Mixing and gain
Gain Structure for Church Sound
Quick answer
Set every church input so the channel meter averages -18 dBFS and peaks no higher than -10 dBFS. On a console calibrated to plus 24 dBu at 0 dBFS that is plus 6 dBu of working level with 18 dB of crest headroom, which is enough for a preaching voice that jumps 12 dB on a single word.
Gain structure is the most valuable hour a church sound volunteer will ever spend, and it is almost never taught because it produces no visible result. Nothing sounds obviously different afterwards. What changes is that the noise floor drops, the compressors start behaving, the feedback margin stops moving around week to week, and the livestream stops arriving at the platform 14 dB quieter than everything else on the internet.
The target is a single number you can put on a card in the booth: average around -18 dBFS on the channel meter, peaks no higher than -10 dBFS, channel fader at or near unity. Everything else in this guide is how to get there for each kind of source, and why the number is -18 rather than something closer to the top of the meter. If you are building a system rather than fixing one, the sizing work comes first: the mixer channel calculator tells you how many inputs you need, and choosing a console covers what to put them on.
On this page
Why -18 dBFS and not -6?
Digital consoles are calibrated so that 0 dBFS, the absolute top of the meter, corresponds to a specific analogue level, and on almost all professional gear that level is plus 24 dBu. Aiming at -18 dBFS therefore puts the average signal at about plus 6 dBu, which is the level the analogue world calls nominal and the level every plugin, every dynamics processor and every converter was designed around.
The other 18 dB is not waste, it is crest factor. A preaching voice is not a steady tone: the difference between a conversational aside and an emphasised word is routinely 12 dB, and somebody who leans into a consonant can produce 15 or more. Music is worse. The peak of a snare hit sits 15 to 20 dB above the average level of the mix. Set the average at -6 dBFS and every one of those peaks clips, and digital clipping is not the gentle saturation of an analogue desk, it is a flat top that sounds like a crackle and cannot be undone downstream.
Undershooting is the other failure. A channel averaging -40 dBFS is not clipping, but the preamp has been left at 20 dB of gain when the source wanted 50, so the volunteer has pushed the fader to plus 8 to compensate. The noise of the preamp is now 30 dB higher relative to the signal than it should be, the gate threshold is sitting in a useless place, and the fader has no travel left for the quiet moments. Gain lives in the preamp. The fader is for balance.
Where the meter should sit at every stage
Gain structure is a chain, and it is only as correct as its worst stage. These are the targets from the capsule to the loudspeaker terminals.
Every stage should carry about 18 dB of headroom above its average level, which on a digital console means an average of -18 dBFS and peaks at -10 dBFS from the preamp all the way to the main bus.
| Stage | Target | The mistake to avoid |
|---|---|---|
| Microphone preamp | Average -18 dBFS, peaks -10 dBFS | Setting gain on the loudest word instead of the average, then finding the sermon inaudible |
| Channel fader | Within about 5 dB of unity | A fader parked at the top, which means the preamp is 15 dB too low |
| Gate or expander | Threshold about 10 dB below the quietest wanted signal | A threshold set for a shouting worship leader that chops the first word of every sentence |
| Compressor | 3 to 6 dB of gain reduction on peaks, 2.5 to 1 ratio on a voice | Ten decibels of reduction, which pushes room noise up between sentences and starts a ring |
| Subgroup or bus | Average -18 dBFS | Eight channels summing to a bus that clips while every channel meter looks fine |
| Main LR bus | Average -18 dBFS, peaks -6 dBFS | Using the main fader as the volume control and running the bus 12 dB hot |
| Amplifier input sensitivity | Set so the amp clips at or just above the system limiter | Amplifier attenuators at maximum, which amplifies every bit of console noise |
| Livestream or record feed | Average -18 dBFS, true peak -1 dBFS | A feed taken post main fader, so the online mix follows whatever the room needed |
Consoles calibrated to plus 18 dBu at 0 dBFS, which includes some interfaces and prosumer gear, shift every figure up by 6 dB. Check the manual before assuming plus 24 dBu.
Preamp gain by source
Preamp gain is not a preference, it is arithmetic: the output level of the source plus the gain has to land at plus 6 dBu. These are the ranges each common church source actually needs.
