Skip to content
Home » Identifying Duplicates in AAF or OMF imports with Pro Tools

Identifying Duplicates in AAF or OMF imports with Pro Tools

The Dilemma

Every dialog editor or post production audio professional has been there: the client gives you an OMF (used interchangeably with AAF for the sake of this article) and it imports into Avid Pro Tools, but the clips in the timeline have no organization, no names, no metadata, and sometimes clips look like they could be duplicates of other clips, even though they have different filenames, levels, etc. These have to be individually checked to see if you’re losing important information, or if they truly are just duplicates.

So, as always, the first thing to do is to clean this timeline up. You have a near-sacred duty towards the following (in no particular order):

  • Director’s vision
  • Editor’s intent
  • Great sounding end product
  • Pre-dubbing and re-recording mixer’s workflow
  • Options readily available on the mix stage

Usually, to fulfill all of these, we will only REMOVE from an OMF transfer clips we KNOW for certain neither mixer will be able to use: blank clips, the boom from half a mile away when there are also radios available, etc. Every usable clip stays available in some form, either in a Junk or X track.But what about duplicates?

Obviously it’d be best to remove all duplicates except for the “best,” i.e. the one with the originally recorded levels, complete metadata, etc. Sometimes it can be hard to tell what is a duplicate and what is just another mic. Now, we could just listen to both. Solo one, solo the other, make a decision. And often that’s what I’ll do. Of course, in doing so, it often is necessary to clip gain them so that they’re the same level, since perhaps one is a boom mix track and one is the boom iso, which means they are the same content with different levels! In fact, there are many instances and reasons why the content may be functionally identical except for levels. Instead of clip gaining them, my dialog editing template now has a set of three tracks that are super convenient for doing this quickly, visually, and for lots of audio.

My Solution

  • Correlation Checker – Stereo Aux Track with iZotope Ozone Imager 2, routed to Dummy bus. This aux is never heard.
    • Correlation 1 – Mono Audio Track panned left 100%, routed to Correlation Checker
    • Correlation 2 – Mono Audio Track panned right 100%, routed to Correlation Checker
Pro Tools screenshot with correlation testing workflow pictured, including iZotope Ozone Imager
Basic correlation testing setup

I keep iZotope Imager open and small on my other monitor. With this setup, I can drop any pair of audio clips into A and B. As long as they’re timed identically, I can then hit play, and look at Imager. If I see a straight line IN ANY DIRECTION, I know the clips are identical. I can now select the “best” one as per metadata, levels, title, etc, and delete the rest. They don’t clutter my Junk or X tracks, and I’ve lost no information.

If I have a bunch of dupe-likely clip pairs on a set of OMF tracks, I can just copy them all into the Correlation Check tracks, and audition them quickly by using Tab to Transient, a jog wheel, or any other navigation method. Pretty soon, I’ve 100% verified the data on all deletable clips!

You could, instead, add sends to your dialog Work tracks that are permanently on, routed to the Correlation Check Aux. Since the aux is sent to a Dummy bus, you’ll never hear it. I’m still learning so we’ll see which I choose!

Obviously you could use any correlation meter. It doesn’t have to be iZotope Imager. I prefer a polar sample display because that straight line just gives me confidence in my decision.

Typical Scenarios

  • If the audio was recorded in camera, the Camera Op will often run the mic into both channels, with the right channel set lower than the left, as a backup in case the left channel clips. The audio is functionally identical except for the levels and possibly a clip, but it may be difficult to be sure of this.
  • The Production Sound Mixer records all the isolated tracks: boom, lav 1, 2, 3, etc to the field recorder but has also created mix tracks. Often, the mix tracks will be split so that Left has boom, right has lavs. This means that the left mix and the boom ISO may be functionally identical. Less discerning mixers (or ones who are hired to boom as well and don’t have time for a mix track!) may not pan the mix tracks, but if there are no lavs, then the Mix.L, Mix.R, AND the Boom ISO may be functionally equivalent!
  • Clips may have been damaged and duplicates extracted and used from another source, with the original still in the timeline.
  • Editor mixed mono and stereo elements in a single track. The mono elements may get duplicated upon import so that they still play at phantom center, depending on the NLE used.
  • Editor baked in gain effects to better hear the elements.
  • Editor (god forbid) duplicated the tracks over and over because “the terrible PSM mixed everything to quiet!” (Yes, I’ve gotten this before… just turn up the gain man!)

I’m sure there are more reasons these issues could occur. These are just some of the ones I’ve run into! Happy editing!