Infrasonic bass below 20Hz: chasing the octave you feel, not hear

There's a moment in the 1977 War of the Worlds "Thunder Child" sequence, and again in the organ pedal on well-recorded Saint-Saëns, where the room stops sounding like it's playing music and starts behaving like weather. I've had that happen in exactly four rooms I can think of, out of hundreds I've sat in over the years. All four had something in common that had nothing to do with the amplifier or the speakers sitting in front of me. They had genuine, controlled output below 20Hz. This is infrasonic bass, and it's the one frequency band in high-end audio that money alone doesn't reliably buy.

Infrasonic, strictly, means below the nominal threshold of human hearing, conventionally cited as 20Hz by acousticians and standards bodies alike. The Wikipedia entry on infrasound is a reasonable primer if you want the physics without the marketing varnish. Below that line you're not really hearing a pitch any more, you're feeling pressure. Chest flutter. A sense that something large has entered the room. It's the frequency range of distant thunder, of a jet taking off three kilometres away, and, not coincidentally, of pipe organ 32-foot stops and a handful of film scores mixed by people who know exactly what they're doing to your sternum.

Why most systems never get there

The honest answer is that most subwoofers are specified generously and perform conservatively. A manufacturer will quote "extension to 18Hz" and mean it at some trivially low output level, often -10dB relative to the driver's midband, measured in an anechoic chamber with no room gain considered at all. In a real room you get help from boundary reinforcement below the first room mode, which is useful, but you also run into the limits of excursion and amplifier current delivery well before most drivers reach their rated low end at any meaningful SPL.

I've measured consumer subwoofers, perfectly competent ones in the $1,500 to $3,000 range, that are down 15 to 20dB by 20Hz despite a spec sheet claiming flat response to below that. That's not dishonesty exactly, it's the gap between a measurement condition and a listening condition. The SVS website's technical documentation is one of the more transparent examples in the category, showing in-room versus anechoic curves rather than hiding behind a single flattering number, which is more than most brands bother with.

Worth being specific about what "-10dB" actually costs you perceptually. A driver that's flat to, say, 90dB at 40Hz and has fallen to 80dB by 18Hz hasn't quietly tapered off, it's lost roughly three quarters of its perceived output at that frequency, because SPL is logarithmic and our pressure sense at these frequencies is already compressed compared with the midband. On paper the spec sheet still says "extension to 18Hz" because that's technically true. In the room, at reference level, you'll feel almost nothing happening below about 25Hz, and you'll assume the recording simply doesn't contain that content. Often it does. The sub just isn't delivering it at a level your body registers.

What it actually takes

Real infrasonic reproduction, output you can measure and feel at 16 to 18Hz without it falling off a cliff, generally needs three things working together: large cone area or multiple drivers to move enough air without excessive excursion, serious amplifier power with headroom well beyond what the sensitivity spec implies, and a sealed or heavily braced ported enclosure design that doesn't run out of linear travel before it runs out of signal. This is part of why the sealed-versus-ported argument changes character at the extreme low end; our piece on sealed versus ported subs in statement systems covers the trade-off in more detail, and the short version is that sealed designs tend to degrade more gracefully as you push past their comfortable extension point, which matters enormously once you're chasing single digits above 0Hz.

None of this is cheap, and it shouldn't be dressed up as more exotic than it is. It's a mechanical engineering problem: large diaphragm area, strong motor, stiff enclosure, enough current on tap. Statement subwoofers that genuinely deliver in this band tend to be heavy for a reason, often well over 50 kilograms for a single enclosure, because mass and bracing are doing real acoustic work, not just signalling price.

To put some rough numbers around it: a 38cm driver with 15mm of one-way linear excursion has considerably more low-frequency headroom than a 30cm driver with 10mm of travel, even before you touch amplifier power, simply because swept volume scales with the cube of the relationship between cone area and stroke. Push that same 38cm driver with an amplifier that can sustain 1,000 watts into a low impedance without clipping, rather than a nominally "500 watt" amp that's actually closer to 300 watts continuous once you measure it properly, and you start to see why genuine infrasonic performance tends to live in enclosures that are large, heavy, and not cheap to freight. There's no shortcut through clever DSP for this part. Equalisation can flatten a response curve that has headroom to flatten; it cannot manufacture excursion or current that isn't there.

Multiple subs matter more here than anywhere else

I'll say something that sits slightly against the grain of how multi-sub systems usually get discussed. Most of the case for multiple subwoofers is about smoothing room modes and evening out seat-to-seat response, and that's well covered territory; see one sub or two: when a second subwoofer actually earns its keep for the mechanics of that. But at the infrasonic extreme, multiple subs matter for a more basic reason: total displacement. Four modest subwoofers moving air in a room produce more true low-frequency output, with less individual strain and less distortion, than one large sub working flat out. It's addition of swept volume, not psychoacoustic trickery.

Run the arithmetic and it's fairly persuasive. Four subs each contributing, say, 400 cubic centimetres of linear swept volume at 18Hz will, combined and summed coherently at the listening position, produce meaningfully more low-frequency pressure than one sub trying to generate the equivalent displacement on its own, and each individual driver is working well inside its comfortable excursion limit rather than at the ragged edge where distortion and compression climb fast. Distortion at these frequencies is also harder to hear directly, which is exactly why it matters more, not less; a driver straining at 18Hz throws up harmonic content at 36Hz and 54Hz that muddies the midbass without you necessarily identifying the cause.