A dynamic handheld on a preaching voice needs 45 to 55 dB of preamp gain, while a stereo keyboard on a direct output needs zero, and a channel set 20 dB away from these figures is the usual cause of a noisy or clipped mix.
| Source | Typical preamp gain | Phantom power | Notes |
|---|---|---|---|
| Dynamic handheld on a preaching voice | 45 to 55 dB | Not required | Distance is everything. A talker at 8 inches needs 6 dB more than one at 4 |
| Dynamic handheld on a strong worship vocal | 35 to 45 dB | Not required | Loud singers routinely need 10 dB less than speakers on the same microphone |
| Lavalier condenser on a lapel | 40 to 50 dB | Required, from the bodypack or the console | Wireless receivers usually output near line level, so gain is lower than you expect |
| Headset condenser at the mouth corner | 30 to 40 dB | Required | The closest capsule in the building, and the quietest preamp setting of any voice |
| Gooseneck on a lectern | 45 to 55 dB | Required | Set for the quietest reader in the congregation, not the pastor |
| Hanging choir condenser | 50 to 60 dB | Required | The highest gain in the building, which is why choir channels ring first |
| Kick drum microphone | 20 to 30 dB | Not required for a dynamic | A kick is 15 dB hotter at the capsule than a vocal |
| Snare and tom microphones | 25 to 35 dB | Not required for a dynamic | Watch the peak, not the average, or you will clip on the rim shots |
| Drum overhead condensers | 25 to 35 dB | Required | Set from the cymbal crashes, which peak 15 dB above the average |
| Guitar amplifier microphone | 20 to 30 dB | Not required for a dynamic | A loud amplifier can need under 15 dB |
| Acoustic guitar through a passive direct box | 30 to 40 dB | Not required | A passive box loses level, an active one does not |
| Bass through an active direct box | 15 to 25 dB | Required by the box | Very hot output. This is the channel most often clipped in a church |
| Stereo keyboard on line outputs | 0 to 10 dB | Not required | Use the line input or the pad, or the preamp will clip before the fader moves |
| Playback from a laptop or media computer | 0 dB with the pad engaged | Not required | Set the laptop output first, then the console, never both at once |
Ranges assume a console calibrated to plus 24 dBu at 0 dBFS and sources at normal church distances. The phantom power chart lists which of these need 48 volts and which are damaged by it.
How to set gain structure on a church console
- Start with every fader down and every preamp at minimum. Not because it sounds better, but because it removes the previous week accumulated guesses from the picture. On a digital console this is what an initialised scene gives you, and it takes about twenty minutes to work through a 24 channel input list from scratch.
- Set the amplifier sensitivity before the console. The last device in the chain sets the ceiling for everything before it. Amplifier attenuators should be set so the amplifier reaches full output at or just above the point where the system limiter engages, not at maximum. An amplifier with its attenuators wide open is amplifying the console noise floor by 20 or more decibels for no useful gain.
- Put the channel fader at unity and leave it there. Unity is the marked zero on the fader, usually about three quarters of the way up. All gain setting happens with the preamp knob while the fader sits at unity. If you find yourself moving the fader to make the meter read correctly, you are setting gain in the wrong place.
- Get the source to make its real sound. Not a tap on the grille and not a count to four. A preacher reading a paragraph at sermon volume, a vocalist singing the loudest line of the loudest song, a drummer playing the chorus. Gain set from a polite sound check is 8 to 12 dB too high for the service and clips the first time the band means it.
- Raise the preamp until the average sits at -18 dBFS. Watch the average, not the peaks, and let the source run for ten or fifteen seconds. Most digital consoles show both a bar and a peak hold. The bar should hover around -18 and the peak hold should land near -10. If the peak hold is pinning at 0, come down 6 dB regardless of what the average says.
- Engage the pad on anything that clips at minimum gain. Bass direct boxes, stereo keyboards and playback channels frequently overload a preamp at its lowest setting. Use the channel pad, usually 20 dB, or take the source into a line input. Turning the source down at the instrument instead is worse, because it drops the instrument below its own optimum output.
- Check the buses and the main last. Twelve channels each sitting correctly at -18 dBFS will sum to a bus somewhere between -8 and -4 dBFS depending on how correlated they are. If the main bus is peaking above -6 dBFS, trim the bus rather than pulling every channel down, and never fix it by lowering the main fader below unity for the rest of the service.
