The master is never touched
Magic Mode separates speech from music, so dialogue is replaced while the track underneath stays exactly as mastered.
For artists & labels
Streaming made the songs global overnight. The interviews, documentaries, behind-the-scenes and campaign content around them are still made for one market. Octavia localizes everything that is not the record, without ever touching the record.
The specific reasons this content usually stays in one language.
Fans in every market stream the release. Only one market gets the interview, the studio breakdown and the tour film.
A release cycle does not wait weeks for localized assets. Content that arrives after the moment has no commercial value.
Years of interviews and documentary footage generate nothing, because localizing them was never economic.
The parts of the platform that matter for this work, rather than the full feature list.
Magic Mode separates speech from music, so dialogue is replaced while the track underneath stays exactly as mastered.
Voice preservation carries an artist’s delivery into other languages, with consent and control over what may be produced in that voice.
Diarization separates interviewer from artist, so each keeps a distinct voice instead of collapsing into one narrator.
Turnaround measured against a campaign calendar rather than a dubbing studio’s schedule.
Burned-in, word-level captions for muted feeds, where most music content is actually discovered.
Batch-process an archive of interviews and documentaries that has never earned anything in a second language.
Four steps from the file you already have to every language you need.
Interview, documentary, EPK or campaign cut — the finished file is enough.
Artist names, track titles and label terms are pinned so no pass tries to translate them.
Prioritise by where the streaming numbers already are, not by population.
Dubbed audio, subtitle tracks and captioned social cuts.
One upload in, every format you need to publish out.
Including the ones where the honest answer is a limitation.
No. Octavia localizes spoken content — interviews, narration, commentary, documentary and campaign material. The music itself is preserved, not translated. Magic Mode exists precisely so the mastered track passes through untouched while speech is replaced.
Technically yes, and it should be governed rather than assumed. Get written consent covering the specific languages and content types, decide internally who may authorise content in that voice, and settle whether audiences are told it is synthesised. Many labels use the artist’s voice for owned channels and a professional voice for paid media.
That is the normal case and it is handled by dialogue separation. Verify the result on passages where the music is loudest, since that is where any artefact would be audible — but for typical interview and documentary beds it is clean.
Follow your streaming and social data rather than intuition. The markets already consuming the music are the ones where the surrounding content converts, and they are frequently not the largest countries by population.
Practical guides from the Octavia editorial team.
The same engine, pointed at a different kind of work.
Start with one interview and two markets, and measure it against the streaming numbers you already have.