For some work, such as interviews, it’s simply not possible or practical to shoot multiple takes to get multiple camera angles. For these situations, it’s necessary to shoot all the desired camera angles simultaneously with multiple cameras, and then sync the resulting files in post to produce the final edit. However, for VDSLR shooters, using multi-camera video techniques can be a means to work around recording time limits.
As great as VDSLRs are at shooting video, virtually all manufacturers have imposed some kinds of recording limits; be it to avoid the European Union’s tariffs on video cameras, to avoid competition with their own video camera offerings, or simply to avoid having to deal with seamlessly recording across multiple files. Either way the results are the same. If you’re recording a 30-second clip or a 10-minute take for an indie movie, you’re fine. However, if you’re trying to record a longer running piece, such as an interview, those recording limits impose interruptions in the flow of the conversation.
Whether you’re shooting multiple cameras to support a more varied final edit, or as a workaround for short recording limits, the process for syncing and editing multi-camera video work is basically the same.
Equipment and Setup
Traditionally multi-camera synchronization has been done using a genlock signal or time code signal that is distributed to keep all the cameras in sync. With matching time codes, it’s trivial for NLE’s to be able to synchronize multiple video streams.
Unfortunately, for the VDSLR shooter, genlock sync is almost universally not a viable option; even many lower-tier professional camcorders lack support for multi-camera genlock synchronization. This is where post-process audio software synchronization comes in.
In this article, I’ll be specifically covering audio driven synchronization using Red Giant’s Plural Eyes software. However, similar functionality is starting to appear in some NLE’s, Premier Pro CC and the latest versions of Final Cut Pro X for example can now do multi-camera video sync using the audio tracks, with varying levels of quality and capabilities.
The exact video hardware you use isn’t that important, at least when it comes to the synchronization side of things. So long as your cameras manage to record decently clean audio, or you can feed them a clean audio signal from your audio recorder, Plural Eyes should be able to sync things.
Personally, I use a Canon EOS 5D mark 3 as my primary camera, and an EOS M as my secondary camera. This isn’t a recommendation for this particular combination; those are just the cameras I have.
That said, there are a few things I try to keep in mind when choosing cameras. First, I try to make sure my cameras have similar color profiles and are set to matching white balances. While this has nothing to do with audio and syncing, it does help with color grading and cutting between cameras. Anything you can do to make the output of the camera’s match in the field will help reduce what you have to do to match the images in post.
Second, I like cameras that have a mic in port, as I can back feed them with audio from my mixer/recorder instead of having to rely on their built in mics. Again, having clean nearly identical audio tracks helps Plural Eyes’s job easier.
Strictly speaking, Plural Eyes doesn’t require an external audio file to synchronize video tracks. It’s capable of sorting things out with the internal audio tracks on each video file. At a minimum, though, you need good audio on all the files, be it from the internal microphones on the cameras or from an external source fed into them.
My audio setup revolves around a dedicated audio recorder, a Tascam DR-60D specifically. Feeding my recorder are a pair of entry-level shotgun mics, a Rode Video Mic Pro, and a Rode NTG2.
With that said, I would strongly recommend using lavalier mics for interview work instead of shotguns. They’ll get the mic closer to your subject’s mouth, and therefore will give you a much better signal to noise ratio than the shotgun mics will. Plus using lavalier mics also means you don’t have stands and booms to worry about setting up or keeping out of frame.
Strictly speaking, you can make your setup as complicated or simple as you want. As long as you have a good audio track on all your video files, Plural Eyes should be able to sync things. I say should because I haven’t actually tried to push the limits to see how well it does with poor or worse quality audio.
I bring audio from my mics into my DR-60D where I record it as a 16-bit 48KHz BWF wav file. If you want to use higher bit-depths and sample rates, that’s certainly your choice. However, given the mics I have recording higher than 16-bit or 48KHz is just a waste of space for no useful gain in audio quality.
