Proxy Assumption

Previously, I talked about creating proxies to use for remote editing. The original footage remains in one location—not linked to the project—and only the proxies are used. I emphasized that you should make sure that the proxy files are created in such a way that they easily and faultlessly link up with the original footage. You can’t just assume that you did it right because if you didn’t, it may not be an easy fix.

Before you start any editing, it’s important to test to ensure everything will link. In addition, it might prevent you from having to take some of the drastic steps I mentioned last time, like changing filenames or timecode.

Testing is simply following the steps you’ll use to finish the project. Ingest your proxies into the edit software that will be used for the offline cut. Put all your proxy clips onto a timeline and then export that timeline via whatever method your finishing software requires: XML, project, etc.

Next, using another edit machine, import the sequence into your finishing software and relink to the original footage. Did the software find all the clips? And did it find the right clips?

Using another machine should accurately simulate what will happen when you finally relink the footage to the sequences. Moving to another machine ensures that all the links are “broken” to start with. But if you don’t have a second machine to simulate the workflow, try things like removing your original footage drive or renaming proxy folders and original footage folders and originals and then see if you can relink.

When you point to different folders, it might take a few steps to relink. Usually, the software finds all the clips in the selected folder and also in subfolders. I don’t consider a few steps like that a failure.

A failure is if you have to manually relink lots of files, one by one. A failure is if files link to the wrong clips. And a failure is if some clips can’t be linked at all. While doing this testing might seem tedious, it’s not as tedious as relinking files one at a time. Believe me.

If you’re using a specific application and have found a proxy workflow that tests well, don’t assume other applications will work just as well with that workflow. In my experience, two applications perform better than others in relinking footage:

  1. Blackmagic DaVinci Resolve does a good job. With its core focus originally on color correction, Resolve, at its inception, had to deal with relinking footage that didn’t originate in the application.
  2. I’ve often heard that one of Avid’s advantages is that it has “a strong and reliable database.” With Avid’s Media Composer being used in features and television series—think lots of footage—the software needs to have a database that’s reliable so that relinking is robust.

But even with those applications, testing is still important.

I’m not through with proxies. Next time, I’ll talk about using them to reduce the performance requirements on your edit machine.