Premiere users are lucky. Its plug-in format has become a standard for professional audio applications (including Macromedia Deck and BIAS Peak), so third-party manufacturers offer sophisticated sound design tools that integrate perfectly with the video editing program. Depending on the size of your budget and hard disk, you can add a full array of audio plug-ins from Waves, BIAS, Opcode, Invision, and other companies ... giving your copy of Premiere the power of a recording studio.
But if you don't want to invest in third-party extensions, you can still get some first-class effects by using the basic audio functions that come with every copy of Premiere. All it takes is a little time, and a willingness to experiment.
While plug-ins allow more precise control and deeper effects, you can accomplish a basic flanging sound with nothing more than Premieres clip speed control and a couple of audio tracks. Compare example audio1 with its flanged version, audio2. We chose the steady sound of an airplane interior so the effect would be obvious, but sound designers often flange dynamic effects such as car passbys to give them even more motion.
The flanging you hear on audio2 is merely the original clip playing on two simultaneous tracks, with one tracking having its in point moved slightly to delay it, and its speed changed slightly from 100%. These changes have to be smaller than what you might expect -- about half a frame delay, at one-half percent speed -- and that takes a little trickery to accomplish:
The same voice clip was copied to three audio tracks, but the in-point was trimmed differently on each. Since we wanted more drama than a simple flange, we used larger speed changes: each clip was set 3% slower than the one above it. (A 3% speed change is roughly one-half-step musically, about the largest variation you can do to a voice without making it sound strange.) This change meant that track X1 took significantly longer than track A. To keep things in sync, we used the razor blade tool to split X1 and remove two frames. The entire effect was mixed down to a single movie, which we then imported to two tracks, flanged, and mixed again.
You can use this effect, without the flanging pass, when you want to thicken crowd scenes. Put the crowd on a few simultaneous tracks, and change the speed on each so they sound like different groups of people. Varying the in-points, and making tiny cuts within the slower tracks, keeps them in unison. It's a lot cheaper than hiring more actors.
The Fill filters direct a clip to just one speaker. (You can also use the Pan filter, introduced with Premiere 4.2.) The speed changes cause each ear to hear a sound thats different but related, and the brain tries to sort it out by creating a stereo image. It's not the true stereo of a music recording, but it sure sounds big. Since the changes are fairly large, mono listeners hear a thickened version rather than a flanging effect.This compatibility is important in many multimedia and broadcast projects, where you cant predict how the audiences speakers are set up. Its a good idea to check stereo effects by previewing them twice: open the Preview Options window, and listen once in a stereo format and once in a mono one. If you don't hear a stereo effect either time, make sure Use Filters is also checked.
Since the clips are playing at different speeds, they end at different times. We added the fades shown at the end of tracks B and XI, to keep the crash in unison.
The stereo movement depends on Premiere 4.2s Pan filter and its ability to change over time. (Pan doesnt have the same Start and End sliders as filter like Camera Blur, but it does respond to the Vary Settings command in the Filters dialog.) The same original sound is on audio tracks A, B, and X1. Track A's clip starts with a pan setting fully to the left and ends 50% to the right. Track B's laser starts 50% left and moves fully right. We then changed the speeds radically, for a big stereo laser that moved from the left to the right. Fantasy sounds, unlike human voices, can take these larger speed variations since the audience isnt used to the sound's natural timbre.
But it wasn't big enough. Thats why we added a third copy of the original clip on Audio X1, playing at a third of the original speed. This lowered the pitch about an octave and a half. We faded both Audio B and X1 to match the faster Audio A, and previewed again.
It still wasn't big enough, so we grabbed all three clips with the Block Select tool and turn them into a virtual clip, which we moved down a little later on Audio A. Then we put a copy of the virtual clip on Audio B, at only 50% of the original speed. Both virtual clips -- or a total of six versions of the original laser -- combine to make a sound that stretches across almost four octaves. Some reverb and compression would have made it even bigger, if wed wanted to use third-party plug-ins.