- Store the scene and write the preamp numbers on a card. Digital consoles recall preamp gain with the scene, which is the whole point of buying one. On an analogue desk, write each preamp setting on tape above the knob. Either way the next volunteer starts from a correct chain rather than from whatever the last service left behind.
The livestream feed is a separate gain problem
The most common gain fault in a modern church is not in the room at all. A stream feed taken from the main mix inherits every level decision made for the room, including the moment somebody pulls the main fader down 6 dB because the front row complained. Online, that arrives as the sermon suddenly getting quieter for no reason.
Take the stream from a separate bus or a matrix output, set it to average -18 dBFS in the encoder with true peaks no higher than -1 dBFS, and leave it alone. Streaming platforms normalise loudness, so a feed that averages 12 dB below everything else on the platform gets turned up along with its noise floor and its room reverberation. The full treatment, including why the room mix and the broadcast mix want different reverb and different vocal levels, is in how to mix for livestream.
A cheap interface such as the Focusrite Scarlett 2i2 4th Gen USB-C Audio Interface at $224.99 is fine for this job as long as you respect its calibration, which is not plus 24 dBu. Check the manual, find the number, and shift your target by the difference. The Behringer UMC202HD Audiophile 2x2, 24-Bit/192 kHz USB Audio Interface with Midas Mic Preamplifiers is the budget version of the same idea.
Sources
- Console calibration of plus 24 dBu at 0 dBFS, manufacturer published specifications
- Crest factor figures for speech and live music, standard audio engineering references
- Focusrite and Behringer published input specifications for the interfaces named
Frequently asked questions
Why is -18 dBFS the target and not 0 VU?
They are the same target expressed differently. Professional digital gear is calibrated so that 0 dBFS equals plus 24 dBu, which puts the analogue nominal level of plus 4 dBu at about -20 dBFS and a comfortable working average at -18 dBFS. Aiming there keeps roughly 18 dB of headroom for peaks, which matches the crest factor of speech and live music.
Should I set gain with the fader or the preamp?
The preamp, always. The preamp sets how much of the converter range the signal uses, which decides the signal to noise ratio and where every dynamics processor threshold lands. The fader sets balance between sources that are already correctly gained. A fader pushed to the top is a sign that the preamp is 10 to 15 dB too low, not a solution to a quiet channel.
How much compression should a preaching voice get?
Around 3 to 6 dB of gain reduction on the loudest words, with a ratio near 2.5 to 1, a fast attack of about 10 milliseconds and a release around 150 milliseconds. More than that raises the room noise between sentences, which both sounds worse and eats into your feedback margin. If you need more than 6 dB, the microphone is too far from the mouth.
What gain does a hanging choir microphone need?
Typically 50 to 60 dB, the highest in the building, because a hanging condenser sits 3 or more feet from the singers and the inverse square law has already taken 20 dB or more out of the signal. That high gain is exactly why choir channels ring before anything else, and why they should be gated or muted whenever the choir is not singing.
Does gain structure change anything about feedback?
It does not change the loop gain itself, but it changes every measurement you take. A channel with the preamp low and the fader high has the same feedback point but a much worse noise floor, and the gate and compressor thresholds sit in the wrong place, which makes the system behave differently week to week. Set gain first, then ring the room out.
Do analogue consoles use the same targets?
The principle is identical, the meter is different. On an analogue desk use the channel PFL or solo to set each preamp so the meter reads around 0 VU on the average and the peak light only flickers on the loudest moments. That is the same plus 4 to plus 6 dBu working level, with the console headroom above it, which is usually 18 to 22 dB on a professional desk.
Researched, not professional advice. This page is compiled from published manufacturer specifications, operator manuals, FCC rules, published standards and owner-review consensus, not hands-on testing. Sound system design, rigging loudspeakers overhead, and any electrical work are jobs for a qualified professional: have flown loudspeakers and their attachment points signed off by a structural engineer or a certified rigger, and have all wiring done by a licensed electrician to your local code. Wireless microphone rules change, so confirm the current FCC position before buying. As an Amazon Associate we earn from qualifying purchases.