What this mic hears
A digital desk laid out like an X32, with a real feedback loop behind it. Gain up a lapel mic until the room rings, find the frequency, notch it out, keep going. The narration is Reboot's own Front of House Mixing course.
To find a ring: turn one band up to +12 with a narrow Q and sweep the frequency. When you hit the ring it gets much worse. Then press FLIP to turn that boost into a cut.
The house PA. Everything the audience hears.
A floor wedge pointed back at the talker. A second, separate feedback path — and usually the one that bites.
Every open mic feeds the same loop. Two open mics instead of one costs you about 3 dB of gain before feedback. Mute whatever nobody is talking into — it is the cheapest fix on the desk.
“Show me” draws where it wants to ring — the answer key. Safety mutes a runaway.
Channel view works the way a digital desk really does: one channel selected at a time, with a screen for detail — the mic's response curve, the EQ graph, routing, the analyser. It is where you go to understand what is happening.
Full desk is the whole surface at once: every channel with its own gain, low cut, four bands of EQ and its sends, laid out like an analogue console. It is where you go to work — you can see all four mics' EQ side by side and reach any of them without selecting it first.
It is one mix, not two. Anything you change in one view is already changed in the other, because both are drawing the same desk. The only thing that lives solely in Channel view is Q (how wide each EQ cut is) — putting freq, gain and Q on the surface would be twelve knobs a channel, so Q stays on the EQ page.
A handheld sits about 3 cm from the mouth and is directional, so it hears a lot of talker and not much room. A lapel sits about 20 cm away on a chest and is omnidirectional, so it hears less talker and all of the PA. You need roughly 8 dB more gain to reach the same level, and you have about 10 dB less room before it rings. In the gym it simply cannot get there — which is the correct answer, not a broken simulator. Some rooms need a different mic, not more EQ.
Listen to the voice change as you switch capsules on the Config page. The lapel loses the top above 4 kHz where speech clarity lives, so the instinct is to boost there — and in a live room that is close to where it wants to feed back.
A lapel in a live room often howls low, not high — in the boardroom it goes at around 133 Hz and in the gym around 143 Hz, because the capsule is flat and the room's low-mid energy wins. The fix there is usually the low cut, not a notch: that is why it now sweeps all the way to 400 Hz. Reach for it before you spend an EQ band.
Watch the ring frequency change too. With the handheld, the boardroom rings at 4.16 kHz, pushed there by the mic's own presence peak landing on a room resonance. Swap to the flat-voiced lapel and the same room rings down at 232 Hz instead — the capsule is no longer lifting the top, so the low-mid room mode wins. The mic decides where a room misbehaves, not just how loudly.
| On this desk | Means |
|---|---|
| Gain | How much the preamp amplifies the mic. The first thing you turn up, and the thing that causes feedback. |
| Low cut | A filter that throws away everything below the set frequency. Cleans up rumble and handling noise. |
| 48 V | Phantom power. Condenser mics (the lapel, the podium) need it; dynamic mics do not. |
| Ø | Polarity flip. Useful when two mics fight each other; it will not fix feedback. |
| PEQ / band / Q | Parametric EQ. Pick a frequency, boost or cut it, and Q sets how wide a slice you affect. High Q = surgical. |
| Main LR | The house PA — what the audience hears. |
| Bus 1 | A separate mix, here feeding a floor wedge pointed at the talker. |
| Gain before feedback | How many dB you can still add before it howls. When it hits zero, it howls. |
Each room is described only by its dimensions and average absorption. Everything acoustic follows from that, using the standard formulas:
Above the transition frequency the room is modelled statistically: dense peaks whose height and sharpness follow the reverb time, so a live room rings harder and narrower than a dead one. One deliberate compromise: a real reverberant peak has a bandwidth of about 2.2/RT60 Hz — at 3 kHz in the gym that is a Q in the hundreds, far narrower than any console filter can touch. That is exactly why you cannot properly EQ a room above its transition frequency. Here they are widened into a console-addressable range so the drill is possible at all.
The Setup page runs the standard sound-system design calculation next to what this room actually delivers:
In the boardroom PAG−NAG says you have about 28 dB of margin with a handheld. The room gives you 14. Both numbers are right. PAG/NAG is purely geometric — inverse-square distance and the number of open mics — and models no reverberation whatsoever. The gap between the two is the room. Use PAG/NAG to sanity-check a design before load-in; trust the room once you are standing in it.
The desk tracks speech clarity: how much importance-weighted level your EQ has taken out of the bands that carry intelligibility (the 2 kHz octave carries roughly 30% of it, 4 kHz 25%, 1 kHz 20%). A 9 dB notch at 2 kHz costs about 2.7 dB of weighted clarity; the same notch at 250 Hz costs 0.8 dB. That is the whole argument for notching outside the speech band wherever the ring allows it — and the reason six deep cuts through the middle leave a voice sounding like a telephone.
The 1–4 kHz band is shaded on the analyser so you can see when you are about to spend clarity, and there is a second line 6 dB below the ring point: that is the stability margin a real system should be run at, not right on the edge.