Audio Post Production Workflow for Video Editors, From Spotting to Final Mix

"Audio question before I start, which export is the locked cut, and who actually signs off the mix, you, the director or the client?" That was the first message my agency got from a freelance sound designer the first time we handed a brand documentary to a proper sound team. I read it twice, because we had sent him three exports with FINAL in the file name and we genuinely had no answer to the second half of his question.

Look at it from the editor's side of the desk and those two questions, which picture he's building sound against and where the approvals happen, are basically the whole audio post production workflow, right. The sound work itself happens in the sound team's own tools, usually Pro Tools, and you don't need to know how to do any of it. What you do need is the stages in order, what you hand over at each one, and the moments where somebody watches and says yes or no.

So this is how my agency runs it now, stage by stage, with the review checkpoint I keep in PlayPause at each one while the audio work stays in the mixer's room.

What an audio post production workflow looks like for video

Sound design and mixing is the last big creative pass in film and video post-production, and it runs in a fixed order for the same reason picture does, right, each stage needs the one before it frozen. The picture locks and gets turned over, the team decides what every scene should sound like, and the dialogue editor cleans the lines. Then the sound designer builds ambiences, effects and Foley, a temp mix pulls it together, and the final mix finishes levels and loudness and prints the deliverables.

To be very honest, when a sound job goes sideways in my agency I can nearly always trace it to a spotting note that never reached the mixer, or a client leaving notes on MV3 after the mixer had moved to MV4, and very rarely to the audio work itself.

To keep the timecodes honest I'll use one illustrative project throughout, a 22 minute brand documentary with interviews and a licensed music bed. It goes to one freelance sound designer who also mixes, with the director and the client's head of marketing both signing off. In PlayPause that film gets one card, and each stage gets a custom status, Turned over, Spotting, Dialogue edit, Temp mix review, Final mix review and Mix approved. Custom statuses come with every plan, so anyone who opens the workspace knows where the sound is without messaging me.

Turning over picture to the sound team

Every picture change after turnover means the sound team conforms their whole session to your new cut, which is hours of work nobody quoted for. So the one rule I give first timers is that a rough cut or a nearly there cut never goes to sound, right, only a cut you're willing to call locked. My longer picture lock to sound handoff checklist covers the detail, and this is the short version.

The burned in timecode matters because a note at 00:12:40:05 then points at the same frame in your timeline, the mixer's session and the review link.

In PlayPause the locked picture with production sound goes onto the card as MV1, the status moves to Turned over, and the sound designer gets one password-protected link. If picture changes later, the new cut stacks on the same card, I revoke the old link instantly and write one comment listing what moved and by how many frames. That's the habit behind keeping a sound designer off an outdated cut. I also settle the second half of that original message here, right, by writing the name of whoever approves the mix as the first comment on the card.

Spotting sessions and how to capture spotting notes

A spotting session is a screening where the director, the editor and the sound lead stop at every moment that needs a sound decision, where music comes in and out, which scenes need more room tone, where a silence is on purpose, and so on. On most of our projects it happens over a video call, and the catch here is that notes from a call live with whoever was typing, right, as a list of timecodes the mixer then cross references by hand.

So I run spotting on the review link itself and write every decision onto the frame while we talk. Range comments carry most of it because sound decisions cover a stretch, for instance a range from 00:04:12:00 to 00:04:51:10 saying "keep the office ambience thin, she's alone for the first time and it should feel a bit cold". A single frame comment handles hard sync like a door slam on the cut. Drawing on the frame helps when the source is ambiguous, say circling the background television that should be playing a news bulletin.

Every note also gets an owner, so I @mention the dialogue editor on noisy lines and the sound designer on effects. I start each comment with DIA, MX or FX, which sounds fussy until you're the mixer looking only for the music notes. There's more on that in keeping dialogue and music notes separate, and the time-coded audio annotation page shows how these comments look on an audio file.

Dialogue edit review against the script

The dialogue editor makes every line clean and continuous, cutting between takes, filling gaps with room tone, removing clicks and hum, and flagging lines that need ADR, where the speaker re-records in a booth. What catches first timers out is that a dialogue edit can quietly change what people say, right, a word clipped in a noise fix or an alternate take with slightly different phrasing. On a brand film a changed word in a product claim turns into a legal conversation.

So on the Agency plan and up I lean on AI transcription, which turns the cut into a clickable transcript. When the dialogue pass comes back as MV2 I read that transcript against the approved script, and wherever a line doesn't match I click it, land on that exact frame and comment there. For instance, if the script has the founder saying "we deliver in two days" and the transcript reads "we deliver in a few days", it usually means the editor swapped to a cleaner take. So my comment names the changed words and asks the director whether the cleaner take or the exact claim wins.

It's AI, so a mismatch means go and listen rather than proof, but trust me on any level, reading 22 minutes of dialogue against a script beats scrubbing it in headphones. The SRT export gives you a head start on captions too. The checkpoint for this stage is an agreed ADR list, one comment per line on MV2 with the director replying approve or reject in the thread, so nobody books a booth until every line on it has a yes.

