Chart: a sub at full drive vs the same sub EQ'd flat to 20 Hz

Why I don't tune subs "flat to 20 Hz"

I get asked this a lot: "I want my sub to play flat to 20 Hz." It sounds like the right goal. But with a big vented sub, making it flat to 20 Hz doesn't add any deep bass. It just turns everything above 20 Hz down. Here's why, and why your room usually fixes the bottom end for you.

The short version
  • A sub's deepest note is set by the driver, the box and the amp. EQ can't push it any lower.
  • "Flat to 20 Hz" means turning 30 to 80 Hz down to match 20 Hz. You give up loudness and get nothing back.
  • Your room adds its own deep bass. A sub that rises gently towards 80 Hz usually ends up close to flat once it's in a room.
  • So I let the sub play its natural shape and only EQ what's left, measured in the room.
  • The end result: both ways end up flat in the room, but my way is about 5 dB louder. Same sub, same amp.

What a sub actually does, flat out

The blue line below is a real simulation, not a sketch: an 18-inch sub driver in a 210-litre vented box tuned to 20 Hz, on one 440 W amp channel, turned up as far as it can go before something runs out of puff. No EQ. That's its natural shape: it's a bit quieter at 20 Hz and rises gently towards 80 Hz.

The orange dashed line is the same sub EQ'd "flat to 20 Hz".

Chart: the sub at full drive rises from about 113 dB at 20 Hz to about 122 dB at 80 Hz. EQ'd flat to 20 Hz it stays at about 113 dB all the way up, so the shaded area above is output thrown away.
One sub at 1 m, outside a room. Blue: full drive, no EQ. Orange: the same sub EQ'd flat to 20 Hz. The shaded area is the output you give up.

Look at 20 Hz: both lines are at the same spot. The amp is already flat out there, so EQ can't lift it. All "flat to 20 Hz" does is pull the rest down to meet it: about 3 dB less at 30 Hz, 5 dB less at 40 Hz and 9 dB less at 80 Hz. Nine dB is a lot. It's most of the punch in a kick drum or an explosion.

In other words, we're not extending the sub deeper. We're turning down the top end, for no reason.

Your room fills in the bottom

Here's the bit most people don't know. Once the sound waves get longer than the room can hold, the sub stops "throwing" bass at you and starts pressurising the whole room instead, a bit like pushing on the air in a sealed car. The deeper the note, the more it builds up. That's called room gain, and it lands right where a vented sub is quietest.

Chart: the sub on its own rises from about 113 dB at 20 Hz to 122 dB at 80 Hz. In a well-sealed 6 m room, the estimate lifts 20 Hz to about 120 dB, so the line is close to flat. A leakier room lifts it less, to about 116 dB.
The same sub, no EQ, on its own (blue) and in a 6 m long room: well sealed (orange) and leakier (green dashed). The room lines are estimates.

In a well-sealed 6 m room, this sub goes from sloping up to nearly flat, with no EQ at all, and it's still running at full drive everywhere.

The end result

So what do you actually end up with? Here's the same sub, set up both ways, in the same 6 m room. Then room EQ does its job and trims each one flat. Room EQ should only cut: the amp is already flat out at 20 Hz, so there's nothing left to boost with.

Two charts. Top: in the room before room EQ, the natural sub is close to flat, about 118 to 122 dB; the sub pre-EQ'd flat to 20 Hz now rises 7 dB towards 20 Hz. Bottom: after room EQ trims both flat, the natural sub sits at about 118 dB and the pre-EQ'd sub at about 113 dB, 5 dB quieter.
The same sub in the same 6 m room (well sealed, estimate). Top: before room EQ. Bottom: after room EQ trims each one flat from 20 to 80 Hz.
  • Pre-EQ'd "flat to 20 Hz": the room lifts the bottom on top of the EQ, so it isn't flat any more. It's about 7 dB heavier at 20 Hz than at 80 Hz: boomy. Room EQ has to cut it all back down to about 113 dB.
  • My way: the room fills in the bottom of the natural slope, so it's already close to flat. Room EQ only trims it to about 118 dB.
  • Both end up flat. Mine is about 5 dB louder from 20 to 80 Hz, with the same sub and the same amp. In a leakier room the gap is smaller, about 3.5 dB, but it's still there.

