Audio workflow in post

[dlevanchuk]

So i'm dipping my toes in a full feature editing and getting a bit terrified with the workflow..

Typically, in my previous short I would create multiple audio tracks for each character, different tracks for room tone, sfx etc. All in one sequence. Then I would lock the video, do OMF export to my audio guy where go goes in to clean up the dialog, adjust/replace sfx, create Foley etc. Then he masters the whole projects, sends me one big wav file, and I drop it into my video..


Now, with this feature beast I'm completely lost.. I'm a 3rd scene into our movie, sorting out all audio and I have about 30 different tracks already! :-S I'm pretty sure feature edit work flow is different from what I ve been doing..
Should I edit each scene as independent sequence, just as if I'm editing a short film, and bringing everything together in the end with more or less managable chunks of audio? Or do I just keep creating 100+ tracks for my poor audio guy?

Please spare an advice for a proper audio track workflow for a feature . I would hate to send my audio tech 80 min OMF file with a ton of tracks and get "Are you f**king crazy?" email back from him :-(



EDIT:
nevermind, find it in older indietalk posts...
http://www.indietalk.com/showthread.php?t=27202

I'll just manage everything in smaller chunks

 

[Alcove Audio]

Let your sound editor do all of the sorting out of the tracks. KISS - Keep It Simple, Stupid! No insult intended; I apply this to myself constantly, except I substitute sh!thead when applying it to myself.

Your sound editor will have his/her own workflow. It is nice to have some basic organization when receiving the OMF, however. With each visual cut put the audio on alternating tracks. Most scenes will have two dialog tracks and a room-tone track with perhaps a few "extra" tracks. Label the tracks DX A, DX B, DX C, DX D, RT, WIFO A, WIFO B, WFX, A, WFX B, Score A, Score B, Music A, Music B. (DX = Dialog, RT = Room-Tone, WIFO = Wild Foley, WFX = Wild Sound Effects.)

The audio clips should have the same names as the visual cuts - Sc-12a_Tk-5_CU, for example. It's also a big help if you have more info in the metadata.

Make sure that your sound editor also gets the audio from the alt/unused takes. Put them in folders by scene.


Why would you end up with 30+ tracks, unless, of course, you are doing the sound edit yourself?


BTW, why don't you email your sound editor and ask her/him what s/he wants from you? You are a filmmaker, and filmmaking is all about communication! Make detailed notes for yourself, and maybe give them to your sound editor, about the characters, the meaning of the film, what each scene is about and why it is important to the film. What do you hear, and, most importantly, WHY you hear these things. Also include, if temps are not in the basic edit, a cue sheet for the score and the music. Give as much information as you can.

 

[AudioPostExpert]

The thread you linked to did not give much useful information.

The track layout Alcove suggested is more along the lines of the track layout your dialogue editor will create rather than necessarily the track layout you should supply to the Dialogue editor. Given the choice, I personally like to have the tracks organised by character in the OMF/AAF. It can be useful to have two tracks for a particular character (if necessary), so the character's dialogue can switch between tracks in line with the scene edits. Also, separate tracks for room tone, wild DX and wild FX as Alcove suggested. I don't like the audio clips to be renamed though but left as the Production Sound Mixer labelled them, provided he/she labelled them correctly/logically in the first place!

The metadata can make a huge difference to the speed/workflow of the dialogue editing process but is rarely, if ever, done correctly at the low/no budget level by the Production Sound Mixer. If a linked AAF is not being used and all the required metadata hasn't been recorded, depending on workflow, budgets and costs, it is possible for the picture editor to do more of the organisational work, which reduces the work and therefore the cost of audio post. For example, creating additional tracks containing alt takes and separated out polywavs (where more than one mic was used to record the takes).

Whether and how you break down the film into more manageable sequences depends mostly on how (and to whom) the film is to be distributed. For example, a film distributed on 35mm would have to be broken down into A/B reels, a DVD or DCP could be left as a single sequence and for TV broadcast the film would have to be broken down into parts to fit the broadcaster's schedule of commercial breaks. Your audio post person will be able to advise how to organise the OMFs/AAFs in all these cases.

As you can see, there are many different potential ways of organising your audio and OMFs/AAFs to ease the transition to audio post. You certainly do not want to be going beyond 100 tracks though (or anywhere near 100 tracks)! By far the best advice is the advice already given by Alcove, discuss all this (in detail!) with whoever is doing your audio post!

