Editing Digital Audio

Bill Davis
Special Issue 2000

Gather 'round everyone, it's time to reveal the secret ingredient that all the pros agree must be added to each of your video recipes if you want to cook up an outstanding program. Ready? The secret ingredient is... great audio!
That's right. To be good at making videos, you must learn to be good at editing audio. If you doubt the truth of the above statement, try this simple exercise. Pop a tape you haven't yet watched into your VCR - fast forward a bit into the heart of the program - then hit Play, turn your back on the set, and listen. You'll probably discover that you can still understand most of what's happening in the program. Now turn around and watch the screen, but turn the sound off.
Unless you're a talented lip reader, your comprehension of what's happening will fall off quicker than a chef's hat in a hurricane! The point of this little exercise is to underscore just how important audio is to the success of any video and to also demonstrate why most of the new digital video editing programs also sport some very tasty audio editing features!

In the old days, sound editors had to rely exclusively on their ears. Today, nearly every digital editing system has the capability of producing a visual representation of what your audio tracks look like. This is a wonderful new ingredient! In Figure 1 you can see an audio waveform as it might be displayed in the timeline of a typical video editing interface.
This particular waveform represents the sound of a voice saying the phrase, "Hi, my name is Bill" (Figure 1A). The peaks and valleys representing the groups of words and spaces are easy to see, and in most editing programs, the process for doing basic audio editing on this kind of dialog is simple and sweet. Using the razor blade tool (1B), you would just click on the audio track in the space right before the word "Bill" (1C). That cut would separate the track into two parts. Making another cut right after the word "Bill" would isolate the word. Selecting and deleting the edited section would remove the word and leave silence in its place (1D).
However, silence is not always a good thing. Play the track and unless you have background music or other sounds to cover the gap, you'd probably notice that the sudden absence of background noise makes for a pretty sour edit. Background noise? Absolutely.
Unless the audio you're editing has been recorded in a super-quiet environment or in a special audio recording booth, it will include ambient background sounds. In audio jargon this is often called "room tone." To avoid sudden silence in an audio track where you've deleted a sound, you need to copy the room tone from another part of your recording and paste it into the gap.
Figure 2 shows another audio waveform that represents the same voice saying the simple phrase, "Tom is not here" (2A). If we cut the word "Tom" (2B) from example 2 and instead of using room tone, paste it into the gap we created in example 1, we instantly create the new phrase "Hi, my name is Tom" (2C). This is digital audio editing at its most basic. This sort of cutting and splicing works great on an audio-only file. But when you're working with files that have both audio and video content, things get a bit more complicated. If you don't separate your audio and video first, even a simple cut like this will cause a major glitch in the video part of your program.
The solution is to "unlink" the video from your audio before you do the edit. The ability to "unlink" your audio and video is a powerful capability, but be careful - it also has the potential to mess up your program in a hurry! That's because if you "unlink" a clip and then paste some new audio into the audio tracks without ALSO adjusting the matching video track, you can throw your program "out of sync" and leave it looking and sounding like a badly dubbed foreign film! This is NOT a recipe for success!

In the real world of "audio for video", things are rarely as simple as our first two examples. For example, when working with video files, replacing the name "Bill" with the sound of the word "Tom" on the audio track causes a problem if we can clearly see Bill speaking on our video - his lip movements won't match what we hear. This is where "B-roll" footage in the form of a video cutaway can save the day by covering up the mismatch between the new audio and the old video.
Look once again at audio sample 2. Do you notice something odd about the script? Here's a hint. How many people do you know who would actually say something as stiff as "Tom -is -not -here." In the real world, we'd be more likely to say "Tom's not here." Use that phrase in the edit we practiced above and the result would be "Hi, my name is Tom's." Now that's going to leave a bad taste with your audience no matter what's showing on the screen!
You might attempt to zoom in to your audio waveform and make a stab at cutting the "z" sound off the end of "Tom's." Sometimes this kind of edit works, often it doesn't. It all depends on how the script was read, and how consistently it was recorded. Which is one of many good reasons to practice consistent audio field recording techniques.
The more uniform all your dialog recordings sound, the better the chances you'll be able to successfully cut and paste things later to repair a copy problem or fix a dialog flub.

• Digg This!
• del.icio.us
• Technorati
• StumbleUpon
• Reddit