Studio credit
Elephant Audio
Audio production · Sound design · Music production
LahorePKEngineering sound since 2018
Loading
Loading
Apple Music requires immersive audio not to exceed −18 LKFS and −1 dB TP, both measured per ITU-R BS.1770-4, delivered as a 24-bit LPCM Broadcast Wave Format ADM file at 48 or 96 kHz. The Atmos master must come from multitracks: upmixing from a stereo release, de-mixing stems out of one, and adding reverb to a stereo mix in the sound field are all disallowed.
Source
Transcribed from the published standard, read 2026-09-19. Check it yourself
Most platform loudness numbers in circulation are normalisation: the platform measures what was uploaded and applies gain on playback, and nothing is rejected. Apple's immersive requirements are the other kind. They are stated as limits a delivered asset must satisfy, in the same provider documentation that governs whether an asset is accepted.
So the practical posture is the opposite of the streaming one. For a stereo release the useful advice is to stop chasing a number and leave true-peak headroom. Here there is a number, it is a ceiling, and exceeding it is a delivery problem rather than a taste problem.
From Apple's Immersive Audio Source Profile, read 2026-09-19. Both measurements cite the same standard:
“A Dolby Atmos track must be created from multitracks or stems created from multitracks. Upmixing from a stereo release is not allowed.”
Apple disallows three separate routes to a fake immersive master: upmixing from a stereo release, extracting or de-mixing stems out of a stereo release, and taking a stereo mix, placing it in the sound field and adding ambience or reverb.
Read together they say something a loudness number cannot. The requirement is not that the result sounds a particular way — it is that the material genuinely exists as multitracks. There is no processing chain, and no separation tool however good, that satisfies a rule written about where the audio came from.
⚠️ That is also why an immersive mix cannot be retrofitted to a finished record whose sessions are gone. The decision to keep multitracks archived is taken years before anyone asks for an Atmos version, and it is the thing that makes one possible.
Stated so the page is not read as more complete than it is:
No — it must not exceed −18 LKFS, which is a different instruction. Anything quieter is permitted. Summaries that render the requirement as a target turn a ceiling into a goal, and mixing up to meet it spends dynamic range that the specification never asked for.
Not for Apple Music. The Atmos track must be created from multitracks or stems made from multitracks, and Apple explicitly disallows upmixing from a stereo release, de-mixing stems out of one, and placing a stereo mix in the sound field with added ambience. If the multitracks no longer exist, there is no compliant route to an immersive version.
Apple requires the immersive and most recently delivered stereo sources to differ in duration by no more than 50 milliseconds, and for gapless records each track boundary must fall within 0.5 seconds of the stereo boundary. In practice that means the immersive mix is conformed to the stereo edit rather than treated as an independent cut.
They are the same unit measured the same way and they are not comparable requirements. Apple's figure is a delivery ceiling for immersive assets. Spotify's is a playback normalisation target for stereo, applied after upload, which nothing is rejected for missing.
This page is the lookup. Broadcast and Streaming Loudness Specs for Commercials is the explanation behind it.
An Atmos version is decided by whether the multitracks still exist, not by processing. If they do, the mix is built from them and conformed to the stereo edit within Apple's tolerance.
Studio credit
Elephant Audio
Audio production · Sound design · Music production
LahorePKEngineering sound since 2018
Last updated: