You’re watching your favorite action flick. The hero has rescued the hostage and they’re blasting through rush-hour traffic on a motorcycle. Close behind them, the bad guys follow in a hot-rodded SUV, shooting hundreds of rounds of ammunition. Cars flip and explode, engines race and yet, you can hear every word of dialog. How is that possible? It’s not, recording audio for video isn’t always a smooth process. Of course, they probably recorded dialog during the shoot, but all the external noise made it useless in the final edit. So the studio used a process called automated dialogue replacement, or ADR, to replace the dialog. Sounds expensive, doesn’t it? What if I told you that your next production could use ADR to create a clean dime.
ADR is everywhere
ADR has many names. Some use it as an acronym for automated dialogue replacement, while others prefer Additional Dialog Recording. Still, others call it dubbing or looping. Regardless, it is the process of re-recording replacement dialog for a production.
ADR is used for many reasons. The most obvious use is the replacement of noisy audio on the day of the shoot or situations where dialog recording simply wasn’t practical. ADR techniques can eliminate profanity for a television audience, or completely replace voices for use in another country. In fact, if you watch TV or movies at all, you hear ADR all the time. Feature-length movies, TV dramas, reality programming and even animation and video games all use ADR at one time or another.
Here’s a typical workflow: after a rough cut is established, the talent comes to a special studio called a dubbing stage. The producer or director shows them a clip of the scene and the actor rehearses their lines along with the clip. The goal is to recreate the emotion, setting and intensity of the original shoot — or possibly improve on it. An engineer hits the record button and, as the scene plays on a large screen. The actors hear their original take through headphones as they deliver the same lines. When everyone is happy with the take, the process repeats for all of the required replacements.
Every actor in the project may visit the dubbing stage before completing the process. Once the producer signs off on the final recordings, the editor or engineer edits the best takes to make them fit the visuals. It’s a long, drawn-out process that often takes months to complete. Why do they go to all that trouble? In a word, control. By recording dialog in a controlled studio setup, the producer and director get exactly what they wanted in the first place, plus they have full control over volume and clarity in the final mix. A final bonus is the ability to remove the voice and replace it with an actor in another language.
You can ADR too
Believe it or not, you really can do this at home with a short equipment list. The first item on the list is multitrack audio recording software – often called a DAW or digital audio workstation. Don’t let the fancy name scare you, this is any audio editor like Sound Forge, Audition, Reaper or even Audacity. Next, you need a decent microphone. This can be a shotgun, studio condenser mic or even a lapel mic. Just make sure you have a windscreen to minimize breath pops. Finally, you need a way to plug the mic into your computer. This may be as easy as plugging directly into your sound card with a couple of adapters. Alternatively, a simple 2-channel USB or Firewire audio interface offers more control and options. The basic techniques for this job are the same as any voice-over recording. See 10 Voice-Over Tips for a more comprehensive outline of the procedure.
Get everything hooked up and working, then identify the clip or clips that need dialog work. Most of the major audio applications allow importing video clips as an audio source, so do that if possible. If not, export the audio from your clips and bring that into the audio editor on its own track. Now, establish a new track for recording and play the clip through several times for the talent. Once they’re confident with their performance, record a few takes and listen to the results. The ability to see the video clip is a bonus for the talent, but watching the existing waveform on the timeline offers excellent visual cues too. Once you’ve got something everyone is happy with, it’s time to clean things up a bit.
If you are truly fortunate, your talent will nail the performance perfectly and you won’t have to edit. In the real world, you’ll probably have to repair a few things. Using the original dialog as a visual guide, line up the first phrase or word. As you look further down the time-line, some phrases will match timing and others won’t. To repair the timing, split the phrases in the quiet gaps and slide each one around until it lines up with the guide vocal. If you recorded in a quiet environment, the only sound on the recording is the talent. This leaves you some wiggle room to cut up the track and move things around until the timing matches through the entire clip.
This is like any other edit job and there is nothing sacred about using everything from one clip. If necessary, feel free to bring in words and phrases from all of your takes. You may find that one little change in inflection or delivery really sells the edit, so use everything at your disposal. Sometimes, the visual reference isn’t enough, so play your edit alongside the original take to verify the pace and timing. The talent may slur some words and run others together, so listen closely to make sure everything flows well. When you’re happy with the edit, export the audio as a WAV file and drop it back into your NLE. Slide the new track around until everything matches, then mute the original audio track. If all went well, you will have a clean, clear, perfectly synchronized dialog track.
Not all automated dialogue replacement sessions involve full dialog replacement. Sometimes, you’re blending existing dialog with bits and pieces from the studio. In this situation, do everything possible to replicate the audio from the original shoot. Use the same microphone at the same distance and angle if possible. This will make the edit transitions smoother and maintain the illusion. Finally, your ADR dialog will be somewhat dry and sterile. Make sure to record some room tone or other location audio to lay underneath the new dialog track. It will cover the fact that the audio was recorded separately and offers another layer of control in mixing for stereo or surround.
Give automated dialogue replacement a try. It’s a little tedious, but with practice, you can patch the dialog in your productions just like the big studios, at a fraction of the cost.