Subwoofers & Bass Management

Everything the earlier parts of this guide assume about spatial reproduction — that the precedence effect resolves direction from the first arrival, that amplitude panning produces a stable phantom, that the direct-to-reverberant ratio sets perceived distance — quietly depends on the system actually delivering those cues at the listener's ears. Nowhere is that dependence more fragile, and more often broken, than at low frequencies. Below a few hundred hertz the wavelength becomes comparable to the room itself, the room stops behaving like a free field and starts behaving like a resonator, and the loudspeaker you carefully positioned for imaging becomes almost irrelevant to where the bass energy ends up. Bass management is the discipline of taking this awkward physics in hand: deciding which loudspeakers reproduce the low end, how the subwoofer hands off to the mains, how multiple subwoofers can be made to fight the room instead of feeding it, and how the dedicated low-frequency-effects channel fits into the picture.

This chapter builds the subject from the wave equation up. It connects directly to time alignment and phase for the crossover region, to equalization and room correction for what to do about modes you cannot cure with placement, and to measurement and calibration for verifying the result. The recurring theme of Part V applies here in its sharpest form: a subwoofer that is loud but uneven, or aligned in level but not in time, actively destroys the coherent arrival and controlled spectral balance the rest of the system was designed to produce.

Why Low Frequencies Are Special

Wavelength and the failure of the free field

The wavelength of sound is $\lambda = c/f$, with $c \approx 343$ m/s at $20^\circ$C. At 1 kHz that is 0.34 m, comfortably smaller than a loudspeaker, a head, or the gaps between room boundaries. At 40 Hz it is 8.6 m — larger than most listening rooms in at least one dimension. A 1 kHz wave radiated into a room undergoes hundreds of reflections that, statistically, average into a diffuse field; geometric and statistical acoustics describe it well. A 40 Hz wave reflects only a handful of times before it has crossed the room, and instead of averaging it sets up standing waves: discrete resonances at frequencies fixed by the room geometry, not by anything you did to the loudspeaker.

For a rigid rectangular room of dimensions $L_x, L_y, L_z$, the eigenfrequencies are

$$f_{n_x n_y n_z} = \frac{c}{2}\sqrt{\left(\frac{n_x}{L_x}\right)^2 + \left(\frac{n_y}{L_y}\right)^2 + \left(\frac{n_z}{L_z}\right)^2}, \qquad n_x, n_y, n_z \in \{0,1,2,\dots\}.$$

The simplest are the axial modes, with two indices zero, at $f = \frac{c}{2L}$ and its multiples along each axis. In a room 6.0 m long, the lowest axial mode sits at $343/(2\times 6.0) \approx 28.6$ Hz, with overtones at 57, 86, 114 Hz, and so on. Between these resonances the response sags; on them it peaks by 10–20 dB at antinodes and collapses to a deep null at nodes. This is the physical reality of bass in a room, and it is the reason this chapter exists. The boundary between "modal" and "diffuse" behaviour is captured by the Schroeder frequency,

$$f_S \approx 2000\sqrt{\frac{T_{60}}{V}},$$

with $T_{60}$ in seconds and volume $V$ in cubic metres. For a 100 m³ room with $T_{60} = 0.4$ s, $f_S \approx 2000\sqrt{0.004} = 126$ Hz. Below $f_S$, individual modes dominate and the response is position-dependent and lumpy; above it, modal density is high enough that statistical reverberation describes the field. Subwoofers live entirely in the modal region.

Why localization is weak below ~80 Hz

The auditory system localizes using two binaural cues, as developed in psychoacoustics: the interaural time difference (ITD) and the interaural level difference (ILD). The ILD relies on the head shadowing the far ear, which requires the wavelength to be comparable to or smaller than the head diameter (~0.18 m, i.e. above roughly 1.5–2 kHz). At 60 Hz the wavelength is nearly 6 m and the wave diffracts cleanly around the head — no shadow, no ILD. The ITD remains physically present, but at long wavelengths the difference is a small fraction of a period and the system's ability to exploit it for a steady low-frequency tone is poor. The practical consequence, confirmed by listening tests behind every consumer and cinema bass-management standard, is that a single band-limited low-frequency source below about 80 Hz cannot be localized by direction in a normally reflective room.

