Gain Structure Matters More in the Ballroom Than at the Rock Show

Most of us learned gain structure mixing bands. Set the preamp, push the fader, turn it up until it sounds big. If the kick is a little hot or the vocal runs close to clipping, the room is loud enough that nobody notices.

A corporate general session is the opposite situation. The room is quiet. The program is mostly speech. The presenters aren’t performers, and half of them have never held a microphone. The same audio is going to the room, the stream, the recording, and an overflow room down the hall, all at once. Every mistake in the gain structure shows up somewhere, and some of them stay on the internet.

That’s why I think gain structure matters more in a ballroom than at a rock show.

What gain structure actually is

Gain structure is setting the level at each stage of the signal path so the signal sits well above the noise and well below clipping, all the way from the microphone to the speakers. The preamp brings the mic up to a healthy level. The fader is for mixing, not for fixing a bad preamp setting. Everything after that, the bus, the matrix, the amplifiers, the record and stream feeds, gets a level it was designed for.

Get it right and the system is quiet, clear, and predictable. Get it wrong and you get one of two problems: noise, because something was set too low and turned up later, or distortion, because something was set too hot and ran out of headroom.

Why corporate rooms are less forgiving

There’s nothing to hide behind. In a band mix, guitars, drums, and a loud room mask a lot. In a general session, the noise floor is right there between sentences. A preamp set too low and made up with 15 dB of fader gain brings up the hiss, the HVAC, and every page turn at the podium.

The talkers are all over the place. A host who projects, a CFO who mumbles into their notes, a panelist who leans in and laughs into the mic. A band is consistent from night to night. Corporate presenters aren’t consistent from sentence to sentence, so the structure needs enough headroom for the loud moments and enough gain for the quiet ones.

Gain before feedback is tight. Lavaliers sit a foot from the mouth, open in a room full of speakers, often with presenters walking right under them. Every extra dB of gain you need because of a bad setting is a dB closer to feedback. Good structure is how you get the quiet presenter loud enough without the room ringing.

There are more outputs than the room. The stream, the recording, the hybrid meeting platform, the overflow room, the press feed. Each one depends on the level coming out of the console. A clipped input is clipped everywhere. A noisy one is noisy everywhere, and the stream encoder makes it worse.

The recording outlives the event. If a band’s snare clips on a Tuesday, it’s gone by Wednesday. If the CEO’s keynote clips on the webcast, it’s in the archive, on the company intranet, and in the highlight reel.

Automixers depend on it. On a panel or a boardroom table, an automixer decides which mic is open based on relative level. If the input gains aren’t matched, it opens the wrong mic, or the loud talker steals the gate from the quiet one.

The operator changes. Breakouts run with techs rotating through. The house system gets touched by venue staff. A structure that’s consistent and documented lets anyone sit down and ride it. One that only makes sense to the person who built it falls apart at the first lunch break.

How to set it up so it holds

• Set gain with the actual talker at their real speaking level. “Check, one, two” at the console doesn’t tell you anything. Get each presenter to say a few sentences at the podium or with the lav on, the way they’ll actually present. If you can’t get them before the show, set conservatively and leave room to come up.
• Leave real headroom. Speech peaks hard on laughter, plosives, and the occasional excited pitch. A sensible starting point on most digital consoles is speech averaging around −20 to −18 dBFS with peaks well under −6. You can always add level downstream. You can’t un-clip a preamp.
• Keep faders near unity. If you set the preamp right, the fader lives near 0 and you have room to ride both ways. A fader parked at −30 or +10 is a sign the preamp setting is wrong.
• Set wireless at the transmitter and receiver, not at the console. Match the receiver output to the console input. Don’t make up for a lav set too low on the pack with extra preamp gain on the desk.
• Give the stream and record feeds their own mix and their own limiter. The room mix is built for the room, with less of the loud talker and more of the quiet one, and it lives with whatever the room acoustics add. The stream needs its own levels and a safety limiter so one shout doesn’t clip the broadcast.
• Treat playback like a microphone. Set the laptop or playback machine at a fixed output (and tape over the volume key if you have to), DI it properly, and set gain from the loudest video in the show, not the first one.
• Write it down. Note gain, transmitter settings, and any presenter quirks in the show file or the console scene. The next operator, or you at 7 a.m. on day three, will thank you.

The short version

At a band show, gain structure is about how good it sounds. At a corporate show, it’s about whether the quiet presenter is heard, the room stays out of feedback, and the stream and recording come out clean, all at the same time and usually without a second chance.

The band plays again tomorrow night. The keynote happens once.