Sound Advice: Editing Narration

Eons ago (that’s about five years or so in computer years) audio engineers recorded and edited narration on tape. It was tedious, caused eyestrain and often required the engineers to twist their bodies into unnatural positions over and over to make the repetitious edits. Worse, the edits were permanent. If you made a mistake, your only recourse was to re-record the narration from scratch. Fortunately, times have changed for the better. Computer-based audio editing has completely transformed the process. Edits are done, undone and redone with just a few clicks of a mouse, and now, the only potential injury is a sore wrist.

With all this newfound power and flexibility, video producers can easily try multiple variations of a given take, or rearrange the script after the voiceover talent has gone for the evening. In this article, we’ll teach you how to make the most of the digital audio tools at your disposal.

File Management

Much like video editing, your audio editing session begins with hardware and software choices. Fortunately, any computer you use to edit video has more than enough power to perform your audio tasks. In fact, virtually any off-the-shelf computer sold today contains more than enough resources to complete your creation. There are several options for recording software – many even free – but a good editing package offers a clean interface and useful editing tools. You’ll spend quite a bit of time with this software, so make sure it fits the way you want to work.

Next, you have to make some decisions about how you will manage the audio files you’ll create. There are two schools of thought: record it all into one big file for easy management or record the project in multiple, smaller files for a list-based approach. Either method is justifiable, but the single, big file approach offers some extra benefits. You’ll have only one file to deal with and won’t need a filing "system." The waveform display on your computer makes it easy to visually identify various takes and extraneous noises between takes. The single file tactic makes for easy application of consistent signal processing. It’s also easier to calculate the total finished length of your project.

Editing the Dialogue

The first editing task is a pass through the entire recording to select the keeper takes and eliminate the unwanted material. The safest way to accomplish this is to open a new, blank file in addition to your recorded narration. As you find the segments you want to keep, copy and then paste them into this empty file. When you’ve finished with this pass, you’ll have two files: the unmodified original and a rough edit that contains only the parts you want to save. Close the original file, but keep it handy just in case you need a replacement segment later (see the Destructive Editing sidebar for an alternative).

Your second pass will focus on the new rough-edit file. As you play through the narration, listen for stray noises such as page turns, coughs, sniffles and large gasps of breath. In most cases, you’ll find that these noises occur separately from the narration, so they are easy to edit out. As you find them, simply razor out the offending sounds and delete them from the file. After each new edit, listen to the narration in paragraph-sized sections, comparing it to the script to make sure you haven’t inadvertently eliminated any important dialog.

During your final pass through the recording, listen for pacing and timing of the various segments. Make sure that each sentence sounds natural, with proper gaps between phrases. If you’ve tightened a bit too much, place the cursor where you need a gap and insert some digital silence: usually a quarter-second or so at a time until it sounds right. When you’re happy with the edit, check the time of the complete project or file. It should be slightly shorter than your video project. If it’s longer, you have a problem. This is a good time to tighten the gaps between sentences and scour the script for sections that you can eliminate without harming the message.

Adjusting Levels

As you listen through the narration you may notice volume differences between the various segments. The first step to even out these differences is to normalize the entire wave form. This process brings the volume of the entire audio segment up without causing clipping. A value over 95% is ideal. Afterward, find those segments that sound lower in volume, highlight them and normalize that section independently. By isolating sections to normalize from louder sections, you can increase the volume without causing problems with the louder sections.

Alternatively, you may find sections of the dialog that sound too loud in comparison to the rest of the recording. You can lower their volume with the normalize process as well: just use a lower value. Continue through the entire recording until you have a consistent volume level from section to section.

Dynamics

The waveform display in your editing software will show a series of spikes followed by areas with a lower volume. This occurs naturally when recording the human voice but, in your finished production, the narration may have to compete with background music, ambient sounds and even sound effects. It’s time to get a handle on the dynamic range of your recording. Dynamic range is the difference between the loudest and softest parts of your recording. Reducing the dynamic range (and then raising the volume overall) makes your narration sound louder and helps it cut through the other audio elements in your video.

The tool you’ll use to tame the dynamic range is a compressor/limiter. This may be a plug-in or simply an option in your software, but most work in a similar manner. The Threshold setting determines how loud the sound must be before the signal processing kicks in.