This is the permission slip for the whole enterprise: if the ear cannot tell where deep bass comes from, we are free to reproduce it from one or a few optimally placed subwoofers rather than from every main loudspeaker, choosing placement for smoothness rather than for imaging.

Key takeaway

Two facts drive everything below: (1) below the Schroeder frequency the room, not the loudspeaker, determines the bass response at any given seat; and (2) below ~80 Hz the ear cannot localize a band-limited source. The first says bass is a coverage problem; the second says we are allowed to solve it with dedicated, well-placed subwoofers.

Bass Management vs the LFE Channel

These two terms are routinely confused, and the confusion produces real calibration errors. They are different things.

Bass management: redirecting the low end of full-range channels

Bass management is a signal-routing function inside a processor or DAW. Each main channel — left, right, centre, surrounds, height speakers — is split by a crossover into a high-passed component sent to the speaker and a low-passed component summed into the subwoofer. The rationale is purely practical: small or mid-sized loudspeakers cannot reproduce 30 Hz at concert level without distortion or excursion limits, while a dedicated subwoofer can. Relieving the mains of deep bass also reduces their excursion-related distortion in the midrange (Doppler and intermodulation), an under-appreciated benefit.

If $S_i(f)$ is the signal in main channel $i$, $H(f)$ the high-pass and $L(f)$ the complementary low-pass, then the speaker receives $H(f)\,S_i(f)$ and the subwoofer receives the sum of the low-passed mains:

$$S_{\text{sub}}(f) = L(f)\sum_i S_i(f).$$

All the bass of every channel collapses to mono in the subwoofer — acceptable precisely because of the localization argument above.

The LFE channel: a separate, dedicated signal

The LFE (low-frequency effects) channel is the ".1" in 5.1, 7.1, 7.1.4. It is not the subwoofer feed; it is a discrete content channel, band-limited at production to roughly 120 Hz, carrying material the mixer deliberately placed there (the cinematic "rumble"). Its defining technical feature is a +10 dB in-band gain offset relative to the main channels. This headroom was chosen so that high-level effects could be encoded without clipping the channel; on playback the decoder applies a corresponding +10 dB to the LFE before it reaches the subwoofer, restoring the intended level.

The subwoofer therefore reproduces two superimposed signals: the bass-managed sum of the mains, and the gained-up LFE. See formats and surround and matrix for how the LFE sits within channel-based delivery.

A routing table

The table below makes the two paths explicit for a 5.1 system with an 80 Hz crossover.

Source	Crossover treatment	Gain applied	Destination
L, R, C, Ls, Rs	High-pass 80 Hz	0 dB	Respective speaker
L, R, C, Ls, Rs (low part)	Low-pass 80 Hz	0 dB	Summed to subwoofer
LFE (.1)	Low-pass ~120 Hz	+10 dB	Subwoofer (added to sum)

A standing trap

The LFE channel and the subwoofer output are not the same node. A mix can have no LFE content and still drive the subwoofer hard through bass management; conversely, "turning off the sub" silences both the bass-managed bass and the LFE. Calibrating the system as if the subwoofer were only the LFE — the single most common conceptual error in bass setup — leads to either a +10 dB error or a misjudged crossover. Keep the two mentally and on paper separate.

To bass-manage or not

In a mastering or reference room with genuinely full-range main loudspeakers, you may choose to run the mains full-range and use the subwoofer only for LFE. This is legitimate but demands that the mains actually reach the bottom octave cleanly and that their in-room low end is itself smooth — rarely true without help. Most rooms, including high-end ones, benefit from bass management because it lets multiple subwoofers smooth the bass for all channels at once.

The Crossover

Slopes, the standard 80 Hz, and what the number means

The crossover frequency $f_c$ is where responsibility transfers from subwoofer to main. The near-universal default is 80 Hz, enshrined in cinema (THX, Dolby) and consumer (ITU-R BS.775 implementations) practice. The choice is a compromise: high enough that even a small main can be high-passed and relieved of damaging excursion, low enough to stay below the ~80–100 Hz localization threshold so the (often single, often off-centre) subwoofer is not heard as a separate spatial source.

