Use case · AI content creation

AI music and sonic branding with clear rights.

AI can compose audio logos, music beds and variations in many lengths and moods, which makes a consistent sonic identity affordable across every market and format. The risks sit in rights and loudness: the brand needs clear commercial licence terms, a motif distinctive enough to protect, masters that meet broadcast and streaming loudness targets, and nothing that imitates a famous track or voice.

Audio logos, beds and the system around them

A sonic identity is a small kit, not a single jingle: an audio logo of a few seconds, a brand theme, music beds in several moods and tempos, and edits for 6, 15 and 30-second ads with clean endings. Deliver stems as well as mixes, so editors can drop the melody under dialogue or extend a bed without a new composition. Keep the motif recognisable across uses, with the same intervals and rhythm whether it is played on piano for a calm film or on synths for a launch. Markets can change instrumentation and arrangement, but the motif should survive, or the identity fragments.

Licence clarity and protection

Before using generated music commercially, read the terms: the brand needs rights for advertising and broadcast, in every territory and for as long as the identity lives, not a personal or limited-use licence. Ownership is a separate question. In the US, material generated without sufficient human authorship is not protected by copyright, so a composer's hand in the core motif makes it easier to defend. Trade mark law offers another route: the EUIPO has accepted sound marks filed as audio files since October 2017, and a distinctive audio logo can be registered in the EU, the UK and the US. File before the campaign, not after a copy appears.

Loudness norms

Loudness is regulated for broadcast and normalised almost everywhere else. European broadcasters work to EBU R 128: −23 LUFS integrated, with true peaks no higher than −1 dBTP. US television follows ATSC A/85 at −24 LKFS under the CALM Act, which requires ads to match the loudness of the programmes around them. Music streaming normalises playback, with Spotify's default at −14 LUFS, so an over-compressed master is simply turned down and sounds flatter. Master each delivery for its destination rather than one loud version for everything, and check the audio logo at each target, because a three-second sting is where loudness jumps are most noticeable.

Avoid soundalikes of famous tracks

Creative teams often cut a rough edit to a famous track and then ask for something like it. That brief is how brands end up in court. In the 2015 'Blurred Lines' case, a US jury found that the song infringed an older one even though, as a dissenting appeal judge later put it, the two differed in melody, harmony and rhythm; the verdict largely stood on appeal. Voices are protected too: in Midler v. Ford (1988) and Waits v. Frito-Lay (1992), US courts held that imitating a famous singer's distinctive voice in an ad was actionable. Never brief generation with an artist's name or a reference track, and have a musicologist check anything that sounds close.

Updated 25 September 2026

Questions

AI music and sonic branding, answered.

Can an AI-generated audio logo be registered as a trade mark?

A distinctive sound can be registered as a trade mark however it was made, provided it identifies your brand and does not copy an existing work. Copyright is less certain for purely generated music, so a registered sound mark plus a composer's involvement in the core motif gives the strongest position.

What loudness should brand audio be mastered to?

To its destination. Broadcast in Europe expects −23 LUFS under EBU R 128 and US television −24 LKFS under ATSC A/85; music streaming platforms normalise, commonly around −14 LUFS; podcast apps publish their own recommendations. Deliver separate masters rather than one loud file, and check the audio logo at each level.

Does the studio include music for video?

Yes. Synthetic White makes audio as well as images, video and copy, so music beds for video can be generated with commercially safe models by default, alongside voice-over and dubbing into 30 languages with timing preserved. For a bespoke sonic identity, the licence terms, trade mark filing and loudness masters described above still apply, whoever composes the motif.

Get started

See your AI studio generating this week.

Thirty minutes. A live studio in your colours, your brand hub loaded, a real campaign brief.

See plans