From the DR-60D, I patch from the camera out port back to my cameras. If you’re not using a DR-60D, you can use the line out port and a pad to drop the signal to mic-level the cameras expect on the mic in. For my two-camera setup, I simply use a Y cable to split the output from the DR-60D and feed it back into the cameras instead of relying on their built in mics.
Feeding audio to the cameras allows me to simplify camera setup and insure the audio matches making the post processing synchronization easier.
As far as simplifying the setup, I’ve already determined what gains I need to have my cameras set to with my DR-60D’s camera out set to 10 to approximately match the levels being recorded by it. This way I should get clean audio on my videos that doesn’t clip or peak any differently than the master track, and if for some reason the card in the recorder fails, I’ll still have reasonably good audio on the camera files themselves.
There is one caveat to this setup. Feeding a mic-level signal back to the cameras may not work well in all environments, especially when the camera is more than a couple of feet away from the recorder. Mic level signals are very low power, and the inputs on VDLSRs are unbalanced, as a result, they can pick up noise from EM sources nearby. As an alternative, it may be necessary to use a line out port and pad the signal down to mic level at the camera end or simply rely on the cameras to record usable sound.
Recording audio for multiple cameras isn’t any different from any other sync audio situation. Adjust the levels on your mics to get the speakers so their average volume is somewhere around -12dB (which is usually marked on the scale). Start rolling audio, start rolling video, and start the interview.
Some people clap after their audio and cameras are rolling to provide a sharp audio spike that can be used to sync things, however Plural Eyes doesn’t require this—though it doesn’t hurt either.
I hate to say it, but you’re on your own here. I find organizing files for video to be a fiddly and annoying process. I try to keep things organized. However, I’m not always successful at that.
For a project that involves a single take from multiple cameras, I often will just dump everything in one folder. For more complicated projects, I’ll try an organize everything into folders for scenes and cameras.
Whatever organizational scheme you choose, my recommendation would be to insure you have groupings by project or interview, then by camera for that interview or scene. When you import files into Plural Eyes, it doesn’t automatically sort files by camera metadata, so you need to know what files belong to what camera so you can place them properly in your project.
Red Giant’s PluralEyes is arguably the industry leader for syncing multiple video files to audio, though similar functionality is starting to appear in NLEs, such as Premier Pro CC. If you happen to have access to audio syncing via your NLE you can skip this step.
Fire up Plural Eyes, and you’ll find yourself staring at a new blank project.
- For each Camera
- Click the new camera button to create a Camera Bin,
- Select the camera in the bin list.
- Click Add Media to open add media window and add your camera files or drag and drop them on the camera.
- Add the audio track to the project
- Click the new Audio Recorder icon,
- Select the audio recorder bin in the media list.
- Click the add media icon and add your audio track or drag and drop it on the audio recorder.
If everything looks good on the time line—the right media files are assigned to the right cameras—then the project is ready for synchronization. To sync the files, click the synchronize button at the top of the window and wait while Plural Eyes lines everything up.
At this point, everything should be in the right places. If it didn’t work, you can try again using the “Try Really Hard” option from the Sync menu. If that doesn’t work, you’ll probably need to refer to the Plural Eyes user guide for fine-tuning and adjusting the sync.
Remember, the biggest disadvantage to this type of post-production synchronization is that it relies on having good audio with every video file. If you didn’t capture good audio at record time with one or more of your files, you may not be able to do anything about syncing them in post. This is why I go out of my way to redistribute the audio from my recorder back to the cameras as best as I can.
After everything is synced, the next step is to export the project data to Premier Pro. If you use another supported NLE, then the steps will be similar.
- From the file menu, choose export. This will open the Export Timeline dialog.
- Pick Premier Pro under Editor Formats.
- Make sure “Final cut Pro XML for Premiere” is checked.
- With the settings checked the way you want them, click Export.