This is also where a lot of statement rooms go wrong. I've sat in systems with four enormous subwoofers that still don't reach convincingly below 20Hz, because they're all crossed too high, badly phased against each other, or fighting room modes that were never measured properly in the first place. Quantity of iron doesn't guarantee extension. Setup does the heavy lifting; the hardware just has to be capable of being set up correctly, which is a different and lower bar than most buyers assume.

The setup sequence that actually works, in my experience, is boring and sequential rather than clever. Measure each subwoofer individually at the primary seat first, with REW or equivalent, to establish its raw in-room response and group delay. Then set relative levels so no single sub dominates the sum. Then check phase and polarity between pairs, which on a lot of gear is a simple 0/180 switch rather than continuously variable, so you're sometimes choosing the less-bad of two options rather than a perfect alignment. Only after that do you touch EQ, and only to flatten what's already reasonably well summed, not to paper over a phase problem that EQ can't actually fix.

The room is the actual bottleneck

Here's the part that the gear reviews rarely spell out. Below about 20 to 25Hz, wavelengths are longer than most domestic rooms, which means you're not really dealing with a loudspeaker problem any more, you're dealing with room modes and standing waves that dominate everything the driver does. A room with dimensions that happen to reinforce, say, 16Hz will give you infrasonic output almost for free in one seat and almost none three metres away. This is precisely the territory our statement sub design piece and the broader bass management content on this site keep returning to: measurement beats intuition every time down here, because your ears are poor instruments for judging absolute level at frequencies this low. You feel pressure changes, but you can't reliably judge flatness by ear the way you can at 200Hz.

The wavelength at 18Hz is roughly 19 metres, which is simply longer than almost any residential room's longest dimension. That single fact explains most of what's strange about bass below 25Hz: there's no true far field in a domestic space at these frequencies, so the room's boundaries are doing as much or more work than the driver. A standing wave with a peak at the listening position can make an ordinary subwoofer sound extraordinary in that one seat, and a genuinely excellent subwoofer sound thin in the seat two metres over. I've had owners swear their system reaches 16Hz based entirely on where they happen to sit, and a measurement sweep tells a different story the moment you move the microphone.

Getting a genuine measurement means an SPL meter with an extended low-frequency calibration, or more realistically, miked measurement software run through room correction, taken at the listening position rather than near the sub. REW, the free room measurement software widely used in this community, will show you exactly where your infrasonic output actually sits rather than where the subwoofer's marketing claims it sits. I'd trust that over a spec sheet every single time, and I say that as someone who spends a fair bit of professional energy taking spec sheets seriously.

Does a statement system actually need this?

Honestly, for most music, no. The lowest fundamental on a double bass sits around 41Hz, a piano's lowest A is roughly 27.5Hz, and even a large pipe organ's 32-foot stops bottom out around 16 to 20Hz only on specific repertoire. If your listening diet is jazz, string quartets, most rock and most electronic music, chasing flat response to 16Hz is largely an engineering flex rather than a musical necessity, and I'd be lying if I pretended otherwise.

Where it earns its keep is home cinema, and specifically a fairly narrow slice of home cinema: films mixed with deliberate infrasonic content, which the Dolby speaker setup guidance touches on when discussing LFE channel handling, even though it stops short of specifying exact extension targets for consumer rooms. The LFE channel on a well-mastered disc or stream genuinely carries content down into the high teens in Hertz on specific effects, explosion sustain, certain creature sounds, the low-end rumble under dramatic reveals. Our bass management by the numbers piece gets into how crossover and slope settings interact with that content, which is worth reading before you spend serious money chasing extension your crossover point will filter out anyway.

There's a trade-off worth being blunt about here too. Pushing a system to deliver flat, undistorted output at 16 to 18Hz at reference level usually means accepting compromises elsewhere, a larger footprint than most lounge rooms comfortably accommodate, enclosures that need to sit against structural walls to avoid rattling cabinetry, and running costs in amplification that a lot of buyers underestimate. If your room is shared living space rather than a dedicated theatre, the WAF, for want of a better term, often rules this out before the acoustics do. That's a legitimate reason to stop at 25 or 30Hz and spend the remaining budget on screen, seating or acoustic treatment instead.

The practical verdict from the rooms I've actually heard this work

All four rooms I mentioned at the start had three things in common, not coincidentally: more than one subwoofer, genuine acoustic treatment addressing the specific room's modal behaviour rather than generic panels bought off a list, and an owner who'd actually measured the result with a microphone rather than trusting their ears or the spec sheet. None of them were the most expensive systems I've sat in front of this year. One of the four was in a converted garage in outer Melbourne running a pair of sealed subs that would retail for less than a flagship cartridge.

If you're building toward this, I'd treat it as the last ten per cent rather than the first spend. Get your mains and your primary sub crossover sorted, get the room's main modes under control, and only then start asking whether genuine sub-20Hz extension is worth the next outlay. For most listeners the honest answer is that a well-placed, well-integrated pair of subs reaching cleanly to 25 or 30Hz will do more for the music you actually listen to than chasing the last octave most recordings never use.

Still, there's something in that lowest band, felt rather than heard, that no amount of midrange precision replaces. It's not for everyone and it's not cheap to do properly. But when a room gets it right, you know within about four seconds of the first note, and that's a rarer experience in this industry than most marketing departments would like you to believe.

— Sofia Laurent, High-End & Statement Systems, Sound Technology