Sound levels are relative....to a point.
Within this simple phrase lies the reason of why backline contamination is such a huge problem for live sound in smaller venues.
First, consider the setting for which amplified sound was first introduced; the big open-air concert. Or, similar in affect but looking completely different, the studio session.
Everything that reaches ears comes through the mixing board. It is pretty much that simple. The musicians play and sing, a selection of microphones (and pickups) take those elements that are deemed essential to create the desired sound, those signals are processed to taste and mixed together, and the final result is broadcast from line arrays...or is compressed for streaming or cut into a master disk or whatever.
And perhaps this gives rise to a problematic philosophy. Sound engineers and designers tend to come from this world of control. There was a point reached in studio sessions when each musician was isolated in a sound-proof booth, unable to see the other players, unable to hear anything but what the engineers sent to his or her headphones. Thankfully, most studios have backed off from that, embracing the interaction and life -- and moving to a philosophy that treats the ensemble as the primary source and spot-mics only to bring out nuance in individual instruments.
But we still have this lovely illusion that, since sound is passing through the mixer, we should be able to control what is heard by the audience electronically. And this just plain isn't so.
In a small house, theatrically-trained actors are heard easily without amplification. So are singers...the only problem can be if the accompaniment is overpowering. Which it can be. Un-amplified, brass, drums, and even piano can be enough louder than even a trained voice to make the result unbalanced.
The problems become even greater in the medium-sized house. Through the range from club-like to 2,000 seats, a significant part of what reaches the audience's ears did not come through the sound system.
Levels are relative. It is as appropriate to say "The band is too loud" as it is to say "The singers are too soft." The problem is, there exists an apparently simple solution to the latter. So the approach in the majority of spaces is to try to deal with the problem by amplifying the singers -- usually via wireless microphones.
In the right situations, all that is required is gentle reinforcement. The microphones near-invisibly add a few more dB, and the singers rise above the accompaniment in a natural way. The experience is acoustic; the sound appears to come from the singer and interacts in a natural way with their surroundings, supporting a sense of reality.
The same measures can be taken when a band is not balancing with itself. In many cases the traps will overpower some of the reeds. And often as not there are keyboards, or electric bass, which don't make significant sound without electronics.
My preference is to treat a pit acoustically; for every instrument playing in the pit to be heard in the pit. Keyboard players have monitor speakers that are turned up enough for the other players to hear them. This allows the pit to adjust to each other and act like an ensemble.
This doesn't work so well for having to amplify some of the elements over others. And it confuses many people tremendously when you do something like mic a drum. Because a "drum" isn't an entity. It produces a variety of sounds that, to sound right and sound real, also have to balance with each other. In short, the drum is so loud you can't hear the drum. So I mic the drum to be able to hear it over the drum.
(Or more specifically, I mic to hear the nuances of the snare and the click of the hat -- sounds which get masked by the volume coming off the shells).
And this gives rise to the perception of panacea, in which every single note you will get from anyone in the production will be, "So and so's microphone needs to be louder." Always louder. Never trimming the competing elements. Never understanding that loud is relative, and that making the chorus softer is a better way to allow the solo to be heard.
Because sound is relative, to a point. The point being there are soft edges pushing up into concrete ceilings. As you raise levels, you approach feedback threshold. Far short of actual feedback, sounds will begin to take on an edgy, brittle shimmer, like they are being played through one of those tin-can-and-string telephones.
And you can push the feedback threshold back through judicious equalization. The problem being that you begin to cut into the sound you want.
Even if you avoid feedback, the room itself has acoustic properties. First you begin to drive the air of the room into resonance. Then all the materials in the room begin to vibrate in sympathy. All of these physical effects generate harmonics of their own. As you increase the level of Sound A higher and higher in the speakers in order to make it louder than Sound B, you also produce a Sound C; the room itself. The louder you go, the louder the room is, until all of these secondary sounds are as much competition as the original problem you were trying to solve.