A setting of -20dB is a good starting place for your compressor. The Ratio adjustment establishes how much compression will occur. Start with a gentle setting of 1.3:1 – use a higher setting if the volume contains a lot of variation. A ratio of 10:1 or higher turns your compressor into a limiter. This requires a higher threshold setting of -3db or so and effectively tames those spikes in your audio. When you’re happy with the results, save the file and get ready for a reality check.

Final Touches

Now is the moment of truth when you combine your carefully edited narration with your video elements. Import the file into your video project and place it on it’s own audio track. Using the Razor tool, split the audio track into segments that fit the flow of your video. Now you can easily slide the audio to the exact time it’s needed in the video. Make sure to lower the volume of any music or other audio while the voiceover plays. You can easily do this with a combination of rubber bands and handles on the various audio tracks. Continue to refine the blend until you’re happy with the finished sound, but don’t forget to check the finished soundtrack on a TV. This is more in line with the way your audience will likely hear your production.

Digital audio editing is very similar to digital video editing. The steps are the same – pre, production and post – and the entire process happens inside your computer. Now every video producer can use these powerful tools to produce and edit narration just like the big production houses.

[Sidebar: Destructive Editing]

When you edit video on a computer, you are performing a “non-destructive” edit. This means that you aren’t changing the original files when you cut and trim, but are instead changing a project file, which you will then render to a new video file. Strangely, audio editing software has been slow in implementing non-destructive editing, so that when you cut a file and hit save, the original file is permanently changed. While non-destructive audio editing is becoming more common in audio tools, it isn’t ubiquitous. Look for that feature before you buy.

[Sidebar: Which Came First?]

Back in the days of tape-based editing, it was common practice to record the narration before editing the video. The completed audio track was laid down on the master videotape and the video elements were inserted to fit. Today, you’re not limited in that way and it’s just as easy, and perhaps better, to do a rough edit of your video before recording and editing the narration. This allows you to make script tweaks and wholesale edits before recording your voiceover, saving time and possibly recording costs.

2 COMMENTS

  1. With regard to the final paragraph of this article, and with all due respect, the narration IS the story of a video. To tell as detailed a story with pictures alone, there would have to be an awful lot of pictures, and title cards would probably also be required.

    As it IS the story, the narration also provides the pacing of a video. The reason narration was recorded first and the video was edited to it was because it just made complete sense.

    The way this worked in the “old days” was that a storyboard was created along with the script so the story could be roughly visualized and, thus, how much video needed to be shot (and we always worked smart by shooting extra footage). The producer would also use the storyboard to show the client what was planned. The subject matter of the story alone determined the speed at which the narration would be spoken and, using the storyboard and the script, we would decide if there were any areas where pauses were required for any PLANNED screen action.

    While the narration track was being recorded, video footage was being reviewed and selected. When the producer and the client were happy with how it sounded, the FINISHED narration track was transferred to the video master, and the footage was edited to the pacing provided by it. Video: done.

    The downside to using videotape was, should something need to be added to or removed from it after completion, it would require a lot of work (aka time and money) to reassemble the video. And everyone involved knew that.

    So, when technology brought us desktop computers and this marvelous thing called non-linear editing, everyone celebrated knowing the inserting or removing of video and/or audio elements was so much easier because we can just drag things around like puzzle pieces.

    Thus, with this truly awesome technology, it amazes me (because it makes no sense whatsoever) that a video would be edited first and then the narration forced to squeeze-fit into sequences not nearly long enough for the words. This results in a video that sounds absolutely awful because the story being told… the very reason for the video… sounds rushed, as if the narrator had to use the bathroom.

    Professional narration is storytelling. It is NOT just reading words, which almost anyone can do. Storytelling requires giving words their proper weight (emphasis), and sometimes, momentary pauses are required to allow the audience to absorb what it being said. When a narrator is forced to fit copy within too-tight spaces, any attempt at giving words their proper weight is brutally killed, and the result is so rushed that the audience can’t absorb anything being said (or remember it later).

    It is beyond me why would anyone want to produce a video where the narration is so obviously rushed that audience members remember THAT instead of the STORY. Technology has given us the wonderful gift of non-linear editing and, for reasons I cannot comprehend, some insist on finalizing five pounds of video and then forcing 10 pounds of narration into it. Total insanity.

    No, it is NOT better to do a rough cut of the video before recording and editing the narration; doing so is a waste of time. Telling the story PROPERLY, by having the narration recorded and approved FIRST, and then editing the video to that FOUNDATION, is the only way to save time and costs… because it is working smart and sensibly.

LEAVE A REPLY

Please enter your comment!
Please enter your name here