G

 

[Alcove Audio]

See, this is where the budget differences between APE and myself show up. APE generally works on decently budgeted projects with editors who have some idea of professional protocols/procedures; I work primarily on low/no/mini/micro projects edited by, ummm, less experienced people (to be polite).

I'm lucky if the sound clips are not spread over two or three dozen tracks with no coherence or organization whatsoever. Quite often there are four or five copies of a line "stacked" to increase the volume, many times separated by several tracks (original on 2, copies on 4, 17, 21 and 29 for example). And forget about room tones. I'm lucky if there's less than a dozen audio tracks with some semblance of organization, much less getting the alt/unused dialog tracks.

 

[AudioPostExpert]

You're right Alcove, up to a point, the AAFs are usually still quite a mess but I've never seen a line stacked, not even twice (thank god!), although I have seen that done occasionally with PFX and temp SFX. And, to be honest, only a handful of times has the picture editor ever broken out the alts/polywavs for me. Finding usable RT is also often a PITA as directors don't seem as diligent in recording decent RT as they were a decade or more ago.

From what I've seen, no one sync-locks the picture and sound during filming on indietalk. That's a bit of an added hassle for shorts but for features, the amount of time this must cost in post is mind boggling!

G

 

[Alcove Audio]

You know, I can't even remember the last time I got a pre-roll or a 2-pop.

 

[knightly]

When I'm sending myself audio, I give each character their own track for dialog (only thing I do in final cut is get that audio mostly done timing wise. I then OMF the project over to Soundtrack Pro and finesse each dialog track automating the levels and cutting out all of the non-dialog bits.
I then lay room tone under each scene.

Then I do a per character pass for footsteps (I don't foley this, I use a library and cut the steps track into individual steps, then match them to the body movement).
Per character "Hand" pass for sounds the produce with their hands by interacting with stuff
Per character Clothing pass for clothing movement sounds

Ambient environment sound FX pass to lay the sound bed for the rest of the work to be done.
FX for specific things like car engines or raptors in the distance.

Music with volume and EQ adjustments (as needed - sparingly) to "Duck" for the dialog and other sound FX -- the composer should add these cues into the music as it's being written/performed, but it's not always done.

Dialog + tone gets sent to a bus with its own fader
Environment to a bus
Steps, Hands and Clothing to its own bus
FX to its own bus
Music to its own bus

This allows me to automate the mix of the smaller components as larger groupings, so I mix within each bus, then mix the bus itself against the others.

I then export each bus as its own entity (AIFF or WAV) and reimport into final cut.

Around the same time, the Colorist is completing the grade and the new video and new audio are returned to a timeline where they are synched back together and exported as their own entity.

 

[AudioPostExpert]

I'm not sure how useful it is but as Knightly's done it, perhaps it would be useful to describe the layout of a fairly typical example of one of my mix sessions. I'm using ProTools, which has some useful features for audio post, one of which is the ability to create logical groups of channels (sub stems I suppose you could call them) and then have a VCA fader assigned to each group which can be automated and globally controls the level of the group. You can nest groups and VCA faders, as detailed below:

DX Group:
Sync DX Group: 4-12 (usually about 6) channels for the production dialogue + 2 Room Tone channels.
ADR Group: Usually 4-6 channels for any ADR + 1 or 2 (mono, stereo or 5.0) channels for walla.
Output to: DX Stem buss, Full Mix buss. Aux sends to 2-6 DX Stem reverbs.

SFX Group:
PFX Group: 2-4 mono plus 2 stereo PFX (production sound effects) channels.
Foley Group: 4-12 Foley channels.
Hard FX Group: 4-10 mono and 4-10 stereo channels for hard (foreground) sound FX.
BG FX Group: 4-10 mono and 4-10 stereo channels for background sound FX.
Ambiance Group: 4-12 stereo or 5.1 channels for ambiance sound FX.
Sound Design Group: 4-12 stereo or 5.1 channels for sound design elements.
Output to: SFX Stem buss, M&E Mix buss, Full Mix buss. Aux sends to 2-6 SFX Stem reverbs.

MX Group:
MX Group: 2-4 stereo or 5.1 channels for the music submixes.
Output to: MX Stem buss, M&E Mix buss, Full Mix buss. Aux sends to 1-2 MX Stem reverbs.

AAF Group: 2-12 mono and stereo channels containing the original picture editor's audio export.
Output to: The studio monitoring buss only!

If delivery specs require undipped stems or additional LtRt mixes for example, then the routing gets a little more complicated.

G