Slope is the rate of roll-off, in dB per octave. A 4th-order Linkwitz–Riley (LR4, 24 dB/octave) is the bass-management standard because the high-pass and low-pass branches sum to a flat magnitude and are in phase at $f_c$. Its branches are each $-6$ dB at $f_c$, summing to 0 dB:

$$|H(f_c)| = |L(f_c)| = \tfrac{1}{2},\qquad |H(f_c)+L(f_c)| = 1 \;(0\text{ dB}).$$

A 2nd-order Butterworth (12 dB/oct), by contrast, has branches $-3$ dB at $f_c$ that are $90^\circ$ apart; they sum flat only if one branch is polarity-inverted, and then exhibit a phase rotation. Steeper slopes confine each transducer to its comfortable band and reduce overlap, but introduce more phase shift (group delay) through the crossover region — relevant when you align the sub, as we will see.

Acoustic vs electrical crossover

The crossover you dial into the processor is the electrical target. What matters acoustically is the acoustic crossover: the electrical filter convolved with each transducer's own roll-off. A subwoofer already rolling off mechanically above ~90 Hz, combined with an 80 Hz electrical low-pass, produces a steeper and lower effective acoustic slope than the filter alone suggests. Likewise a small main is not flat to 80 Hz; its own high-pass behaviour adds to the electrical one. The lesson: set the electrical crossover, then measure the acoustic result. Two systems with identical 80 Hz LR4 settings can sum very differently because their transducers differ.

Summation at the crossover frequency — a worked example

At $f_c$ both the sub and the main are radiating at roughly equal level. Whether they add depends on their relative phase at the listening position, which includes the filter phase, the transducers' own phase, and — critically — the path-length difference from each to the listener.

Worked example. A subwoofer 4.0 m from the listener and a main 2.5 m away, both crossed at 80 Hz LR4. The acoustic path difference is $\Delta d = 1.5$ m. The extra propagation delay is

$$\Delta t = \frac{\Delta d}{c} = \frac{1.5}{343} = 4.37\text{ ms}.$$

At 80 Hz the period is $T = 1/80 = 12.5$ ms, so the path difference corresponds to a phase angle of

$$\phi = 360^\circ \times \frac{\Delta t}{T} = 360^\circ \times \frac{4.37}{12.5} \approx 126^\circ.$$

LR4 branches are nominally in phase at $f_c$, so the net misalignment here is dominated by that $126^\circ$ of path-difference phase. Two equal-amplitude signals (each at level $A$) summing $126^\circ$ apart give a resultant

$$|A + A e^{j\phi}| = 2A\left|\cos\!\frac{\phi}{2}\right| = 2A\cos(63^\circ) = 0.908\,A,$$

i.e. about $-0.8$ dB relative to one source alone — a partial cancellation right at the crossover, producing a notch instead of the intended flat sum. Had the phase reached $180^\circ$ the two would have cancelled to a deep null. This is precisely why the next section exists: alignment is not optional polish, it is what makes the crossover sum correctly.

Why 80 Hz keeps the sub invisible

At an 80 Hz crossover with an LR4 slope, the subwoofer's output is already $-24$ dB by 160 Hz. Almost no localizable energy escapes the subwoofer, so even a sub placed asymmetrically does not pull the image of, say, a kick drum toward its corner — the localizable harmonics stay in the mains. Push the crossover up to 120 Hz and this guarantee weakens; the sub starts to become audible as a place.

Sub-to-Main Alignment: Delay and Polarity

The summation example showed that two correctly crossed sources still cancel if they arrive misaligned. Alignment has two knobs — polarity (coarse, $180^\circ$) and delay (fine, continuous) — and the goal is to make the sub and main arrive in phase through the entire crossover overlap region, not only at the single frequency $f_c$. This is the same discipline developed in time alignment and phase, applied to the lowest band.

Polarity: the coarse adjustment

If the sub and main are roughly a half-period apart at $f_c$, they subtract. Flipping the polarity of the subwoofer (a $180^\circ$ rotation, flat across frequency) can convert that cancellation into addition. Polarity is a discrete choice — try both, keep whichever gives more output through the crossover. But polarity alone cannot fix an arbitrary phase offset; it only rotates by $180^\circ$. The continuous fix is delay.

