Imagine your China production partner has delivered the launch film. The voice recording and music are approved, and a regional media team now wants a radio spot and a Spotify audio ad. Before commissioning two exports, ask which receiving specification each file must satisfy and who will verify it.
ONCE recommends separating the editable audio asset from each destination master. That distinction lets an international brand reuse good production work while paying explicitly for adaptation, measurement and any necessary remix.
Two targets, two scopes
Checked on October 31, 2026, Spotify's US Audio Ads help page specifies integrated loudness of −16 LUFS, with a tolerance of ±1.5 LUFS, and a true peak ceiling of −2.0 dBTP. The ceiling is a maximum, not a level that every file must reach. This is an audio advertising page, not a universal music or video delivery specification. No publication date is displayed.
The EBU R 128 publication, version 5 from November 2023, recommends −23 LUFS programme loudness for broadcasting. Its full recommendation directs advertising and other short-form material to supplement R 128 s1. A buyer must still obtain the actual receiver's requirements; this is not a delivery contract automatically adopted by every broadcaster.
These documents support a production decision: check the destination before approving the master. They do not establish advertising performance, or show that raising a broadcast file by seven decibels will satisfy Spotify's requirements.
Average loudness does not resolve peak headroom
An integrated reading describes loudness across the item. True peak examines its highest waveform excursion. A restrained voice-led mix can contain a sharp product sound that leaves little room for a global gain increase.
Measure first. If a gain adjustment satisfies all applicable limits while preserving the intended sound, export and verify it. If it cannot, inspect the mix and treat the relevant peaks before measuring the new export. Pursuing a number without listening can leave the narration dull or the music pumping.
Passing a meter also does not demonstrate that a listener can identify the model name or follow the instruction. Technical acceptance and message comprehension need separate checks.
The ONCE destination, measurement and listening workflow
Confirm the receiver before finishing the mix
Record the market, platform, buying route, ad type, duration and channel layout. Attach the specification, retrieval date and the media owner's confirmation. A label such as “international master” is too broad. Where a broadcaster requests R 128, obtain its short-form and encoded-file requirements rather than relying on a headline target.
Measure the file being handed over
Keep a short report for each master: filename, language, duration, integrated loudness, maximum true peak, meter and delivery profile. After changing narration, shortening the spot, replacing music or re-encoding, recheck the affected export. A report for yesterday's file cannot certify today's revision.
Listen and assign the correction
At a comfortable fixed playback level, review the full spot and then its important words and strongest effects. Check headphones and a relevant everyday playback device. Write time-specific feedback, such as “the drum hit masks the final syllable of the product name.” If a platform preview is available, listen to that too. Record platform acceptance separately from audio approval.
This is a proposed workflow. ONCE has not performed a customer listening test or account upload for this article, and no conversion result is claimed.
Scope the order as adaptation, remix or a new audio execution
- Adaptation: the existing mix works, and gain adjustment can meet the complete destination profile. Budget for export and verification.
- Targeted remix: an effect or music peak prevents compliance. Retain the concept, revise the relevant passages and check the resulting voice balance and dynamics.
- New audio execution: the film depends on visible product actions that its soundtrack never explains. Rewrite and record a self-contained audio message before mastering it.
A useful estimating model is: total fee = base audio production + adaptation and checking for each delivery profile + agreed remixing or additional recording. This is a scoping model, not an ONCE rate card or an industry benchmark. Two specifications may share the same underlying session. Language changes and new edits should be priced for their actual production and verification work.
Make the next market easier to service
For one approved campaign, ask the production partner to hand over source recordings, suitable dialogue and music/effects stems, the mix session, destination masters and measurement reports. Asset delivery does not itself expand music or voice usage rights; confirm those separately.
This turns a one-off commission into a maintainable content service. When another market joins, compare its delivery profile with the approved set, then decide whether to export, remix or record. Brands gain a clearer acceptance process, and production partners can price continuing service against visible work. Explore ONCE's video production guidance when building that handover into a wider campaign brief.