Sound design and temp mix review

Sound design gives the film its world, the ambiences, hard effects, Foley footsteps and cloth, and on a documentary, right, it's mostly room tone and a sense of place. After a stretch of quiet work you get a temp mix, a rough balance of dialogue, music and effects so the director hears the direction before the finishing work starts.

The trap is that the client hears the temp and spends their notes on loudness that was always going to change. So when the status moves to Temp mix review I pin a comment saying levels will move and we want notes on choices, like whether the music drops out before the founder's story. The sound designer adds the temp as MV3 through guest version upload, which on the Agency plan and up needs no account, only a name and email. Then I open side-by-side version compare with MV2 against MV3, mostly to confirm the picture under the mix is the locked picture, because a mix laid back onto the wrong export happens more often than anyone admits.

A usable mix note names the stretch, the element and what you want instead, for instance a range from 00:17:10:00 to 00:17:45:00 saying "the strings fight the interview, dip the bed until she finishes". I broke that skill down properly in feedback a mixer can act on.

Final mix approval and delivery

The final mix happens in a calibrated room, where the mixer balances everything and prints the deliverables, usually the full mix, separate dialogue, music and effects stems, and an M&E if the film might be dubbed. Loudness is the technical half of final mix approval, right, and the target comes from the delivery spec you sent at turnover. European broadcast under EBU R128 asks for minus 23 LUFS integrated, US broadcast under ATSC A/85 sits at minus 24 LKFS, and each streaming platform publishes its own. The mixer measures against that, and I'm very very clear with clients that a browser link is where they approve choices, never loudness.

The final comes back as MV4, and before chasing anyone I check the who-watched analytics, which show who opened the link, when and from which city. That tells me whether silence from the head of marketing means no notes or nobody pressed play. The approval is written as a comment on MV4, "MV4 approved for delivery", the status moves to Mix approved, and any late fix stacks on top as the next version so the approved record stays intact. If you run a whole slate of films like this, the post-production house and film and documentary production pages go into how bigger teams set it up.

A mix can stretch past a month once legal gets involved, so a 90 day link on the Agency plan at $19 a month outlives the schedule. Agency also includes Playbooks, so I keep a brand Playbook per client with the permanent sound rules, like the founder's lines never sitting under music and how stems get named. Each film then gets a folder Playbook whose deliverables checklist isn't done until every item is ticked.

Mistakes I see on a first sound turnover

The most expensive one, and I'm pretty sure every editor makes it once, is turning over too early and then sending two frame trims the sound team has to conform under, until the goodwill is gone. Right behind it is the unnamed mix approver, the exact gap that original message exposed, which always surfaces on the day the director loves the final and the client wants the music louder. The one that stings at the end of the day is forgetting to quote stems and an M&E, because sound teams price those separately, so I write them into the statement of work before turnover rather than discovering them on the invoice.

If you cut in Premiere Pro, the PlayPause Premiere Pro panel shows the sound team's picture notes inside your timeline, so when the sound designer flags that shot 42 is two frames short, you click the comment and the playhead jumps to that frame. In any other editor you export, upload, review and then work through the notes in your own timeline, as my Final Cut Pro to DaVinci Resolve workflow shows for the grade. The post production checklist template and the structured comment workflow for sound editors go deeper on the picture stages and the notes. Does that make sense, right, the sound team owns the sound, and your job is making sure they build it against one picture, one set of notes and one named approver.

Frequently asked questions

What are the stages of an audio post production workflow?

For most video projects it runs turnover, spotting session, dialogue edit, sound design, temp mix and final mix, then delivery of the full mix, stems and an M&E if needed. Small teams merge stages, right, so one person might edit dialogue, design and mix, but the order holds because each stage needs the previous one frozen.

What should an editor send the sound team at turnover?

A locked reference video with burned in timecode and a two pop, an AAF or OMF with handles, the production sound and sound reports, the script with suspected ADR lines, a music cue list, and the delivery spec. I add one review link to the locked picture, with the mix approver named in the first comment.

How do I organize a sound post workflow for a small team?

I keep one card per film with a custom status for each stage, so every handover has a visible checkpoint. The mix approver's name goes in the first comment, every note is tagged DIA, MX or FX with its owner @mentioned, and each pass stacks as the next MV, so the mixer always knows which version is current.

Can a client approve the final mix from a review link?

They can approve the creative choices from a link, whether a line is clear or the music drops out in the right place. They shouldn't approve loudness or low end from a browser, because the link plays a streaming copy on speakers nobody has checked, so I treat final mix approval as two sign offs, creative in PlayPause and technical by the mixer.

If you're handing your first film to a sound team, set up the card, the statuses and the review link before turnover, and the later stages mostly fall into place. You can see which plan fits your team on the PlayPause pricing page, and every plan starts with a 7-day free trial, so you can try the whole setup on a real project first.

So yeah. That's my way of saying it.