FIG. 1: Add polish to the mix by splitting the breaths and music from the dialog and placing them on their own tracks for individual processing.
Screencasting programs such as ScreenFlow for the Mac and Camtasia for Windows let you simultaneously capture screen movement and audio. Here's how to use your DAW to edit and polish your screencast's audio track. I'll cite Digidesign Pro Tools 8 in my examples, but most DAWs offer similar techniques (see “Step-by-Step Instructions” below).
Separated at Birth
It's often difficult to use a pop filter when creating screencasts. It can block the view of your monitor and obscure any printed text you're using. A windscreen is less visually obtrusive, but it will provide only limited protection from plosives. Fortunately, you can correct those and other inherent problems after the fact.
The first step is to temporarily export your captured material and import it in your favorite DAW. Low-resolution QuickTime output will do fine visually, but make sure to export the audio at full fidelity. In your DAW, extract the audio to its own track and, below it, create four extra audio tracks for your edits (see Fig. 1).
Because screencasts primarily contain dialog, the imported audio will serve as your main dialog track. Insert a low-cut filter with a moderate slope and set it to about 100 Hz. This should catch most of the plosives that have sneaked past your windscreen. Nasty peaks will most likely have to be dealt with at the waveform level or with fades.
Reducing the level of breaths on the main dialog track greatly enhances the overall quality of the finished product. The goal is to reduce, not eliminate. Set the level of the first extra audio track to between -12 and -16 dB, depending on how heavy a breather you are. Use your DAW's selection tools to select and cut the breaths on the main track and then drag them to the track below. They will instantly be reduced by 12 to 16 dB.
It may be necessary to add slight fades before and after the cuts to smooth the transitions. Pro Tools' Separate Region function (Command + E on the Mac; Control + E on Windows), along with the Smart Tool, lets you switch between these tasks with minimal wasted movement. It's easy to unintentionally cut off the ends of words, but with experience you'll learn to visually recognize when consonants are fully completed and to establish your selections efficiently.
Trim the Fat
Tightening up dead spaces and eliminating hesitations or unintentional mouth noises in the original performance will improve the overall presentation. These are easy to select and delete, but you may need to tighten up the holes that are created. Because screencasts generally don't contain mouth movement on camera, slipping dialog out of sync by small amounts usually won't cause a problem. But when shifting any of the dialog in time, check it against the video to make sure it is consistent with the visuals.
Tightening up loose dialog in this fashion may lead to small but extremely conspicuous amounts of dead space. To fill this sonic vacuum, record some room tone with your mic set in its original position and the signal path settings exactly as they were. Create a recording of 30 seconds or so, place it on the second spare track and loop it for the duration of the session. Trimming a second or two from the head and tail of the room tone file will help avoid any glitches at the loop point.
Use that track as “sonic glue” where empty spaces occur on the main track.
Separate Tables, Please
Use the same select and cut techniques you used for the breaths to isolate any music or other non-dialog audio. Place this on the third additional track to allow for sonic manipulation to match the EQ and levels with the rest of the content.
Set the record path on the final spare track to match your original recording. Use that track for the dialog replacement that will inevitably be necessary. Match the record levels as closely as possible with your original recordings, and re-record or punch in bits that are not salvageable or that you want to perform better. You can move those fixes to the main dialog track, but treating them separately allows for more accurate sonic matching.
Once all the audio has been separated and edited, the next stage is sweetening. Use EQ, compression and limiting as necessary to even everything out for a balanced mix. Finally, bounce all tracks of edited audio to a stereo file and bring it into your screen-capture program. Mute the original audio track and output your final version.
Eli Krantzberg is a Montreal-based bandleader, musician and sound engineer. You can find his training DVDs atwww.groove3.com.
STEP-BY-STEP INSTRUCTIONS ON NEXT PAGE
Import the QuickTime video into your DAW. Place the audio on its own track with a low-cut filter set to about 100 Hz.
Set up additional tracks for breaths (lowered between -12 and -16 dB), room tone, music and dialog replacement.
Select and split the breaths and music contained in the main track onto their own tracks.
Add fades where the breaths are split from the main track and loop a short recording of room tone throughout the session.
After sweetening the tracks with EQ, compression and mix-bus limiting, bounce the final edit down to a stereo file.
Bring the edited audio track back into your screen-capture software, mute the original and output your final version.