Speaker and subwoofer placement that measurably helps

Move the subwoofer before you change a setting in the receiver menu. Placement outranks every knob you have, and in the bass it is not close: standing waves put pressure nulls at fixed spots in your room, and in the rooms Harman's engineers measured, the dip at those nulls runs about 25 dB deep. The front speakers at least have an international standard telling you where to put them. The subwoofer does not, which is why the accepted method for finding its spot is a crawl on the carpet.

The 60 degree triangle is an actual standard

Recommendation ITU-R BS.775 is the reference for multichannel loudspeaker placement, and the current edition is BS.775-4, approved December 2022 and listed by the ITU as in force. Its wording is direct: "the left and right frontal loudspeakers are placed at the extremities of an arc subtending 60° at the reference listening point." That is 30 degrees each side of centre, and the equilateral triangle instruction in setup guides traces back to that line.

The table inside Figure 1 is just as usable. Centre: 0 degrees horizontal, 1.2 m high, 0 degrees inclination. Left and right: 30 degrees, 1.2 m, 0 degrees. Side and rear: 100 to 120 degrees from the centre front reference, at least 1.2 m high, inclined 0 to 15 degrees down. The body text adds that "the height of side/rear loudspeakers is less critical," while the front three belong near ear height. Genelec puts monitors at ear level, "usually between 1.2 and 1.4 metres from the floor," and warns against lifting them so high that more than 15 degrees of tilt is needed.

Where BS.775-4 puts each loudspeaker, seen from above A plan view of a listening room from above. The listening seat is at the centre of an arc. The centre loudspeaker sits straight ahead at 0 degrees. Left and right sit 30 degrees each side of centre, so the two of them subtend the 60 degree arc the recommendation specifies. Side and rear loudspeakers fall anywhere in a band from 100 to 120 degrees off the centre front reference, marked on both sides. The recommendation places the front three near ear height at 1.2 m and gives no horizontal toe-in angle at all. BS.775-4 loudspeaker angles, seen from above 60 degrees Listening seat Centre 0 degrees Left 30 degrees Right 30 degrees Side and rear 100 to 120 degrees Side and rear 100 to 120 degrees Front three at ear height, 1.2 m. Genelec puts monitors 1.2 to 1.4 m from the floor. Side and rear at least 1.2 m high, inclined 0 to 15 degrees down. No toe-in angle is given.
The side and rear positions are a band you can land anywhere inside, while the front three are single angles; that asymmetry is the standard telling you which placements it considers critical.

Notice what the standard does not say. Its only orientation column is vertical inclination. Horizontal cabinet rotation, what everyone means by toe-in, is nowhere in that table. BS.775 fixes where a loudspeaker sits on an arc around the listener, not which way it points. Its 30 degrees is a position, not an aiming instruction, though in a symmetric room aiming each speaker at the seat lands on the same number.

Toe-in is a trade you make, not a value you look up

Speakers "radiate more strongly forwards than to the side or rear" at midrange and high frequencies, which DALI notes is where vocal material and cymbal detail live. Acoustic Frontiers gives the mechanism: toe-in "changes the amount of high frequency energy above about 4kHz in the direct sound," and more toe-in "minimizes the effect of reflections because the reflections are lower in level." Lower in level, not gone.

The trade is imaging precision against spaciousness. DALI describes more toe-in as producing "tighter stereo image focus at the hotspot," less as the reverse. Acoustic Frontiers cites research collated in Floyd Toole's Sound Reproduction: listeners often like strong lateral reflections, which increase perceived soundstage width and envelopment. Neither answer is objectively right, so a guide quoting one universal toe-in angle invented it.

KEF's method is the practical one. Start with zero toe-in, flat from the seat, then angle each speaker slightly inward. Follow one instrument rather than the whole mix. Repeat while it keeps improving; when it stops, go back to the previous angle. Expect to finish asymmetric in an asymmetric room: if one speaker sits closer to a side wall, KEF notes, "the reflection boundaries between each will be different so one loudspeaker may need more toe-in than the other."

Find the reflection points with a mirror

Early reflections, per GIK Acoustics, "are reflected sound waves that reach your ears shortly after the direct sound." GIK puts the window at "roughly 30 to 40 milliseconds" in most rooms, one manufacturer's working figure rather than a psychoacoustic constant. They "interfere with your brain's ability to pinpoint where sound is coming from," and recombining with the direct sound they comb filter it into narrow peaks and dips, most audible in the midrange.

Sit in your listening position, have a helper slide a mirror flat along each side wall and across the ceiling, and mark every spot where you can see a driver in it. GIK adds the caveat that keeps the trick honest: sound does not behave like a laser beam, and reflections "occupy areas, not dots."

Treat those zones with broadband absorption, the broader the band the better; thicker panels reach further toward the bass. Genelec ties it back to geometry: "The most accurate stereo imaging can be achieved when the reflections are similar for the left and the right monitor." An asymmetric room is a placement problem before it is a panel problem.

Room modes decide your bass before your equipment does

The source is Todd Welti and Allan Devantier of Harman International, "Low-Frequency Optimization Using Multiple Subwoofers," Journal of the Audio Engineering Society, Vol. 54, No. 5, May 2006. Its abstract: "the frequency response changes significantly from one listening location to another; therefore the system cannot be equalized effectively."

