In Part One, we looked at how to hide your interview subject with the camera. In Part Two, we looked at how to hide your interview subject with software. But there is one thing that can still identify your interview subject – his or her voice.
Sounds Like Fingerprints
Voices can identify someone without even having to look at their face. Fran Dresher has a distinctive voice. I immediately recognize Jeff Bridges doing voiceovers for commercials. I know the sound of my wife’s voice, my child’s voice, my parents’ voices. Think of all the people you know. No one has the exact same voice. If you know someone well, one of the ways you identify them is by their speaking voice. Tone, pitch and rate of human speech makes people’s voices as easily identifiable as fingerprints.
When it comes to an interview subject who wishes to be hidden, you need to ask them if they want to have their voices disguised as well. Sometimes your subject will not be worried about their voice. Your interview subject may just not want people walking down the street to recognize them. Others are worried that an acquaintance may know who they are by the sound of their voice, and want you to disguise it.
This is where pitch comes in.
Almost every TV news station I’ve ever worked at or been in deals with images, i.e. video, the exact same way. Audio is different… When shooting, sometimes your A-Roll audio will be on Channel 1. Other stations prefer Channel 2 for A-roll and Channel 1 for nat sound. A lot depends on what format you are shooting on. Because of all of these variables, I’m not going to give you a step-by-step checklist for how to disguise your interview subject’s voice. Instead, I’m going to stick with theory. This will work, but you may need to ask your chief editor or edit-bay engineer for specifics for your station.
When I do a project in Adobe Premire Pro, I prefer to have seperate mono tracks to work with. My Channel One is my interview subject, in mono. My Channel Two is my nat sound mic. As I work in Premiere Pro, I have set my tracks to be mono tracks. My output is set to make a stereo mix of all the channels. My Odd channels, tracks 1, 3, 5, et al., are set to the Left in my final stereo mixdown. My Even channels, tracks 2, 4, 6, et al., are set to the Right in my final stero mixdown.
When I want to disguise a voice, I first make a copy of their audio track from my Channel 1, which on my timeline is Track 1. I paste that copy onto my timeline in Track 3. When I do my final output, Track 1 and Track 3 will be merged on the same output track, forever mixed together.
Then I look through my audio effects until I find Pitchshifter.
Pitchshifter does exactly that – it changes the pitch of the audio in your clip. I add the effect to my Track 1 and my Track 3. On Track 1, I adjust the pitch higher. On Track 3, I adjust the pitch lower. Once the piece is cut and I’ve done this to each clip that needs disguising, I export it. In my Premiere Pro setup, my odd audio tracks, 1, 3, 5, etc., are mixed down to one audio track.
Other non-linear editors have different titles for their pitch adjustment. If you don’t know what it is, ask a colleague or check the manual.
Some of you may be asking, “J, why do two tracks? Can’t I do this with one and just make her voice higher or lower?” You can. Unfortunately, anyone can record your story, dump it into their computer, and use a similar program to adjust the pitch of the person interviewed until they come close to who it could be. That could be disastrous for your interview subject, which in turn could be disastrous to your credibility and career. By adding an opposite second instance of that voice, no one can re-engineer the pitch of the voice to it’s original form. They can come close, but there will always be a higher / lower instance that will make 100% identification of your interview subject nearly impossible.
If the audio is difficult to understand after it has been disguised, subtitle the interviewee’s dialogue.
Good luck, and happy cutting.