1
System EQ with RTA
System EQ tunes the PA to the room, and the method is measurement, not taste: play pink noise through the PA, capture it with a measurement mic at the mix position, and read an RTA (real-time analyzer) to see the actual frequency response.
The golden rule is to cut, never boost, you correct the room's peaks and resonances down toward a smooth target curve. Boosting just feeds the resonances and burns headroom.
KNOW WHAT AN RTA CANNOT TELL YOU, because this is the beginner workflow and its limits are the reason the next two cards exist. An RTA shows everything arriving at the mic added together: direct sound, reflections, and room noise, with no way to separate them. So a peak on the trace may be a genuine system response you can fix, or a reflection arriving at that one spot, which EQ will not fix because you cannot equalise a time-domain problem out of a room. Two consequences follow. Move the mic and take three or four positions, because a correction derived from one seat can make every other seat worse. And treat broad shapes as actionable and narrow spikes with suspicion: broad tendencies survive across positions, narrow ones usually do not.
The grown-up version of this measurement is a dual-FFT transfer function, which compares the system input to the mic and gives you coherence, a direct readout of whether the trace is trustworthy at all. Learn the RTA first, then stop trusting it alone.
In practice: A boomy 200 Hz and harsh 3 kHz in a venue: the RTA shows the peaks, so take a second and third mic position to see which ones persist, cut those broadly, and leave the spike that only existed at one seat alone.
System EQ: measure with pink noise and an RTA, cut the peaks that survive across several mic positions, and remember the RTA cannot separate the room from the system.
2
The House Curve: Why Flat Is Not Flat
A perfectly flat measurement does not sound flat, so the target is not a flat line but a "house curve": slightly elevated bass and a gentle high-frequency rolloff above about 4 kHz.
The reasons, in order of how much they actually matter. Air absorption is real and large: high frequencies lose several dB over a long throw and more as humidity drops, so a system flat at the mic is dull at the back and bright at the front. Speaker directivity narrows with frequency, so an on-axis flat measurement is already bright relative to what most of the room hears. And listener preference, measured repeatedly, lands on a gently downward-tilted curve rather than a flat one.
Equal-loudness (Fletcher-Munson) gets cited for this and it is the weakest of the reasons, so do not lean on it: the contours FLATTEN as level rises, so at concert SPL the equal-loudness argument for a house curve is much smaller than at living-room level. It explains why a quiet mix sounds thin, not why a 100 dB system needs a tilt.
In practice: A PA tuned dead flat that the band calls "thin and brittle": add a few dB of low end and roll the highs off gently, the house curve fixes the perception gap.
Tune to a gently tilted house curve, not flat. Air absorption and directivity are the real reasons; equal-loudness is the weakest one.
3
Graphic EQ vs Parametric for System Tuning
Two EQ types do system tuning. A graphic EQ has many fixed bands at even spacing, good for broad, general room correction. A parametric EQ lets you dial the exact frequency, bandwidth (Q), and amount: good for a surgical, narrow notch on a specific room mode or feedback frequency.
Broad problem: graphic. Specific, narrow problem: parametric. Modern digital systems give you both.
In practice: A sharp room-mode resonance falling between a graphic EQ's fixed bands: a parametric notch lands exactly on it where the graphic could not.
Graphic EQ for broad room correction; parametric for surgical notches.
4
Delay Alignment for Distributed Systems
Sound travels at roughly 1,130 feet per second, slow enough that a fill speaker far from the stage reaches a listener noticeably before the main PA does. Without correction the audience hears two arrivals: an echo-like smear.
Delay alignment fixes it: delay the fill speaker so its sound arrives in step with (or just after) the mains. Add a few extra ms so the Haas effect makes the sound seem to come from the stage, not the nearby box.
In practice: A delay tower 150 feet out: about 133 ms of delay (plus a few ms) lines its sound up with the mains so the audience hears one coherent source.
Delay fill speakers to match the mains' arrival: distance / ~1,130 ft/s.
5
Sub Alignment: Phase at Crossover
Subwoofers and the main tops overlap around the crossover frequency, and if they are out of phase there, they cancel rather than add, leaving a hollow notch right where bass meets low-mid.
Sub alignment makes them sum: adjust delay on the subs (and sometimes flip polarity) until the crossover region adds up. It must be measured, ears cannot reliably judge phase at low frequencies.
In practice: A "hollow" gap where the bass should be full: subs and tops are cancelling at the crossover, delay-align them so the region sums.
Phase-align subs and tops at the crossover or they cancel into a notch.
6
Crossover Frequency Selection
The crossover is the frequency where bass handoff moves from the tops to the subs, typically 80-120 Hz in a live PA. The choice is a trade: a lower point (80 Hz) lets the tops carry more low-mid for vocal warmth; a higher point (120 Hz) offloads bass work from the tops for more headroom.