Delay: the fine adjustment

Physically, the sub usually sits closer to a boundary or further from the listener than the main, and its steep low-pass filter adds group delay. The combined effect is an acoustic time offset $\tau$ between the sub's and main's arrivals. You compensate by delaying whichever source is early so both arrive together. Required delay equals the arrival-time difference:

$$\tau = \frac{d_{\text{sub}} - d_{\text{main}}}{c} + \Delta\tau_{\text{filter}}.$$

Using the earlier geometry ($d_{\text{sub}}=4.0$ m, $d_{\text{main}}=2.5$ m) the path part alone is 4.37 ms; if the LR4 low-pass adds, say, another ~2 ms of group delay at the crossover relative to the main's high-pass, the sub is "late" overall and you would delay the main by the net amount so they realign. In practice you do not compute this blind — you measure (see below) and adjust delay to maximize summation.

The measurement-driven method

The reliable procedure uses a dual-channel FFT analyzer (Smaart, REW's overlay, etc.):

1. Measure the main alone through the crossover region; note its phase trace.
2. Measure the sub alone; overlay its phase trace.
3. In the overlap band (roughly half an octave either side of $f_c$, ~55–115 Hz for an 80 Hz crossover) adjust sub delay until the two phase traces lie on top of one another. Parallel phase traces = time-aligned.
4. Try polarity both ways; keep whichever, combined with the best small delay, yields the highest, smoothest summed magnitude.
5. Verify by measuring sub + main together: you want a smooth sum with no notch at $f_c$, ideally within $\pm$2 dB of the individual levels.

Quick polarity-and-delay check without a dual-FFT rig

Play band-limited pink noise or a slow sine sweep covering 50–120 Hz. With an SPL meter or RTA at the listening position, sweep the subwoofer's delay (or phase control) while watching the level at the crossover frequency. Maximum level = best alignment; a sharp dip means you have found anti-phase, so flip polarity and re-optimize. It is crude but catches gross errors that wreck the bottom end.

Multiple Subwoofers for Mode Control

Why one subwoofer gives uneven bass

A single subwoofer excites the room's modes from one location. Its coupling to each mode depends on where it sits relative to that mode's pressure pattern: a sub at a pressure antinode drives the mode hard; at a node it cannot drive it at all. The result is a low-frequency response that is strongly peaked at modal frequencies and full of nulls, and — worse for a multi-seat room — different at every seat, because each listener sits at a different point in each mode's standing-wave pattern. You can EQ the response flat at one chair and make three others worse. EQ moves energy in frequency but cannot fill a spatial null; only changing the spatial excitation of the modes can. That is what multiple subwoofers do.

The principle: shaping the modal excitation

The pressure at a listener from several subwoofers is the complex sum of each sub's contribution. By choosing sub positions (which set how strongly each sub couples to each mode) and, in active schemes, each sub's gain, delay and polarity, you can arrange for the subs to drive the problem modes out of phase with one another, cancelling the peaks and partially filling the nulls — and crucially, doing so consistently across the seating area. Welti and Devantier's work at Harman (JAES, 2006) showed quantitatively that a small number of well-placed subs dramatically reduces seat-to-seat variance, and that symmetry is the key: subs placed symmetrically with respect to the room's symmetry planes excite the lateral modes in a balanced way.

Common multi-sub schemes

Mirrored pair (two subs). One at each side, symmetric about the room's centre line (e.g. at the mid-points of the front and back walls, or both at front-wall quarter points). The symmetric pair tends to cancel the odd lateral modes and smooth left-right variance — a large, cheap improvement over a single corner sub.

Four subs at wall mid-points. Welti's recommended robust configuration: one sub at the centre of each of the four walls. This drives the lowest axial modes in a way that produces low seat-to-seat variance across a wide area, with little or no per-sub tuning required. The mid-wall position sits at the node of the first mode along that wall, suppressing its excitation.

Distributed / "sound field management." Geddes' approach: place several subs asymmetrically and at differing distances (and often differing levels) so that no single mode is reinforced coherently across listeners — the modal peaks of one sub fall on the dips of another, and the spatial average flattens. Asymmetry here is deliberate; it spreads the modal contributions in frequency and space.

Double bass array (DBA). An advanced scheme: a grid of subs on the front wall radiating a plane wave down the room, and a second grid on the rear wall, delayed and inverted, that absorbs the wave (active absorption) so it cannot reflect and form length-wise modes. Done well it yields an almost mode-free, very even response, but it demands many subs, precise delay, and a rectangular room.

