In the past I have used FCP with cameras that had their audio pre-synced (XL2, HMC-150). Now I've moved onto the 7d and things are a little different.

In Avid my workflow used subclips to create synced takes, but now that I'm in FCP I don't exactly know what to do to sync.

I've been using Pluraleyes, which is fantastic at syncing, but the post syncing workflow leaves a little to be desired. I've been using Philip Bloom's workflow (Sync with Pluraleyes, cut off the heads/tails, delete camera audio, create reference files for the synced clips). This isn't awful, but when you have 150-300 clips it can make it seem like Pluraleyes was doing very little work, by comparison.

Does anyone use a similar workflow on narrative film, and have they found a way to speed up the process?

I use SequenceLiner.

Everything is based on time-of-day timecode.

Audio is recorded to an Edirol R-44 via a wireless mic system.

The Sony EX1 also records audio via a camera mounted AT-4051a.

In post it's all synced up witih Sequence Liner. It gets me in the ballpark, then I manually refine the sync for groups of clips. Finally, in Soundtrack Pro, I'll select which tracks I want to hear, and subdue those I don't. The on-camera mic usually provides reference for fine-tuning sync. But the bulk of the audio I use is recorded with the R-44.

It's not hard to use one you understand the process. And I'm doing this solo.

I use a slightly streamlined process compared to PB's:

1. Sync with pluraleyes
2. Drag synced audio onto the camera audio tracks to replace them
3. razor blade synced audio at video in/out points
4. select each pair of synced audio/video clips and hit apple-L to link them
5. drag all newly synced & linked clips from timeline into a new bin

This skips the exporting a reference movie step, which both saves time and maintains the individual clips - including their file names in case you ever need to go back to the camera originals.

I also generally don't stop/start the audio recorder at the same time as the camera. For instance, in an interview I may start and stop the camera between each question, but the audio runs continuously. This way everything is in a single sequence after pluraleyes is done with it (or at least only one sequence per long audio take) which speeds up the process a bit.

It's still several steps, but it's easy to get into a flow with it and I've found I can process a couple hours of source footage in 10 minutes or less.