Nobody leaves a video because the grade was a shade cool. They leave because the room is ringing, the level jumps between speakers, or the music is sitting on top of the dialogue. Audio here is treated as its own discipline with its own session, not as something applied to a video timeline at the end of a long day.
Five stages. The first is the only one that cannot be redone later, which is why most of the judgement goes there.
Location recording with directional and lavalier microphones, and controlled-room recording where the source deserves it. Rooms are listened to before anything is plugged in: a hard-surfaced meeting room with a glass wall is a worse problem than any microphone can solve, and the fix is moving, treating or rescheduling rather than repairing it afterwards. Multitrack and isolated channels throughout, so one bad take does not cost the session.
Narration, voiceover, corporate and e-learning reads, podcast and interview dialogue. The editing is the unglamorous part: breath and plosive control, mouth-noise removal, timing and pace, and the level matching that makes two people recorded in two rooms on two days sound like one conversation. Noise reduction applied conservatively, because heavy-handed denoise is more obvious than the noise was.
Recording in person or remotely with local double-ended capture rather than the call audio, which is the difference between a podcast that sounds produced and one that sounds like a call. Per-episode edit, music beds and idents, chapter markers, and delivery at the loudness the podcast platforms expect. Set up once as a repeatable template so episode forty sounds like episode one.
Balance, dynamics, corrective and tonal EQ, spatial placement, and the music-under-dialogue judgement most rough cuts get wrong by about six decibels. Mixed to translate: checked on monitors, on headphones and on a single small speaker, and checked in mono, because a large share of the audience will hear it from a phone lying flat on a desk.
Final level, tonal balance and limiting to the loudness target for the destination. Streaming, broadcast, podcast platforms and music services all want different numbers, and delivering one master to all of them means three will turn it down or push it up. Targets are agreed in the brief and measured on delivery rather than eyeballed.
Public-domain reference frames. Sources and licences.
Audio sessions are recorded, not photographed — there is rarely a spare pair of hands to make pictures with, and a session is not the moment to find one. These licensed photographs show the room the work happens in.
Reference frames — not this studio’s work. Ten licensed photographs of recording and mixing. Creators, licences and sources.
All three are listed rates: per episode, per voice session and per mastered track.
Taken on for the studio’s own films and for video finished elsewhere. The deliverable is not just a mix: it is a mix plus the separated stems that let the piece be re-versioned, translated or re-cut later without going back to the beginning.
Audio post against a locked picture is quoted with the edit, since the two share a timeline. Music mixing and mastering is a listed per-track rate.
For recording work, a photograph of the space and a thirty-second phone recording of it sitting empty tells more than any description — enough to say whether it will work, what it needs, or whether somewhere else would be cheaper than fixing it.