Editing Audio - workflow, what processes to use and in what order?

[ronin99]

Hello,

I was wondering what is the traditional workflow for editing audio (cleanup, amplify, etc..), the process used as well as the best order:

Would you Normalize first, then noise reduction, then EQ, then amplify? I just don't know what the best procedure is.




Also if you have a track where the dialogue audio levels vary wildly, some quiet, some loud, what is the best process or processes to use that will result in a consistent audio level?

 

[Alex H.]

Everybody's workflow will be a little different, but the basics are all pretty consistent.

First, I organize my audio. Nothing frustrates me like a messy timeline, and I cannot tell you how many times I have wasted the first hour of my session dealing with a sloppy OMF. Some video editors get it, and will keep their audio clips arranged per track, but others will simply ignore target tracks and let audio falla where it will as long as the video goes in the right spot. So if you are editing the video as well, do yourself a favor and make sure your audio tracks are organized as you go. My typical order looks like this:

A1: VO
A2-A4: Dialog
A5-6, A7-8: stereo FX pairs
A9-10: Stereo ambient sound
A11-12, A13-14: stereo music pairs

This is, of course, quite simplified. There may be more layers of SFX. There may be more layers of dialog. It depends on the project, but that's generally the order that I use no matter how many tracks of each kind. There will be more tracks in my ProTools session than what the NLE uses.

I'll usually go ahead at this point and set up any subgroups or aux patches for effects plug-ins (reverbs, delays, etc.) so that my mix session is ready when I get to that point.

After that, I will go through and clean up the dialog edits by smoothing fades and crossfades, adding or tweaking room tone, etc. Some EQ if needed, and compression if needed (moreso in broadcast than in "film"). If there are pops or clicks in any of the audio clips, I'll clean those up. And if noise reduction is needed on any of the dialog, I'll do that as well.

After the dialog edit, I'll move into creating the workd around the dialog through SFX, ambient sound beds, etc., and getting any effects plug-ins adjusted to taste.

From there, I smooth out any edits in music tracks, and replace/add music where needed.

Once all the pieces are in place, I start my mix passes. First, a mix pass (or two, or three...) on he dialog. Then, I bring in the ambient beds and SFX. Last, music. After the last mix pass on music, I set up any plug-ins I need on the master fader (compression, limiting... Again more for broadcast than "film") and do one (or two) more passes from start to finish just to make sure everything is in place and nothing is fighting with dialog.

This is still a bit of a simplified description, but it's the general order that I use. I and I rarely, if ever, normalize.

 

[Rick R]

I concur with Alex on arranging and cleaning up the timeline first. I've received nightmare OMFs as well.
Audio post has been discussed here many times previously (search). So I'll just paste in what I just stated on another thread over on the DVi forum:

It depends on the source material.. typically for dialog, I would first clean up the tracks (surgical editing and noise reduction); leveling volume envelopes; broadband or multiband compression; EQ (includes HP/LP filters).
All of the above are on a 'as needed' basis
Notes:
- If an audio track is very low, I would first 'peak normalize' to bring up the overall level.
- Surgical repair edits are usually easier/better with dedicated audio editing software (Sound Forge or Wavelab for instance).
- A brick wall limiter is almost always put on the master buss followed by an ATSC or EBU LUFS meter for broadcast projects.