 dB TP limit only for Dolby Atmos and 5.1/7.1 deliveries. The −16 LUFS often attributed to Apple appears in none of these documents; AES TD1008’s track figure is also −16.

Two, as of 19 September 2026, both at −14 LUFS: Spotify, which normalises to it, and SoundCloud, which recommends it and says it normalises without stating the level. AES TD1008 recommends −16 for track normalisation and −14 for the loudest track of an album, but it is a technical document addressed to distributors, not a platform. For a stereo track, YouTube, TIDAL, Amazon Music, Deezer and Apple publish nothing — TIDAL’s help centre returns zero articles for a search on “loudness”.

No. Turning a −8 LUFS master down to −14 does not produce a −14 LUFS master: the limiting that got it to −8 is in the file and survives the gain change. Normalisation is also not always on — Spotify’s documentation says it is off on its web player and on third-party devices. How loud to master is a real decision with real consequences.

−1 dBTP is the general recommendation: AES TD1008, Spotify and SoundCloud all state it. Spotify and SoundCloud both ask for −2 dBTP if your master is louder than −14 LUFS, because, in Spotify’s words, “louder tracks are more susceptible to extra distortion when encoded for streaming.” TD1008 notes that high-rate encoders, such as 256 kbps, may manage with −0.5 dBTP. Every loudness choice here uses −1.0 dBTP by default. The −0.1 dBTP option trades that margin for peak room, so expect an MP3 or a stream of that file to peak higher than the WAV.

Because the platforms change them, quietly, and most of the figures in circulation were never published in the first place. Every quote on this page carries the date it was checked for that reason. If you are reading this long after September 2026, check the source documents linked above rather than trusting the table.