A comparison table

Scheme	Subs	Tuning effort	Seat-to-seat evenness	Notes
Single (corner)	1	Low	Poor	Max output, worst smoothness
Mirrored pair	2	Low	Good	Symmetry cancels odd lateral modes
Four mid-wall	4	Low–moderate	Very good	Robust, little per-sub tuning
Distributed (Geddes)	3–5	Moderate	Very good	Asymmetric on purpose; level/position varied
Double bass array	8+	High	Excellent	Needs rectangular room, precise delay

The "sub crawl" — free smoothing for one sub

If you are stuck with a single subwoofer, place it temporarily at the primary listening position and play steady bass. Crawl around the room boundary on hands and knees, listening (or metering) for where the bass sounds most even across the frequencies you care about. By acoustic reciprocity, the spot where the seat-borne sub sounds best is where a sub placed there will sound best at the seat. Put the sub there. It will not match two well-placed subs, but it beats guessing.

Setting Subwoofer Level and the In-Room Target

Modal coupling and boundary gain

A subwoofer's effective output rises as you move it toward boundaries, because reflections from nearby surfaces add in phase at long wavelengths — "boundary loading." In free space the sub radiates into full space ($4\pi$ steradians). Against a wall it radiates into a half-space ($2\pi$), roughly $+6$ dB; into a wall–floor edge ($\pi$, two boundaries) about $+12$ dB; into a tri-corner ($\pi/2$) about $+18$ dB. Idealized, each boundary adds ~6 dB at the lowest frequencies:

$$\Delta L \approx 6\,n\ \text{dB},\qquad n = \text{number of adjacent boundaries.}$$

A corner thus gives maximum output — and maximum modal excitation, hence maximum unevenness. This is the central tension: the loudest position is usually the worst-sounding one. Choose position for smoothness, then make up any lost level with gain or more subs, never the reverse.

The in-room target curve

A subwoofer is not aligned to a ruler-flat magnitude. Because of the way the ear and the diffuse field interact, the accepted in-room target rises toward low frequencies. Cinema (the "X-curve" relatives) and well-regarded music-room practice use a gentle downward tilt from low to high overall, meaning the bass sits a few dB above the midrange anchored at the listening position. A common target is roughly +3 to +6 dB of subwoofer region relative to the mains' midband for a small room, tapering by personal taste; cinema rooms run a defined house curve. The point is that "flat sub" usually sounds thin; a modest low-end lift is normal and intended. Set this with equalization and room correction in mind — but fix placement and alignment first, because EQ cannot repair a spatial null.

Matching sub level to the mains

With bass management, the sub must reproduce the low-passed mains at the same acoustic level the mains would have produced there, so the crossover is seamless. The standard reference, following ITU and cinema practice, calibrates each main to a defined SPL from a band-limited pink-noise signal, then sets the subwoofer so its band-limited level matches — with the deliberate house-curve offset applied. We return to exact numbers in the LFE section and the step-by-step procedure.

Calibrating LFE Level and Headroom

The reference levels

Channel-based calibration (cinema and the ITU-derived consumer convention) sets each main loudspeaker to produce a reference SPL from a standardized pink-noise signal. In cinema, each main reads 85 dB SPL (C-weighted, slow) from a $-20$ dBFS RMS pink-noise band per channel. The subwoofer is set so the LFE path reads +10 dB, i.e. 89 dB SPL in band, reflecting the channel's built-in +10 dB gain. (The figure is "in-band" because the LFE occupies only ~20–120 Hz; you measure the band, not broadband.)

Consumer/home practice following ITU-R BS.1116 reference conditions uses a lower absolute reference (commonly each main at 79 dB SPL for a $-18$ dBFS signal, or per the receiver's internal calibration), but the relationship is identical: the bass-managed mains match each other, and the LFE carries its +10 dB.

Separating the two subwoofer contributions

Because the subwoofer reproduces both the bass-managed sum and the LFE, calibration is done in two conceptual steps:

1. Bass-managed level. Set the subwoofer gain so the low-passed contribution from a main matches that main's level through the crossover (this is the seamless-crossover goal above, with the house-curve offset).
2. LFE level. Verify that the discrete LFE path, after the decoder's +10 dB, lands at the +10 dB target relative to a single main. Many processors handle the +10 dB internally; the error to avoid is double-applying it or omitting it.

