(Note: I originally posted this on my website in 2004. A few folks have asked about it since then, so with a few tweaks and updates for 2012, here again is my take on why we record with more than just two microphones.)
Recording Classical Music: Microphones and Multi-tracks
Some would say the only true way to record classical music is just one pair of microphones. It’s a great idea, and at first glance, it makes perfect sense: use two high quality microphones, (preferably omni capsules) spaced the same distance apart as the human ears, find the best seat in the house, and voila – A perfect stereo recording!
Well, yes and no. It’s not quite that simple. Keep an open mind, and read on…..
Unlike the experimental nature of some jazz, rock and pop recordings, classical music requires a different approach. It is not an overdubbed, highly processed sound like some other genres. Any experienced engineer who works in any of these styles will tell you the “Classical” approach is different right at the start of the process in that musicians will always prefer to get it right in the first place: onstage, as an ensemble. Be it a concert or recording session, the performance is not some producer’s computer-sequenced dream or built from the ground up with a drum-machine click track. Frankly, the conductor/music director should have as much or more control of the sound than anyone else in the entire process. Very often, it is incumbent on the recording engineer and producer NOT to ruin a perfectly good performance with overproduction or any heavy handed processing.
More than any other listening experience, classical music still reigns supreme in that it is a highly focused, detailed experience for its audience, whether heard live or on a recording. The classical audience comes to expect perfection, as well as a quiet, calm comfortable listening environment. It is that very environment (and performance discipline) that dictates this different approach than all other music recordings.
Today, the line continues to blur between the various approaches to recording all genres of music, including classical. As the tools get better, those who use them acquire greater skills in editing and mastering, while still allowing unwavering faithfulness to the original performance. (To those whom more has been given….) Nevertheless, the days of getting it all in perfect one take, as commendable as it might be, are dwindling (even Toscanini and Ormandy edited their recordings) and it’s now become quite acceptable for even classical musicians to build the perfect work of art, with all the mistakes removed; the performance as flawless as the score. Granted, miraculous, error-free performances still happen all the time in classical music, but it is ever more desirable to create the perfect masterpiece with these new tools, well within the discipline of classical music.
But that’s getting a little bit away from the topic here: Single-point stereo microphone recording of classical music, or multiple microphone/track use.
It has always been a long-held ideal to find the perfect “sweet” spot in the audience, in order to best experience and/or record a live performance. This makes perfect sense, just as it applies to watching a movie in a theater. Find the middle of the middle section, about one-third of the way back in the audience, near the center aisle. (Hint: All big budget movies are mixed this way, in the same spot, in mini-theater/studio control rooms out in Hollywood.)
In a perfect world, and a perfect acoustic space, this would work perfectly.
But just as with movies, operas, and plays, there is another sensory input coming into play during a live concert, and it is missing when just listening to a recording after-the-fact. It’s the visual impact that glosses over (and forgives) so many imperfections going on around the listener in a hall, church or even home listening space.
Here’s a way to understand some of the strength of the visual’s impact: Turn on your favorite Cable Channel (PBS, Palladia, Bravo, etc.) broadcast of a good music performance on your television. Record it on your DVR, and watch it – enjoy it all as it plays out. (No tricks here – just enjoy.) Now, take that same recording and turn off the picture, plug in headphones or play just the audio on your hi-fi system. (Even better: play it elsewhere from you television-viewing environment. If your home theater or viewing area is like mine, it’s a completely different space and experience than my “audio” listening area. Perhaps you even have a “music room” per se.)
When you play just the audio from the DVR or DVD, you’ll probably be shocked at what you thought was a great recording. Without going into a whole sidebar on the bad things that happen to good audio in television broadcasting, cable distribution, etc., what you’ll probably notice is a less than perfect stereo image, compressed (and sometimes hissy or lumpy) audio, and even occasional distortion on the peaks. Why, you may ask, didn’t you notice that when it was on TV!?!?
The answer, of course, is that the visual component is so strong, it lessens the impact of the audio. The viewer simply doesn’t notice all the imperfections, at least at first. Of course, many audio professionals and musicians are trained to sort through this distraction anyway, but the perception remains, and this partially explains why sound-for-TV has been allowed to be so bad for so long. A great-looking television show overcomes a badly recorded audio production anytime. (When was the last time you heard someone complain about the sound quality of the CNN feed from the middle of a hostage crisis or helicopter rescue?) Next time you watch TV, close your eyes for a while, and notice what you’re really hearing.
