Dialog Editing Part 3: The challenges of dialog editing & mixing

This post may contain affiliate links. Please read my disclosure for more information.

Working with vocals (for music) is similar to working with voiceover. Both are recorded in a recording studio with a stationary mic appropriately placed for the best sound quality. There’s not a lot of background noise or bleed of other sounds and if there is, the engineer can usually do something about it.

Production dialog is more bit like working with a live music recording. People move around (on stage or on set) and you can’t always put mics where you want them. You have little or no control of other sounds being picked up by a mic. When you’re mixing music that was recorded live, one common challenge is mic bleed (such as hearing drums in every mic). With production dialog, you’re fighting nature bleeding into every mic.

Production dialog can pick up extraneous sounds like:

  • Constant background noise (like nearby traffic, a river, ventilation or equipment humming)
  • Varying background noises (dogs, birds, other people, plane flying over, car horns or cars driving by at close distance)
  • Room reflections/reverb

On top of it, boom operators and actors move around so the sound of a character (eq, balance, presence etc) can change during a scene. A boom op has to keep the mic out of the camera shot – which affects placement. Lav mics are typically placed so they won’t be seen which can sometimes mean a small sacrifice in sound quality to look better.

The challenges of editing production dialog

The real challenge of editing production dialog is the background noise. Check out a little of this video of an interview happening in a glass recycling facility. How hard is it to follow the information they’re trying to give when there’s glass being poured and trucks driving around and beeping?

Did you catch the bad audio edit of the glass around 3 minutes? That’s a perfect example of what can go wrong when you’re dealing with noisy production dialog.

This interview was likely cut down from it’s original length (by a picture editor). It could have been an hour interview that they’re condensing into ten minutes. This can wreak havoc on the background noise which now is coming and going between edits. To add to the challenge, you may not be able to use the sound that happened right before or after (called “handles“) to smooth a bad edit. You can mask the problems sometimes with added sound fx (more glass) but that could make it harder to understand. Ultimately, you still have to find a way to make the dialog edit work on its own and sound sound natural (as much as it can).

The challenge having lots of material

Another challenge dialog editors and mixers face is editing between a lot of material. For example, in a documentary, a filmmaker may interview the same person multiple times but on different days and different locations. In the edit of the film, they might use a line from one interview and the next line from another interview of the same person in a totally different location.

When this happens, the goal isn’t necessarily to have two interviews that both sound as good as possible. It might be to make them sound closer – similar enough in sound that it doesn’t draw attention to the viewer. It’s possible the two interviews sound drastically different depending on if it’s the same mic, the distance the mic was placed, if it was indoor or outdoor, how much background noise or reverb there is, etc.

One job of the dialog editor is to help take attention away from the fact that these locations sound totally different. Sometimes you need a quick crossfade and sometimes a really long one. Maybe nudging the timing a frame here or there will help (if it doesn’t affect sync). One of the main concerns for the dialog editor is the background noise – is it coming and going? Does the sound change? Does the shift happen in a way that distracts the audience or can you hide it in a way they won’t notice?

Dialog intelligibility

The number one goal of dialog mixing is intelligibility. What’s tough about dialog editing and mixing sometimes is what we consider intelligible (or not) may be different from someone else.

Dialog intelligibility is probably 60% dialog mixing and 40% what’s going on around it. There’s people who aren’t bothered at all by background music and can understand the dialog just fine. But, there’s people where any background music interferes with their ability to understand dialog.

Music examples

A good analogy would be lyrics in music. Here’s a couple examples of songs who’s lyrics are very hard to understand:

The issue isn’t just the band. “Louie Louie” doesn’t have compression on the vocals so some words get lost under the instruments. “Rock the Casbah” has a reverb and a delay on the vocal. It’s an interesting effect for a song but it compounds the problem of intelligibility because of the singer’s strong accent.

Here’s an example in post-production. The BBC received complaints about the show Happy Valley being difficult to understand. Check out these two trailers for the show. Do you find one easier to understand than the other?

It’s an interesting example because some of the soundbites are the same. Personally, I find the second one easier to understand. I find the music (especially the rhythmic base of it) competes with the dialog in the shorter trailer (the first one). The music in the second is mixed lower and is less busy of a track.

There’s no wrong answer, though… If you found the opposite or can understand both equally it demonstrates how intelligibility is subjective!

There’s only so much a mixer can do to make something intelligible. The same could be said of an editor – you have to do the best you can given the mics, actors, accents, noises, and whatever other challenges come at you.

Dialog intelligibility is something to keep in mind as you’re starting to edit. The solution to a problem may not always be mic that has the best quality. It might be the mic where you can understand the dialog the best.

Leave a Comment