Drums in the Rear Channels?
I was talking with John Atkinson on the phone about surround-sound audio formats, and he was telling me about a Manhattan presentation run in parallel with Home Entertainment 2001 in May. In remixing the Van Morrison catalog for DVD-Audio release, veteran producer Elliot Scheiner had demonstrated how in some songs, such as "It Stoned Me," he placed the drum tracks in the rear channels. Unfortunately, this little example of revisionist mixing might become the multichannel norm if record labels don't get a grip on how and why music was recorded in the first place. What's next? Phil Spector's classic mono mixes in seven-channel surround?
A couple of days after JA's call, David Chesky sent an e-mail suggesting that I should remix Alternesia, a stereo CD of mine released a while back by M•A Recordings, as a surround DVD-A or SACD. I had recorded Alternesia, with care, on a modified analog 2", 16-track MCI tape machine, and it has always been ripe for remixing and remastering to a higher-resolution format. Although the album consists almost entirely of analog recordings of real instruments, its "ambience" is not based in any real-world acoustic space, and so would easily lend itself to a trippy surround mix.
Or so I thought.
After doing a mental exercise on how Alternesia could be reworked, I quickly came to the realization that, whether consciously or subconsciously, the knowledge that I was writing for a two-channel medium influenced the originally compositions and how they were written and arranged, let alone mixed. A particular density in the mix was important; had I known I was writing for a multichannel medium, I would have written and recorded different parts from the start. For the surround version, in addition to the obvious mixing changes, I would have wanted to rewrite some of the parts and record quite a bit of new stuff.
One example: The first track, "Premonition," opens with some random tuned metal sounds that are slowed down a tad and arranged to fill the stereo soundstage from left to right. I recorded three parts (mixed hard left, center, and hard right) and added enough out-of-phase ambience to throw the image as deep into the room as possible. For this to work in surround, I would need to add enough new parts to encircle the listener with the same apparent density of sound, but this time from the rear and sides as well.
The percussion parts on most of Alternesia were built up until the complexity of the two-channel mix felt correct—many parts never made it to the final CD, or were simplified. Merely throwing the existing parts around the room would disturb this balance. But adding parts to the rear—doubling some parts, writing new ones—could move the soundscape into the room without weakening the overall blend. The Alternesia that would likely emerge from the multichannel process would be an entirely new work.
That got me to thinking about other albums I'd recorded, and some of my favorite discs from other artists. I'll go out on a limb and suggest that a sizable chunk, if not most, of popular-music recordings created in studios in the last 40 years, having been written and arranged specifically for one or two channels, would probably have come out musically different—not just spread to more channels—had the artists, arrangers, and producers known that their works would be first released in a surround format.
Classical music, on the other hand, as well as most jazz, blues, and bluegrass, might be a completely different story. David Chesky says that he writes his symphonic works for the stage and the audience, and postpones the stereo and/or surround decisions until the recording is finally made—for him, the "medium" is the live event, and any resultant recording is the "document." Whereas, for the late-20th-century pop/experimental artist, the studio is the performance space, and the resultant mix is the "event" and the document that reveals the final work.
I've heard plenty of record execs claim that improved sound quality alone is not enough to justify the introduction of new high-resolution media. Taking parts originally laid down on multitrack tapes mixed to mono or two-channel stereo and dispersing them among four or more channels might call attention to DVD-A's or SACD's differences from those stodgy old CDs, but it could, in the process, corrupt the magic of the original release. I'll bet even a group like Pink Floyd, an obvious candidate for surround-sound revisionism (assuming they could agree to cooperate long enough to get it done), would have trouble spreading out their "classic" stereo mixes to four or more channels without the urge to tinker more than just a bit with the actual music and arrangements.
There are exceptions, of course; Mike Oldfield's Tubular Bells was originally recorded and mixed as both a two- and a four-channel work, and it succeeds nicely as a Virgin SACD surround disc. Warners is also to be commended for the DVD-A treatment of the Doors' classic "Riders On the Storm," in which the music remains locked up front while ambient rain and thunder fill the entire room. And I can easily imagine a surround recording of Symphonie Fantastique in which the orchestral bells' echo part in the final movement is heard from offstage, as Berlioz intended.
But why muck up the vast majority of recorded mono/stereo popular musical history with witless second-guessing? Note to record labels: Unless artists are involved in a fundamental way in reworking their albums to their satisfaction, and unless the results are subsequently labeled and marketed as "rewrites," leave the two-channel stuff alone. If you want to prove that surround sound is the next big thing, commission new works written specifically for four or more channels.
Even then, it might take a while to get surround right, and not all forms of music will benefit. Not until recording artists sit down to a blank hard drive with visions of surround sound dancing in their heads will we discover just how much the medium can enhance the message.