Now let’s take the same concept and plug it into a “live” concert experience, where there are usually no HD video cameras or glossy post-editing. Even without that: YOU – the listener – are seated at a concert, let’s say the best seat in the house. All around you, there are still distractions, both visually and acoustically, some you may or may not notice…. The hall may be brand new or wonderfully ancient; magnificent to look at, or maybe in disrepair, plaster chipped and falling on the seat next to you. If it’s a big event, the audience may be in formal wear, or in attire appropriate for on the weather outside…. umbrellas, trench coats, even hats, scarves, boots, etc. You also have a program booklet to read, and a cell phone to turn off. Distractions are everywhere!
Once things have settled down and the audience is seated and relatively quiet, you’ll find many other distractions lurking along the way: If it’s an old church on a steam heating system, the pipes may be banging long like some old ghost is trapped in the basement. If it’s got a conventional heater or air-conditioning system in place, there may be a constant hum or whine that’s on even before you arrive. The light dimmers may buzz, as well. Very often, the noise is so constant that it sinks into the background, unconsciously accepted by the audience because it’s been there the whole time. (In many halls where we record in regularly, we’ve “sampled” these kinds of background sounds/interference and have them on file, ready to be digitally removed – “dialed-out” – of the final mixes.)
What many recording engineers know setting up ahead of time is that there can often be a bombardment of unacceptable sound going on (even during a performance) that the audience otherwise misses at the time, due to so many visual cues and other distractions. There are coughing and sneezing audience members, food wrappers, chairs creaking, and the afore-mentioned steam pipes to endure; the day-to-day distractions are legion. It is a very noisy world in which we live.
But, thanks to all of our sensory input (including whether we’re too hot, too cold, or just-right in our seats), many things get glossed over as we settle down into an otherwise enjoyable concert performance. They are all around us, but in most cases we unconsciously push them aside to let the music get into our brains. It’s selective input, and it simply shows how subjectively we humans process so many things at once.
A well placed, single-source stereo pair of omni-directional microphones knows no such selective/human filtering.
They are deadly accurate, and will record EXACTLY what they are “hearing,” warts and all. Aside from a closed session with no audience present this (less-than?) “perfect world” listening environment can have no other distractions or sonic input, other than what is coming from the stage area – along with any reverberant sound around the microphones in the hall.
Now, considering how direct sound drops off in intensity the further one goes from the source, it is entirely possible that a throat-lozenge wrapper ten feet away from these microphones can sound as loud (or even louder) than the triple-piano (ppp) solo violin or woodwind passage arriving at the same point in time from the stage thirty feet away. Certainly, the listener in the hall can grudgingly discern (or unconsciously ignore) the difference, but the recording has now captured it all: ambient sounds as well as the music from the stage, due to the microphones’ unwavering and brutal accuracy. Very often, the relative amplitude levels of desirable vs. undesirable sounds are now skewed way out of proportion, discerned and alleviated only by our visual cues.
One can see the violinist or tenor soloist, while the wrapper noises are invisible and therefore not necessarily “heard.” Our eyes give us some input, while the ears add to the rest of the experience. Live, it’s one thing; recorded, it’s quite another. If you don’t believe this is true, imagine if they only sold the first few rows of any given concert because anyone who couldn’t get a seat up front wouldn’t be interested in attending. Of course, we know this isn’t so, as long as folks can see something, they are often happy enough. The audio component completes the experience, with the brain having the final say on what is necessary and important. The rest gets filtered out as non-essential.
Ambient Sound levels all around the listener increase proportionally to the distance from the stage, but visuals often overcome the distraction.
Like the DVR/DVD experiment, this type of recording – 2 omni microphones in the audience – will have a dramatically different impact after-the-fact, in an audio-only listening test. With no visuals now to distract the engineer and producer, the candy-wrapper, the coughing patron, and the enthusiastic clappers suddenly loom quite large on the sonic landscape now, getting your attention as much as the music. Remember that in most cases, they are closer to the microphones than the music itself, and lopsided levels are the result.
Out at the microphone location, someone merely applauding at the end of a string quartet or vocal/piano duet sounds like a thunderclap compared to the actual music, but again, our senses accommodate for this in person. Not so with microphones and electronics with ruler flat response. Overall levels must be set to accommodate this dramatic disparity in sound levels, (often ten to twenty or more decibels apart) if only to avoid saturating the signal chain in the recording during the loudest passages (that usually being the applause, not the music!)