Headroom

The +10 dB exists for headroom. If your subwoofer and amplifier cannot deliver +10 dB of clean output above a single main's contribution across 20–120 Hz, loud LFE-heavy content will clip or compress the sub. Sizing the subwoofer for this headroom — not for "how loud can it go in isolation" — is the engineering point. A useful sanity check: with all channels driven plus LFE at reference, the subwoofer should still have visible excursion margin.

Bass management in immersive systems

Immersive formats (7.1.4, 9.1.6 and object-based layouts; see object-based audio and formats) add height channels and more surrounds, but the bass-management principle is unchanged and arguably more important: there are now many small loudspeakers — height units especially are often compact — none of which should be asked to make deep bass. All bed and height channels are high-passed and summed to the subwoofer(s). There is still one LFE channel (the ".1" or ".2"); height beds do not get their own LFE. With two subs in a ".2" arrangement, the LFE and the bass-managed sum are distributed across both, usually as a correlated (mono) feed optimized for room coverage by the multi-sub logic of the previous section, not as a stereo bass image — low bass cannot be imaged anyway.

Do not let immersive multiply your bass

A common immersive mistake is running some channels full-range and others bass-managed, or letting each height speaker try to reproduce its own low end. The bass then arrives from many positions with random relative phase, summing to a comb-filtered mess that no single EQ can fix. Decide a single, consistent bass-management policy — almost always: high-pass everything, sum to the managed subwoofer system — and apply it uniformly.

Step-by-Step Sub Integration Procedure

The following assumes a measurement setup (calibrated mic at the primary seat, dual-FFT or REW) and a processor with per-output gain, delay, polarity, and crossover. It ties together every section above.

1. Set the crossover. Choose $f_c = 80$ Hz LR4 as default (raise only for genuinely full-range, smooth mains; lower only if mains extend cleanly below). Apply the high-pass to all mains and the complementary low-pass to the sub feed.

2. Place the subwoofer(s) for smoothness. Use symmetry (mirrored pair, four mid-wall) or the sub-crawl for one sub. Avoid the deepest corner unless you must have the output. Measure each candidate position's low-frequency response at all primary seats; pick the configuration with the lowest seat-to-seat variance, not the flattest single seat.

3. Rough level. Set the subwoofer gain so its band-limited pink-noise level is in the right neighbourhood of the mains (you will fine-tune after alignment).

4. Time-align. Measure main-alone and sub-alone phase through the overlap (~55–115 Hz). Adjust sub delay so the phase traces are parallel; try both polarities; keep the combination that maximizes the summed level at and around $f_c$.

5. Verify the sum. Measure sub + main together. Confirm a smooth crossover with no notch at $f_c$ (target within $\pm$2 dB). If a notch persists, revisit polarity and delay; a deep, narrow null usually means anti-phase, a broad dip means a delay error.

6. Multi-sub optimization (if applicable). With more than one sub, set per-sub gain/delay/polarity (manually by measurement, or via the processor's multi-sub optimizer) to minimize the spatial-average response variation across seats. Re-verify the crossover sum afterward — multi-sub tuning shifts the combined sub phase.

7. Apply the house curve and EQ. Only now, with placement and alignment fixed, apply gentle EQ to tame remaining broad modal peaks (cut, don't boost, into nulls) and to realize the target tilt. See equalization and room correction.

8. Set final level and the LFE. Calibrate the bass-managed level to match the mains plus the house-curve offset; verify the LFE path lands at +10 dB relative to a single main.

9. Listen and cross-check. Use familiar program with known bass, and verify across all seats. Confirm with measurement and calibration that the result holds for the whole audience, not just the sweet spot.

Order is everything

Placement, then time/polarity alignment, then EQ, then level — in that order. Each step assumes the previous one is locked. EQ-ing before you align is fitting a filter to a response that will change the moment you touch the delay; setting level before you align means re-doing it. Resist the temptation to "fix it with EQ" early.

