Bookshelf vs Floorstanding: Where the Low End Stops and a Subwoofer Starts

Two things fall straight out of that table. Moving from a standard bookshelf to a large standmount buys about 8 Hz of -10 dB extension, and moving from a large standmount to a mid tower buys roughly the same again. Neither jump is a cliff.

The cliff is between a 100 mm woofer and everything above it. A compact bookshelf is not a small tower; it is a midrange with a bit of upper bass attached, and every one of them wants help below 80 Hz.

Extension is the least trustworthy figure on a spec sheet, and it is the one the whole purchase turns on.

Some manufacturers publish it anechoic, some in-room, and plenty publish a bare frequency range with no tolerance attached at all. A cabinet listed as 45 Hz to 22 kHz with no dB window is telling you nothing. That 45 could be the -3 dB point, the -10 dB point, or the frequency below which the woofer produces no measurable output at any level. Where a tolerance is stated it is usually plus or minus 3 dB, which is a 6 dB window wide enough for a genuinely lumpy response to pass.

When two spec sheets disagree, compare internal volume and port tuning instead. Those are physical and they cannot be spun. A 12-litre box tuned to 52 Hz will not make usable 35 Hz output regardless of the copy, because below tuning a ported cabinet unloads, the acoustic damping disappears and cone excursion runs away.

Sensitivity carries the identical problem. Into 8 ohms, 2.83 V is 1 W, so for most home speakers the two reference conditions land on the same number. Into a nominal 4-ohm cabinet, 2.83 V is 2 W, and the printed figure flatters the speaker by 3 dB. Check which condition is quoted before you size an amplifier around it.

There is a get-out here worth knowing. Play a 31 Hz note through a cabinet that stops at 45 and you still hear the pitch, because the ear reconstructs a missing fundamental from the harmonic series above it. It works better than it has any right to, and it collapses at volume, because the harmonics that are doing the work are also the ones the woofer is straining to produce.

Anechoic numbers understate what you will hear, and the gap is not small.

Below the frequency where the room's longest dimension equals half a wavelength, a reasonably sealed room pressurises rather than propagating waves, and output rises. In a typical 4 by 5 metre lounge that transition sits somewhere around 35 to 45 Hz, and the lift below it runs roughly 3 to 6 dB per octave. A cabinet measuring -3 dB at 48 Hz on a test bench often reads close to flat at 40 Hz once it is in a real room.

Boundary reinforcement stacks on top. One wall behind the speaker adds about 3 dB in the bass, a wall plus the floor about 6, a corner about 9. That is free output, and it is also exactly why a speaker that sounded lean in a demo room turns boomy at home against a bookcase.

None of it helps above roughly 80 Hz, and none of it rescues a cabinet that has genuinely run out. Room gain lifts what is already there. It cannot lift zero.

Four rules cover almost every system.

• ▸Set the crossover at the speaker's -3 dB point plus 10 Hz, or at 80 Hz, whichever is higher.
• ▸For any compact bookshelf with a 100 or 130 mm woofer, use 100 to 120 Hz whatever the -3 dB figure says. The limiting factor there is excursion, not extension, and taking the bottom two octaves off it buys headroom you will hear immediately.
• ▸Set every channel to Small in the receiver, towers included. A subwoofer with its own amplifier and its own position in the room almost always does better below 80 Hz than mains asked to do it from wherever the imaging wants them.
• ▸Never cross above 120 Hz. Past that the subwoofer becomes locatable and you start hearing male voices coming out of a corner.

Crossover points are quoted as the -3 dB point of the filter, and in a bass-management chain the low pass is usually 24 dB per octave while the high pass is 12. Those slopes are designed to overlap, so the acoustic sum stays flat through the handover. If the room measures a dip right at 80 Hz, move the crossover point rather than pushing the subwoofer level up to fill it.

Towers earn their footprint in exactly four situations, and every one of them is physical rather than a matter of taste.

• ▸Rooms over about 35 square metres. The extra 3 to 5 dB of sensitivity and the larger total radiating area mean the speaker is not working hard to fill the space, and a driver that is not working hard distorts less.
• ▸Two-channel systems with no subwoofer and no plan to add one. This is the clearest case on the list: a tower with a -10 dB point at 28 Hz is doing a job nothing else in that room will do.
• ▸Installs with nowhere sensible for stands. A pair of 60 cm stands occupies the same floor footprint as a slim tower and always looks like it was added later.
• ▸Systems on a low-power amplifier, anything under about 30 watts per channel. Here the sensitivity gap matters far more than the extension gap, because 3 dB of sensitivity is worth a doubling of amplifier power.

Set against all that, a standmount on a proper stand plus a subwoofer beats a tower on placement flexibility every single time. The sub goes where the room supports it, the mains go where the imaging works, and those two positions are almost never the same square metre of floor. A tower forces them to be.

Get the tweeter to seated ear height and most of the argument between the formats evaporates.

Seated ear height on a normal sofa runs 95 to 110 cm off the floor. A bookshelf cabinet with its tweeter 25 cm up the front baffle therefore wants a 70 to 85 cm stand, which is taller than the 60 cm stands most shops keep in stock. Measure yours before ordering. A tweeter sitting 20 cm below your ears costs more in tonal balance than any cable ever will.

Fill hollow stands to about 80 percent with dry kiln-dried sand or steel shot, then stop. Packing them solid is worse, because it turns the column into one resonant mass instead of a damped one.

Spikes under a tower are about stability on carpet, not about draining vibration into the floor. On a suspended timber floor, spikes couple the cabinet directly to the joists and frequently make the bass worse; compliant feet or a slab of paving stone under the speaker usually measures better. On a concrete slab it makes very little difference either way.

Fit outriggers to anything over about 110 cm tall with a narrow baffle. A tall cabinet rocking two or three millimetres at 40 Hz smears its own output, and it is a tipping hazard in a house with children.

• ▸Floor-to-ear height at the seated listening position, to the nearest centimetre.
• ▸Distance from the front baffle to the wall behind it, with the speaker where you can actually live with it. Under 30 cm and a rear-ported cabinet will boom; sealed or front-ported is the answer, not a foam bung.
• ▸Room volume in cubic metres, which is just length times width times height. Under 40 and a compact cabinet plus a sub is plenty. Over 80 and you want the sensitivity a tower brings.
• ▸The lowest fundamental in your listening, taken from the source table above.

Those four numbers choose the cabinet class in about thirty seconds, and they do it without reference to price, brand or anybody's description of how a speaker sounds.