Equation 3 gives a rectangular room's standing wave frequencies as f = (c/2) · sqrt[(nl/l)² + (nw/w)² + (nh/h)²], with c the speed of sound, "typically 344 m/s," l, w and h the dimensions in metres, and each n an integer from zero. Axial modes, set by a single dimension, "dominate the low-frequency performance" in most rooms. Your dimensions predict which frequencies misbehave, not where a null lands; that depends on position within the pattern, whose maxima always sit at the room boundaries.

This is why correction DSP does not replace moving the box. A null is a physical cancellation, around 25 dB deep in the rooms that paper investigated, so boosting it asks the amplifier to raise something not present at your seat and can wreck other seats. Position first, correction second.

Corner placement is the convenient mistake. Every adjacent boundary reinforces low frequencies; Genelec quotes "up to 6 dB higher sound level" for a monitor placed against a solid wall. Axiom Audio, which sells subwoofers, states the cost: corner loading produces "a frequency response curve that looks like hills and valleys," and "more uneven bass than even if you pull the subwoofer out two or three feet from that corner."

Sub positionOutputSeat-to-seat evennessPick it when
Hard cornerHighest, three boundariesWorst, deep peaks and nullsMax output, one listener
2 to 3 feet off the cornerHighBetter than the corner, per AxiomDefault start for one sub
Wherever the crawl landsVariesBest single-sub result, one seatOne sub, one main seat
Two subs, opposing wall midpointsHigh combinedNear best in Harman's modellingSeveral seats must match

The last row is the one worth acting on: two subs exploiting destructive interference cancel odd-order modes, an idea the paper credits to Floyd Toole. A third at the centre sits in the null of those modes; raise its gain 6 dB and the second-order mode cancels too. Sweeping 100,000 configurations, the authors found wall midpoints optimal for one to four subwoofers: four best in practice, two at opposing midpoints nearly as good with stronger low frequency support.

Candidate subwoofer positions in a rectangular room A plan view of a rectangular room from above, front wall at the top and the seating in the middle. Three candidate placements are marked, and they are alternatives rather than a set to use together. One box sits hard into the front left corner, where three boundaries give the most output and the least even bass. One box sits two to three feet out from that same corner, the default start for a single subwoofer. A pair sits at the midpoints of the two opposing side walls, which was near best in Harman's sweep of 100,000 configurations. A shaded band just inside the walls marks the boundaries, where standing wave maxima always sit. Candidate subwoofer positions in one room front wall opposing wall midpoints Sub 1 Sub 2 Hard corner three boundaries 2 to 3 feet out the default start for one sub seating Shaded band: standing wave maxima always sit at the room boundaries. Two at opposing wall midpoints were near best in Harman's sweep of 100,000 configurations. Genelec quotes up to 6 dB from one solid wall. Nulls in the rooms Harman measured ran about 25 dB deep.
These are alternatives, not a shopping list. The first two are the same box moved a few feet; the third is a different strategy, which is why it is the only one that changes what happens at the other seats.

The paper then undercuts its own rule. In one of three real rooms the best pair was not at the wall midpoints, and the authors conclude it is always better to measure and pick positions from that. All of it is about seat-to-seat consistency. One subwoofer can be optimized for one chair. It cannot be optimized for a sofa.

The subwoofer crawl, as KEF publishes it

You need a long cable and a bass-heavy track you will not tire of, because it loops. Put the subwoofer on the seat where you normally sit, start the track, then move at floor level around every spot where the sub could go.

Crawling rather than walking is the step people skip, and KEF says why: "When you stand you'll be off-axis and the sound will be different." Stop at each candidate and compare volume, depth, punchiness and definition. Mark the winner with tape and put the subwoofer there. The swap works because a linear acoustic system behaves the same way when source and receiver are exchanged, which KEF does not spell out. Set crossover and gain afterwards: "A properly tuned subwoofer will not draw your attention."

The crossover, and the 120 Hz number that is not one

Yamaha's RX-V685 manual defines the control. The crossover "sets the lower limit of the low-frequency components that can be output from a speaker whose size is set to 'Small'." The steps on that receiver are 40, 60, 80, 90, 100, 110, 120, 160 and 200 Hz. The page states no factory default, so distrust any article that quotes one for your model.

Eighty hertz is the conventional starting value. Ben Hagens, KEF's senior product training specialist, reports the rationale usually attributed to THX: "In the 1980s, THX determined 80Hz to be their standard with the rationale that 80Hz downwards becomes difficult to localize." Hagens then corrects the folklore built on it: bass being hard to localize "is not the same thing as steering of sound," and "where you end up will depend on a combination of your system and the room." Treat 80 Hz as a default to beat, and raise it when your mains cannot cleanly reach that far down, as small satellites and slim bookshelves often cannot.

Do not confuse it with the other number in these conversations. BS.775-4 recommends the LFE channel "should be band-limited to its nominal frequency band (up to 120 Hz)," and that a broadcast carrying LFE "should not transmit any information on that channel above the nominal 120 Hz cutoff frequency." That governs how the .1 channel is authored, not a setting you select.

One check before blaming the subwoofer. If a main speaker is still set to Large, the crossover does nothing for that channel and the mains keep attempting their own deep bass. Depending on the receiver's bass output option, the sub may then get only the LFE channel rather than redirected bass. Bass that got worse after a subwoofer arrived is worth checking there first.

Sources