Mixing audio in a separate location or using different software from the video edit is called sweetening, regardless of how sophisticated or simple the process is. As more video is edited on non-linear, computer-based editors, audio editing is being integrated into video editing. FCP 5 includes Soundtrack, a sophisticated audio editing program that serves as a "plugin" for FCP, importing and exporting video files and their associated audio.
In linear editing you can use a video dub to combine audio from 3 tracks (two on dub, one on master) and mix down to the other track on the master tape. The two video tapes must be in sync for this to work, however.
We usually use the two tracks available in DVCPro 25 for a voice track (Track 1) and a music and effects (M&E) track recorded on Track 2. Because video/television is now frequently stereo, the 4 channels available on most broadcast formats (including our new DVCPro 50, P2 camcorders) allow for left and right voice tracks and left and right M&E tracks.
The first step in any type of audio postproduction is to complete the video edit. Audio is recorded "dry," which means at full level without fades, equalization, or other processing. The edit master is brought to the sweetening facility or imported into the sweetening software. This process is called the "layover."
The synchronous audio will be combined with music, pre-recorded sound effects, Foley effects (SFX made to sync with the video) and any other desired elements. Equalization will be used to correct any problems with the original audio quality, and various types of processing may be employed to achieve the desired sound.
Once all the elements are assembled on the multi-track audio master, they are mixed down to two or four tracks as they are recorded back onto the video edit master. This process is called the "layback." The stereo matrix is applied at this time if the final product is to be in stereo. Music is usually just panned to extreme left and right to preserve the matrix established by the music producer. All other elements are panned appropriately to place them in proper perspective with the video. Note that for most video projects, voice and effects should not be panned very far apart, as this will lead to a "hole" in the sound from a typical stereo TV set.
Audio mixing must be done by ear. There is no meter that will tell you your audio sounds "right." The sweetening process keeps the individual elements separate, facilitating changes, but the process is much the same if done directly during the on-line session.
While we have an extensive sound effects library, there are limits to pre-recorded SFX. Foley effects are sound effects recorded while watching the video so that they sync. Foley only requires a microphone, a VCR, and some imagination. Often Foley effects can be mixed with canned effects to achieve much more convincing audio "events" than is possible with pre-recorded effects alone.
Lip Syncing.
Lip syncing is easier than it seems, and is best accomplished with a CD player at the shoot. CDs run at a regulated speed, so the CD should match perfectly when played in your editing bay.
On-line Video Postproduction.
An accurate log of all raw footage is critical before beginning an on-line editing session in a linear A/B roll editing suite. This is usually done using a window dub (timecode shown in a visible window over footage) on either VHS tape or DVDs. For projects with large amounts of raw footage, a searchable database or spreadsheet program and laptop computer may be helpful.
The "magic" of A/B roll is all done in the switcher. The editor is just a recorder, taping whatever comes out of the switcher. For this reason it is essential that the input on the editor be set to Analog and Composite (CMPST) for all A/B editing sessions. If the editor appears to perform cuts instead of dissolves or other effects, check this setting.
Tech standards are very important for tapes being sent to a broadcast facility or dubbing house. I prefer a 20 second slate, and DVCPro tapes probably only require 30 seconds of bars and tone. We usually use 30 seconds of bars and tone, 20 seconds of slate and 10 seconds of black (for a total of one minute). Dubbing houses prefer black to countdowns and broadcasters no longer need them, so they are disappearing from the industry.
Using a 1,000 Hz tone, audio levels on the record machine should be set at -20 db corresponding to 0 VU on the mixer before each session. Then maintain accurate levels for playback machines, making sure that the EXT. MON. switch is in the up position.
A/B roll editing employs proc. amps for each source machine. Begin your session with these proc. amps in the preset position (knobs pushed in) so you can see your video as it was shot. Then, if necessary, use the proc. amps to improve image quality.
Sync roll editing is a quick way to switch a show that was shot with multiple VCRs jam synced to a single time code source. If we had 3 source machines, 3 raw footage tapes could be synced and rolled simultaneously. Then the editor could "edit" the show manually using the switcher. Soap operas are sometimes edited this way when time is short.
FCP 5 has added a feature that supports multi-camera editing. While it currently has the drawback of requiring the footage to be digitized (unnecessary when working with tape in an on-line suite), this is not necessary with P2 cards or video acquired directly onto hard drives, so non-linear editors may soon replace tape editors for "sync roll editing."
Inserting new audio or video in a previously edited program requires some trickery in an on-line suite to match the new footage to the old. A 10 frame audio dissolve rather than a cut may be used to soften an audio transition.
Graphics are combined with video and recorded on another tape. We try to use original footage to minimize generational loss. Digital formats minimize this problem, but do not eliminate it entirely.
In linear editing we usually develop an "edit master" with different audio on the 2 channels. Never send this master to a TV station or dubbing house. The odds of getting the audio mixed correctly is less than one in three. Make a regular dub, mixing the audio the way you want it to be heard. Sometimes additional graphic enhancements such as keyed tags are added as this dub is being made. The result is a "release master," representing exactly what you want the ultimate viewer to see and hear.
In this class all assignments must be submitted on DVD, which becomes the "release copy." Only the DVD will be graded, as your video isn't finished until it is in the form that will reach the ultimate consumer.
You are almost done, but be sure to label your masters and copies clearly before you let them out of your hands. Then it's party time!