Worked Example: Align, Level, Then Add a Second Sub

Stage 1 — one sub, one main

Room 5.4 m × 4.2 m × 2.7 m. The lowest length mode is $343/(2\times 5.4) = 31.8$ Hz, the lowest width mode $343/(2\times 4.2) = 40.8$ Hz. Primary seat is 3.2 m from the main and 4.6 m from a single subwoofer placed at the front-wall mid-point. Crossover 80 Hz LR4.

Delay. Path difference $\Delta d = 4.6 - 3.2 = 1.4$ m, giving propagation delay $1.4/343 = 4.08$ ms; the sub arrives later than the main by this amount plus, say, 1.5 ms of low-pass group delay relative to the main's high-pass, so the sub is late by ~5.6 ms total. To realign, we delay the main by 5.6 ms so both arrive together. (Equivalently, in a system that only delays subs, we accept the sub's lateness as small and verify the phase.)

Phase check at 80 Hz. After delaying the main, the residual phase offset measured in the overlap is near zero. Suppose before correction it was the raw $4.08$ ms path part: $\phi = 360^\circ \times 4.08/12.5 = 117.5^\circ$ — a partial cancellation, $2\cos(58.8^\circ) = 1.03$, only $+0.3$ dB over one source, i.e. badly short of the $+6$ dB ideal coherent sum. After aligning to $\phi \approx 0$, two equal sources sum to $2A = +6$ dB: full coherent addition, the flat crossover we want.

Polarity. With the delay set, we test polarity: normal polarity gives the smooth sum; inverting drops the level at $f_c$ by ~15 dB (a null). Keep normal polarity.

Level. Band-limited pink noise: the main reads 85.0 dB SPL (C, slow) at the seat. We set the sub so its band level reads 85.0 dB for the bass-managed path, then apply a +4 dB house-curve lift, landing the sub region at ~89 dB at the lowest frequencies tapering to match by $f_c$. The LFE path, with the decoder's +10 dB, is verified at ~95 dB in band relative to the single-main 85 dB — i.e. +10 dB above one main.

Result at the seat: smooth, but a sweep reveals a $+9$ dB peak at 41 Hz (width mode antinode) and a deep null at 63 Hz at the second seat 0.9 m to the side. One sub cannot fix the second-seat null — it is a spatial node. Enter sub two.

Stage 2 — add a mirrored second sub

Place sub two at the opposite mid-point (rear-wall or symmetric front-wall position), distance 4.9 m from the primary seat. We now have a mirrored pair.

Per-sub alignment. Align sub two to the main exactly as in Stage 1 (delay $\approx 0.3$ m / $343 = 0.9$ ms more than sub one, set its delay accordingly; same polarity verified by summation).

Modal effect. The symmetric pair drives the lowest width mode (41 Hz) more evenly: at the central seats the two subs' contributions to the antiphase lateral mode partially cancel, dropping the 41 Hz peak from $+9$ dB to about $+3$ dB, and — the real prize — the 63 Hz null at the side seat fills to within $-3$ dB because the second sub couples to that mode where the first could not. Measured seat-to-seat standard deviation across four seats falls from ~6.5 dB (one sub) to ~2.5 dB (mirrored pair) in the 30–80 Hz band.

Level re-trim. Two subs sum to roughly $+6$ dB if correlated; we reduce each sub's gain by ~6 dB to keep the combined bass-managed level at the calibrated target, then re-verify the crossover sum with the main (it shifted slightly when sub two was added). Re-confirm the LFE offset.

Metric (30–80 Hz, 4 seats)	One sub (corner-ish)	Mirrored pair
41 Hz modal peak	+9 dB	+3 dB
63 Hz worst null	$-12$ dB	$-3$ dB
Seat-to-seat std. dev.	~6.5 dB	~2.5 dB
Max output (combined)	reference	+~6 dB

The pair is smoother and louder — a strictly better outcome obtained by spatial means, with only modest EQ left to do.

Common Mistakes and Limits

Common mistakes

Sub too loud. The most frequent error, because boundary-loaded bass feels impressive and because the modal peaks make some notes jump out. A subwoofer set 6–10 dB hot smears the direct-to-diffuse balance — masking midrange detail, exaggerating distance cues, and fatiguing the listener. Calibrate by measurement to a defined target; do not set level by ear on a bass-heavy track.