Even in a perfect room, with a perfect system...say, if you gave each audience member a pair of personal headphones, physics still does not allow you to arbitrarily increase volume. Physics -- and biology. The human ear is non-linear, and begins to distort at higher sound pressures. The ear accommodates quickly; what was nice and loud two minutes ago sounds normal now, and ten minutes later begins to sound wimpy and soft. The ear in fact begins to shut down after sufficient exposure to higher levels of sound. First the high end rolls off, meaning everything sounds dull, then the perceived volume drops.
No-one ever wins in volume wars.
So what does this have to do with the backline?
The problem is simple. The leakage from the pit -- loud acoustic instruments like brass and drums, and the monitor levels of keys and bass -- is heard by the audience. As a mixer, you are trapped between two absolutes; the highest practical level you can amplify any sound, and the existing sound that is in competition.
Backline leakage is a problem in almost every way. First, it is sheer volume. Weak singers may not be able to be heard over the natural, un-amplified sound coming out of the pit. Second, it is unbalanced; the backline emphasizes certain instruments at the expense of others. Third, it has a poor spectrum.
This takes a little more explanation. Sound is semi-directional. For a given radiator, the pattern approaches omnidirectional as the frequency lowers. Frequency dependence also counts in reflection; given the scattered surfaces of a typical sunken orchestra pit, the higher frequency content bounces around and gets lost, with less of it escaping the pit. The lower-frequency content treats obstructions like a river treats a small rock; it flows around, and escapes the pit rather less attenuated.
This should be simple to understand. It ever boggles my mind why even many musical directors don't get it. The sound of a band on stage is like a friend across from you at a table. The sound of a band in a pit is like the sound of your friend on the other side of a door. And it isn't made better by asking the friend to talk louder!
This is why, for any situation but the smallest or most open, a pit band won't sound its best without a small amount of carefully selected amplification. Not to make them LOUD. But to make them CLEAR.
Given this, the amplified sound of the band is up against...the leakage from the pit. Just like trying to power up singers over the band via wireless microphones, you are trying to power up the "good" sound (the softer instruments, the nuances of specific instruments, the higher frequencies and other subtleties of performance) above the low-frequency, time-smeared, unfocused mush that makes up most of the backline leakage.
Again, this isn't something the band can do themselves. If you hit the drum louder, the "click" of the stick gets louder, but so does the "thooooummmp" of the shell. Because hearing is non-linear and increased volume can lead to increased resolution you will get a slightly more defined drum sound if you just increase the player's volume. But it isn't anywhere as nice as the amplified sound that selectively takes just one element of the sonic picture and presents it to the audience without any of the filters of the local geography between the drummer and the audience ears.
And bands, too, drive the rooms. The louder they play, the more the set walls, the other instruments, the air itself vibrates in sympathy. All these extraneous and distracting noises get louder and louder as well -- and in a non-linear fashion.
This is why backline leakage is the bane of sound techs in every medium-sized and smaller venue. In clubs, it is near-impossible to fix a band's sound via external electronics. If the guitarist insists on turning up his cab, then loud guitar will be all anyone hears -- the rest of the band might as well go home.
In the theater, in the pit, it isn't quite as dire. But the basic simplicity remains; if the band plays loud, if their monitors are loud, then the sound will suck.
Because the mixer is up against the concrete wall of sonic maximums. When the band is loud, it leaks into the very microphones that are on the singers. I've had plenty of shows in which bringing up the chorus was exactly as if you'd turned up the band 5-10 db. There are times when the drums are so loud they are -- quite literally -- louder in the singer's microphone than the singer is. You would get the singer "louder" (relatively, that is), only by turning them down.
To get the singers to sound decent you need to support them over the total sound of the band. To get the band to sound decent, you need to support them over the distracting leakage from the pit. And you have an absolute limit as to how hot you can run.
Really, it would be better if the band could be more controlled. But that is something that does not seem to happen.