The “Create a sequence with audio content replaced in video clips” has Plural Eyes creates a sequence where the video files have the master audio track replacing their recorded tracks in addition to the default where the video audio tracks aren’t replaced.
To summarize the steps in Plural Eyes:
- Create a new camera for each camera and add your media to it
- Create a new audio track for the master audio track and add your audio file to it
- Click the Sync button to synchronize the project
- Export the Project to FCP/Premiere by going File->Export then choosing Premiere Pro and checking “Final Cut Pro XML for Premiere” and clicking the export button
- Note where Plural Eyes saves the exported file
Now we want to get the Plural Eyes export into Premiere Pro so we can edit it. To do so we’ll need to first start Premier Pro and create a new project if we don’t already have one. When the project has opened up, click Import from the File menu, then navigate to the XML file Plural eyes created, select it and click open.
You should have a new sequence in your project window, along with the media files organized in bins the same way they were in Plural Eyes. There will be one or two sequences, depending on whether you checked the export with replaced audio option in Plural Eyes, in the _FCP_Premiere bin. If you open either of these sequences, you should see each of your cameras in plural eyes laid down as a separate video track.
Now for the editing part. You could edit this project the same way you would edit any other multi-video project, cutting and hiding content in the timeline. However, you can also edit this as a multi-camera sequence using Premiere’s mult-camera editing tools. One word of note, the steps described here applies to Premiere Pro CS6; Premiere Pro CC has changed the multi-camera interface again.
Create a new sequence in Premiere Pro with the appropriate settings for your content. Since the demo video along this article was shot with 2 1080p cameras and 1 720p camera all at 29.97 FPS, I’ll create my new sequence as a DSLR, 29.97 FPS, 1280×720 resolution file. If you’re working with 720p output, you may have to set up the sequence settings manually (remember to save a preset for the future) on the settings tab instead of picking from the presets list. Again, keeping the camera settings as similar as possible goes a long way towards helping make things smoother and easier.
Now drag and drop your “Synced Sequence” from the Plural Eyes bin, into the video 1 track on the timeline for your new sequence. Now you need to enable the multi-camera editing on the nested “Synced Sequence” in the time line. To do this you simply right click on the nested sequence in the time line, find the multi-camera sub-menu, and click enable.
The final step to begin editing with the Multi-Camera Monitor is to go up to the Window menu, and select the “Multi-Camera Monitor” option to open the window. From here, you can play through your video in real time, making cuts between cameras by either clicking on the track in the preview or by pressing the keys 1-9 corresponding with the camera/track you want to switch to.
You can also edit the multi-camera tracks on the timeline by making cuts, and changing the camera displayed by way of the multi-camera sub-menu when you right click on the sequence.
To summarize the steps in Premiere Pro:
- Import the Synced video file from Plural Eyes
- File-> Import
- Find the XML file that Plural Eyes generated
- Click open
- Create a new sequence that matches the sequence generated by Plural Eyes in resolution and frame rate
- Drop the sequence created by Plural Eyes (called either Synced Sequence or Synced Sequence Replaced) into the timeline for the new sequence
- In the timeline change the sequence to a multi-camera sequence
- Right Click the sequence,
- Select multi-camera sub menu
- Click enable in the multi-camera sub menu
- To bring up the Multi-Camera Monitor
- View menu -> Multi-Camera Monitor
- To Edit in the Multi-Camera monitor
- Click the record button, then play to start the video playing
Before I run, here are a couple of quick tips on the multi-camera editing thing.
If you need to adjust an edit point, you don’t have to play through the whole clip again in the Multi-Camera monitor. You can use the Rolling Edit tool to drag the cut on the timeline to move it around.
If you want to undo a whole pile of edits, you can simply delete the segments on the timeline and drag out one of the surrounding segments to fill the gap then redo your cuts.
All told, adding a second or 3rd camera can go a long way towards improving your interviews and as a whole, but can also be used to work around the short recording windows of VDSLRs in some situations.