Techniques For Editing Speakers and Conversations

[Andrew Rendell]

I find that you can usually get in at the start of a word within a sentence without too much difficulty, as long as the word is spoken reasonably clearly. Very often you need to do a fast fade up, FCP’s audio rubberbanding is frequently adequate, but sometimes you need an audio editor that goes to sample accuracy to cut finely enough. The end of a clip is usually a far bigger problem as you need the last word to be spoken as if it’s the end of a sentence. The only way that I have to do that reliably is to cut to the last word, or sometimes just the last syllable, from somewhere else in the conversation where it is spoken with the right inflection. Having an accurate transcript of the conversation is invaluable for that task as you can search for suitable substitutes without having to listen very carefully to every second of a recording. Again, having fine enough control of the edit point makes doing that in an audio editing program better than a video editing one, but whatever kit you’re using this takes quite a lot of practice to “get your ear in”.

Once you’ve got the audio sounding right, you then have to either cut the vision slightly off the audio cut or go to a cutaway. Occasionally putting a speed ramp on the vision over the edited sound can work, but you may have to juggle how the interpolation is done to get it to work (it takes a lot of practice to be able to judge when that trick’s likely to work for you).

With a group of speakers, I go for a single track for each speaker. There’s no advantage to be had from using stereo pairs unless the audio was mixed to stereo on location (which is very unusual in my experience).

[Steve Crow]

“With a group of speakers, I go for a single track for each speaker. There’s no advantage to be had from using stereo pairs unless the audio was mixed to stereo on location (which is very unusual in my experience)”..…

That’s interesting…what I have been doing is taking my mono signal from the wireless lav, duplicating it into 2 tracks and then setting it as a stereo pair within FCP….seems like you are saying this step is unnecessary.

I can see that with your technique I could play with the controls that determine if the sound is coming from the right or left speaker for each audio track (each person speaking) to give the soundtrack more of a sense that one person is sitting to the far right, another is in the middle somewhere and then there’s one person seated on the left.

[Bret Williams]

He’s just putting them on separate tracks for organization purposes. There is no reason to have mono sound duplicated onto a dual stereo track. All you do is increase the volume.

[Andrew Rendell]

yes, duplicating the tracks in that way is the same as centre pan and add 6dB of gain to a single track.

On an organisational note, I stick to one track per person, e.g., Bob on track 5, Fred on track 6, Sue on track 7, etc, and don’t put anything else on a track that’s been designated to one of the contributors. When you have lots of audio sources on a timeline it really helps to keep things as logical as possible, so use as many tracks as you need but don’t add them unnecessarily.

[John Fishback]

Another technique to cover an audio edit is to show a reaction of another speaker. Cut-in a second of two before the edit. You can add a bit of room tone so the edit is properly paced. You might be lucky and the reaction comes from the person speaking next. If you cut in some reaction shots here and there even when you don’t need them it sets the style so when you need the reaction to cover an edit it doesn’t stand out.

John

MacPro 8-core 2.8GHz 8 GB RAM OS 10.5.8 QT7.6.4 Kona 3 Dual Cinema 23 ATI Radeon HD 3870, 24″ TV-Logic Monitor, ATTO ExpressSAS R380 RAID Adapter, PDE enclosure with 8-drive 6TB RAID 5
FCS 3 (FCP 7.0.3, Motion 4.0.3, Comp 3.5.3, DVDSP 4.2.2, Color 1.5.3)

Pro Tools HD w SYNC IO & 192 Digital I/O, Yamaha DM1000, Millennia Media HV-3C, Neumann U87, Schoeps Mk41 mics, Genelec Monitors, PrimaLT ISDN