Five dB is roughly what you'd get from adding a second, identical sub and amp. Getting it for free just by not pre-EQ'ing is the whole point.

What I'd do instead

  • Size the box and tuning for the driver and the room, not for a flat line on a chart.
  • Let the sub play its natural shape. Don't pre-EQ it flat before it's even in the room.
  • Measure in the room, at your seat, then trim the peaks with room EQ. Cutting peaks costs nothing; boosting dips burns amp power and cone travel.
  • Two subs beat one. They give about 6 dB more and smooth out the peaks and dips from seat to seat.
  • Want more at 20 Hz? That takes more cone area, more box or more amp. EQ can only shape what the sub can already do.

If you're not sure what that means for your build, send me your room size, amp and what you want it to do (sales@nbaudio.com.au), and I'll give you my suggestion. Advice on picking parts is free.

For the technically minded: the numbers behind it

Where room gain starts

Room gain kicks in below roughly the frequency whose half-wavelength fits the room's longest dimension: f ≈ c / 2L, where c is the speed of sound (343 m/s) and L is the room's length in metres. A 6 m room starts at about 29 Hz; a 4 m room at about 43 Hz; a 9 m room at about 19 Hz. Below that, the sub can no longer radiate as a free wave; it compresses the air in the room and pressure builds.

Chart: estimated room gain for 4, 6 and 9 m rooms. Gain is small above the start frequency and rises steeply below it; the 4 m room starts highest and gains the most.
Estimated room gain in a well-sealed room, by room length. Smaller rooms start higher and gain more.

How much

In a perfectly sealed room the gain tends towards 12 dB per octave below that point. Real rooms leak through doors, windows, open plan and flexing plasterboard, so most get less, and some open-plan spaces get very little. I model it as 10·log₁₀(1 + (f₀/f)⁴) for a sealed room (3 dB at f₀, heading to 12 dB/octave), and half that in dB as a rough leakier room (heading to 6 dB/octave).

6 m room, this sub 20 Hz 25 Hz 30 Hz 40 Hz 80 Hz
Sub on its own, full drive 112.7 114.2 115.6 118.0 121.8
"Flat to 20 Hz" cut 0 −1.4 −2.9 −5.3 −9.1
Room gain, sealed (estimate) +7.1 +4.3 +2.6 +1.0 +0.1
Room gain, leakier (estimate) +3.6 +2.2 +1.3 +0.5 0
Sub in a sealed room 119.9 118.5 118.2 119.0 121.9
"Flat to 20 Hz" sub in a sealed room 119.9 117.1 115.3 113.7 112.8
End result after room EQ: my way 118.2 118.2 118.2 118.2 118.2
End result after room EQ: "flat to 20 Hz" 112.8 112.8 112.8 112.8 112.8

dB SPL, one sub at 1 m (room rows are estimates). Two subs add about 6 dB.

How the sub was simulated

  • Driver: an 18-inch sub driver (SB Audience ROSSO-18SW1000D), 210 L net, vented, tuned to 20 Hz.
  • Amp: one 440 W channel into 8 Ω. The drive level is the highest one where nothing runs out between 20 and 80 Hz: amp power, the driver's power rating, cone travel (Xmax) and vent air speed (17 m/s). Here the amp runs out first, at 20 Hz.
  • A standard vented-box simulation from the driver's Thiele-Small parameters. Free-field, 1 m, no EQ, no room.
  • "Flat to 20 Hz" is the same curve with everything above 20 Hz cut to the 20 Hz level, which is exactly what an EQ would have to do.

What room gain doesn't do

  • Above the room-gain region (roughly 30 to 120 Hz in a living room) you get room modes: big peaks and dips that change from seat to seat. That's what room EQ and a second sub are for.
  • The room-gain curve is an estimate. Your own measurement at the listening seat is the real answer.

Related

Back to blog