Before You Approve a Music Master, Match the Playback Loudness

Imagine receiving two masters of your new song. The second feels more immediate, so you approve it. Later, at a different playback level, you prefer the first version’s vocal and the way its chorus opens out. What exactly did the second master improve?

An independent musician needs a way to answer that question before release. Comparing versions at matched loudness creates a more useful listening condition: you can describe the musical differences without leaving a large volume difference tangled into the decision. The aim is to approve a sound you can explain.

The listener may hear a different playback level

Spotify’s loudness guidance, checked on 23 October 2026, says normalisation happens during playback. Its Normal setting targets −14 LUFS; louder masters receive a downward gain adjustment. Album playback preserves relative levels between tracks, while shuffle and mixed playlists use track-level adjustment. Spotify also says its web player and third-party devices do not use normalisation.

Apple’s Sound Check explanation, published on 26 March 2025, describes adjusting loudness between songs and shows listeners how to switch the feature on or off. These are documented playback behaviours. They give no basis for assuming every listener, device or service will reproduce your master at one fixed loudness.

The practical implication is to audition more than the untouched file at its original level. Include a comparison in which the versions have been level-matched, then consider the listening contexts that matter to this release.

Understand the two measurements you are requesting

Ask for integrated loudness and maximum true peak alongside the audio. In the EBU’s November 2023 R 128 recommendation, programme loudness is measured across the programme’s duration and expressed in LUFS, meaning loudness units referenced to full scale. Maximum true peak concerns the waveform’s highest level. These answer different technical questions; a peak reading alone cannot establish equal programme loudness.

Spotify currently recommends a master at −14 integrated LUFS with true peak below −1 dBTP. For masters louder than −14 integrated LUFS, it recommends true peak below −2 dBTP to reduce encoding distortion. Read those figures as Spotify’s published mastering guidance, with a defined context, rather than a universal artistic specification.

A measurement also leaves creative questions open. You still have to decide whether the lead vocal tells the story, whether percussion has the intended character and whether the transition into the chorus works. Write those priorities down before comparing revisions.

Set up a comparison with one clear purpose

For a small release, use the following listening procedure as an editorial starting point. It is a proposed workflow, not a laboratory test or a promise that everyone will prefer the same version.

• Keep both delivered masters unchanged. Make a separate comparison session or ask your mastering engineer for level-matched audition files.
• Use one playback route and leave the speaker or headphone volume at a comfortable, fixed setting while switching between versions.
• Match playback loudness using gain, supported by a loudness meter. Ask the engineer to explain the method if you are unfamiliar with the tools. Simply putting both files’ peak meters at the same value answers a different question.
• Compare the same passages: an exposed vocal, a dense section and an important transition. Then hear each version from beginning to end.
• Write a preference with a location and a reason. Record uncertainty when you cannot identify a meaningful difference.

Whole-track matching is a starting condition. A quiet verse and a loud chorus can still differ between versions that share an integrated reading. If one passage seems inconsistent, compare that passage separately with your engineer and record the additional adjustment. This follows from the distinction between the EBU’s integrated, short-term and momentary measurements; one whole-track figure does not describe every moment.

A hypothetical decision about one chorus

Suppose a fictional singer receives Master A and Master B. B arrives at a higher overall loudness. After level-matching, she finds that B makes the backing vocals easier to follow, but she prefers the beginning of the snare hits in A. These are imagined listening observations, not reported research findings.

She can now ask a useful question: can the revision retain B’s backing-vocal clarity while recovering the percussion character she preferred in A? The engineer may explain that the difference involves several interacting decisions and offer another option. Either way, the conversation has acquired a concrete musical purpose.

The singer need not select the quieter original or reject a dense, forceful sound. She needs to know what she prefers once overall level has been accounted for. Approving B remains a valid creative decision if its character serves the song.

Describe the change without guessing its cause

Apple’s Digital Masters technical brief describes compression, limiting and equalisation as different ways of controlling audio. Compression changes dynamic range; equalisation changes the level of particular frequencies. Playback gain is a separate operation. The implication is that reducing an already compressed master’s volume does not recreate dynamic contrasts removed during mastering.

That distinction helps you write better feedback. Try: “At the matched playback level, the final phrase is harder to follow in B; please compare the section around 2:10.” This gives a collaborator something to investigate. “Remove the limiter” prescribes a cause you may not have established.

Use the same approval criteria whether a revision comes from a mastering engineer, an automated service or your own session. If a service supplies a numerical score, ask what the score measures before allowing it to decide whether the music communicates your intention.

Approve a specific file and carry the decision forward

Keep a short approval note containing the exact filename, version, date, measurements, comparison conditions and any remaining concern. Distinguish the audition copy from the file approved for delivery. If another revision arrives, review that version before distributing it.

Once the audio decision is settled, connect it to the rest of the release. MaryChuks’ guide to building five useful assets from one song explains how an approved source record supports the listening page, clips, story and credits. Include the approved audio version in that record.

For your next release, ask for a level-matched comparison and prepare three musical questions before listening. You should finish with an identifiable file and a reason for choosing it that still makes sense when the volume changes.

Sources

• Spotify for Artists: Loudness normalisation — guidance checked 23 October 2026.
• Apple Support: Sound Check — 26 March 2025.
• EBU R 128 and EBU Tech 3341 — November 2023 editions.
• Apple Digital Masters technical brief — accessed 23 October 2026.