Steeper filter slopes give cleaner separation but demand more precise phase alignment at the crossover.
In practice: Tops struggling for headroom at a loud show: raise the crossover toward 120 Hz so the subs take more of the bass load.
Sub/top crossover sits at 80-120 Hz, lower for warmth, higher for headroom.
7
System Tuning Workflow
Tuning a PA in a new room follows a fixed, repeatable order: play pink noise, measure at the mix position, correct the room's problems with system EQ, re-measure to verify the correction worked, then check with familiar music and walk the room for coverage.
The measure-correct-verify loop is what makes the result reliable rather than a guess, and listening to known music at the end catches what the meter misses.
In practice: An hour to tune a reverberant church: pink noise, measure, cut the room modes, verify, then walk the room with a vocal reference for intelligibility.
Tune in order: pink noise → measure → correct → verify → check with music.
8
Smaart Transfer Function: Measuring What the Room Actually Does
A Smaart transfer function is the standard professional measurement for tuning a PA. Dual-FFT: simultaneously capture the reference signal (direct from the console) and the room measurement (mic) and compare them. The difference is the system's actual response, magnitude, phase, and coherence.
Coherence is the key signal a plain RTA does not give you. A low coherence reading means the measurement is dominated by noise or reflections, not direct sound, so the magnitude trace at that frequency is unreliable. Phase across the crossover region tells you whether subs and tops are time-aligned.
READ THEM IN THE RIGHT ORDER, because the order is most of the discipline. Coherence first, since it decides whether the other two traces mean anything. Phase and time next, because a magnitude dip caused by a misaligned crossover vanishes when you fix the time and is uncorrectable when you do not. Magnitude last, and only after you have captured several positions and averaged them, because the system response is what persists across the room and everything that moves as you walk is the room itself.
Match the filter to what you found. Broad tendencies get broad, gentle cuts, and that is most of what a real system EQ contains. A narrow high-Q notch is for a narrow resonance that held still at every position, which is rarer than it looks.
Used correctly, Smaart takes the guesswork out of PA tuning. Used incorrectly, which nearly always means EQ-ing the magnitude trace at one mic position before checking coherence or time, it gives confident-looking false answers and a room that is flat in exactly one seat.
In practice: Tuning a new room: coherence first to see what is trustworthy, phase next to align subs to tops, then 4-6 averaged positions before a single filter goes in.
Smaart gives magnitude, phase and coherence. Read them in that reverse order: coherence, then time, then tone, and only after averaging several positions.
9
SMAART and SysTune: What Live Engineers Actually Measure
SMAART and SysTune are the two standard live-sound measurement platforms, both do dual-FFT transfer-function analysis showing magnitude, phase, and coherence, plus impulse response and delay-finding.
The workflow is the same regardless of tool: connect a reference feed and a measurement mic, find the delay, confirm coherence, read magnitude and phase, then make EQ and delay corrections, and measure across multiple positions before deciding.
In practice: Setting up SMAART in a venue: mic to one input, a console feed to the other, run the delay finder, watch coherence jump to 0.7+, then read the response.
SMAART/SysTune: connect reference + mic, find delay, verify coherence, then correct.
10
Critical Distance: Where EQ Stops Helping
Critical distance is the distance from a source where the direct sound and the room's reverberant sound are equal in level. Inside it you are hearing the PA. Outside it you are hearing the room.
This is the single most useful number in live sound because of what it implies: past critical distance, turning up does nothing for intelligibility. Level raises the direct and reverberant fields together, so the ratio between them, which is what intelligibility actually depends on, does not move. Every engineer who has ever pushed a vocal harder into a gymnasium and made it worse has met this.
Two things change where critical distance falls. A more reverberant room pulls it closer, sometimes to a handful of feet in a hard, empty hall. A more directional loudspeaker pushes it further out, because a tight pattern delivers its energy to the audience instead of to the walls and ceiling.
So in a room you cannot treat, your levers are all ratio levers: tighter directivity, careful aim, and getting sources physically closer to listeners with delays and fills so that more of the room sits inside critical distance rather than outside it.
In practice: A vocal that is clear at row 8 and mush at row 25: rows past 25 are outside critical distance. Delay fills and tighter aim fix it. More gain does not, because it lifts the room by exactly as much as it lifts the voice.
Past critical distance, level buys you nothing. Only the direct-to-reverberant ratio buys intelligibility: directivity, aim, and shorter distances to the listener.
These lessons are the teach side of knowledge cards from the Live Sound deck. In the daily plan each one is followed by a quiz, spaced over weeks, so it stays known rather than read once.