Thus very often the gain structure for optimal signal to noise ratio is seriously skewed, favoring the near-field applause vs. the more distant on-stage music. When the applause is edited (and subsequently brought down) the music must be brought up for optimum playback level….along with the noise floor, and any other gremlins in the room or the electrical/signal path that have crept in during the recording process. To address this problem any other way would require artificial gain processing like compression and/or limiting – totally unacceptable in Classical/Audiophile music recording.
To add one more log to the fire: Binaural, single-point recordings can only be truly experienced with headphones
Think about this one for a minute: If the microphones have already captured the best seat in the house with all the natural reverberation, ambience, room noise, (and coincident mouth-breathers and sneezers all around), then the only possible way to accurately put this sound into your brain is to generate the sound as close to, or literally next to your eardrums. Short of a cochlear implant, for accuracy’s sake, there can be no other acoustic distractions now. (The microphone placement has already done that for you!) The speakers in your listening environment, in essence, are now almost doubling the distance from your ears to where the microphones picked up the sound.
To put it another way, listening with speakers (instead of headphones) you’re potentially listening beyond the pickup pattern of the microphones, effectively increasing the distance from the actual performance: It’s a case of stage-through-space-to-microphones, then back out through more space to your ears via your living room speakers. Very often, the result is a more distant, less-detailed listening experience; almost listless, because it’s now been pushed through two sets of listening environments. Remember, there are no longer any visual imagines to fill in the blanks, and therefore, you now have one less sensory input at this point in the overall emotional experience. It’s twice the airspace, twice the room ambience, and no visuals to distract you.
What to do, then? What IS accuracy in classical recording?
The concept of recording in surround sound (5.1 and others) very smartly addresses many of these issues and more, but for now, we’ll stay with a stereo, 2-point recorded image.
For decades, we (as well as all of our contemporaries in the recording business) have recorded with a slightly modified version of this same “Best-seat-in-the-house” 2-microphone concept, (and of course the 3-microphone Decca-Tree technique) straight to analog tape or DAT, often simply due to budget limitations or the type of music at hand. Multi-track recording was simply not necessary, even overkill in some instances, and almost always beyond the budget of most orchestras and ensembles. Aside from the need for an occasional solo mic, this method still works amazingly well. While it can sometimes put the microphones dangerously close to (or in) the audience, it retains all the dynamics and interaction of the small (and not-so small) ensembles. Archival recordings done this way are usually quite good, suprisingly, even with just two microphones.
An omni-directional stereo coincident microphone pair is still the main component of our live recordings. But when required, it’s a missed opportunity to not expand the process further with spot mics, sectional, sub-group (and choral) mics, and even ambient mics out in the house – often dedicated to the rear, or ‘surround” component of 5.1 mixes, or just for natural reverberation and applause mics. Modern electronics, preamps and balanced cables with lower noise floors, with unlimited additional “virtual” digital tracks in the recording process all add up to more flexibility, and zero sonic tradeoffs. With today’s hi-resolution digital recording technology, affordability, and downright amazing editing capabilities, the choices for creating a good, solid, and exciting classical recording continue to expand, with multiple microphone techniques once shunned and abhorred by audiophiles and purists.
One final argument for using just two microphones – time delays between overly distant microphones – is now moot; a thing of the past. With digital editing, a simple timeline adjustment of the track(s) in question can restore perfect time-alignment & phase coherency, and is one less issue to contend with for the ideal combination of performance space via microphone pickup. A simple impulse recording prior to the session or concert creates excellent alignment templates to eliminate all time-delays arising from multiple microphone use.
Multiple microphones and tracking can indeed be tricky and downright wrong in some classical recordings, but when used for the right reasons, it is not only desirable but essential for creating the REAL “best seat in the house” recordings that will stand the test of time, long after today’s “modern” expectations come and go. Without a doubt: in the wrong hands, too many microphones or improper use can wreck havoc on an already balanced performance that the conductor and musicians have created. But in our experience, judicious microphone selection, placement and blend all combine to create a more desirable, detailed and natural listening experience for the serious music lover/audiophile.
Mixing and editing between takes or performances is now seamless in the digital domain, allowing the classical music world unprecedented results. These combined processes retain all of the purity and integrity of the classical genre, while at the same time granting it the flexibility and power of its modern music cousins.
In conclusion, single-point stereo ambient mic placement, while once a great idea that still has merit, only hints at what is now possible for a complete and satisfying Classical Music listening experience.
Joe Hannigan, Producer
©2004-2012, Weston Sound & Video