Wrong polarity. An inverted subwoofer cancels the mains through the crossover, producing a hollow, weak upper bass that listeners then "fix" by turning the sub up, compounding the modal problem. Always verify polarity by maximizing the summed level at $f_c$.

One sub jammed in the corner. Maximum output, maximum modal excitation, worst seat-to-seat variance, and (above an 80 Hz crossover) the bass audibly pulls toward the corner. Acceptable only when a single sub and maximum SPL are non-negotiable — and even then, prefer a mid-wall position and make up level elsewhere.

Confusing LFE with the sub feed. Treated above; it produces $\pm$10 dB level errors and wrong crossover assumptions. The subwoofer reproduces bass-managed mains plus LFE; the LFE is one band-limited content channel with a +10 dB offset.

EQ-ing before placing and aligning. Fitting filters to a response that placement and delay will change wastes effort and, when EQ is used to boost into a spatial null, burns amplifier power and excursion for no audible benefit — you cannot EQ energy into a node.

Crossover too high. Pushing $f_c$ to 120 Hz to "help small mains" sends localizable energy to the (often asymmetric) sub, making it audible as a place and undermining imaging.

Boosting into a null wastes power and risks damage

A 12 dB EQ boost is a $\times 16$ power increase. Aimed at a modal null, it barely moves the seat response (the cancellation is spatial, not spectral) while driving the subwoofer and amplifier toward excursion and thermal limits — and over-energizing other seats where that frequency is not nulled. Cut peaks, never boost nulls. Fix nulls with placement and additional subs.

Limits

No amount of subwoofer integration overturns the physics. Below the Schroeder frequency the room owns the response; you can make it even across seats, but a small room simply cannot be made anechoic-flat at 30 Hz everywhere at once. Multi-sub schemes reduce variance and tame peaks, but they trade some maximum output per sub and require symmetry or careful tuning that real, furnished, non-rectangular rooms only partly afford. The double bass array's near-ideal result demands a rectangular room and many channels of delayed, level-matched amplification — rarely practical outside dedicated installs. Active and passive absorption (bass traps, membrane absorbers) attacks the modal energy at its source and is complementary to multi-sub methods; for the deepest modes, physical absorption of the required magnitude is bulky. Finally, calibration optimizes for chosen seats: a configuration tuned for four central chairs will still be uneven in the back corners. As always in Part V, the goal is not perfection but delivering the assumed cues — coherent arrival through the crossover, controlled low-frequency balance, and even coverage — to the audience that matters.

The throughline

Bass management succeeds when the listener never thinks about the subwoofer: the low end is even from seat to seat, the crossover is inaudible because sub and mains arrive coherently, the overall balance supports rather than swamps the midrange, and the LFE has the headroom to land its effects. Achieve that and the spatial cues built up across the rest of this guide survive contact with a real room's worst frequencies.

References

1. Toole, F. E. Sound Reproduction: The Acoustics and Psychoacoustics of Loudspeakers and Rooms, 3rd ed. Routledge/Focal Press, 2017 — chapters on room modes, multiple subwoofers, and bass management.
2. Welti, T. and Devantier, A. "Low-Frequency Optimization Using Multiple Subwoofers." Journal of the Audio Engineering Society, vol. 54, no. 5, 2006, pp. 347–364.
3. Geddes, E. R. Audio Transducers and the associated work on distributed multiple-subwoofer (sound-field management) systems, 2002.
4. Davis, D. and Patronis, E. Sound System Engineering, 4th ed. Focal Press, 2013 — crossover summation, alignment, and low-frequency coverage.
5. ITU-R Recommendation BS.775 — Multichannel stereophonic sound system with and without accompanying picture (channel layouts and the LFE/bass-management context); and ITU-R BS.1116 reference listening conditions.
6. Dolby Laboratories. Bass Management and LFE technical guidelines, and Dolby Atmos Home/Cinema calibration documentation.
7. THX Ltd. and SMPTE practice on the +10 dB LFE in-band offset and 80 Hz bass-management crossover; SMPTE recommended practices for cinema sound calibration.
8. REW (Room EQ Wizard) and Rational Acoustics Smaart documentation — measuring sub-to-main phase, delay alignment, and modal response.

← Back to Systems & Calibration