Short answer: for choosing a box, the numbers that matter most are Fs, Qts, Qes, Vas and Xmax. Divide Fs by Qes to get EBP. Roughly, under 50 leans sealed, over 100 leans ported, and in between it'll do either. Then put the real numbers into box-modelling software, because rules of thumb only get you to the right ballpark.
The three things a spec sheet tells you
It's easier if you split the sheet into three questions:
- How does it behave in a box? The Thiele-Small (T/S) parameters: Fs, Qts, Qes, Qms, Vas.
- How hard can it go? Xmax, Sd, power handling.
- How loud is it for a given input? Sensitivity and impedance.
The parameters, one by one
| Spec | What it is | Why you care |
|---|---|---|
| Fs (Hz) | Free-air resonance: the frequency the driver "wants" to move at, hanging in the air. | A rough floor for how low it plays. Lower Fs generally means deeper bass potential. |
| Qts | Total Q: how strongly the resonance is damped, mechanically and electrically combined. | The big one for box choice. Higher (around 0.4 and up) suits sealed; lower (around 0.2 to 0.4) suits ported. |
| Qes | Electrical Q: damping from the motor (magnet and voice coil). | Used for EBP. A strong motor gives a low Qes. |
| Qms | Mechanical Q: damping from the suspension. | Mostly feeds into Qts. You rarely use it directly. |
| Vas (litres) | The volume of air with the same springiness as the driver's suspension. | Bigger Vas usually means a bigger box. |
| Re (ohms) | DC resistance of the voice coil. | What your multimeter reads. Always lower than the nominal impedance. |
| Le (mH) | Voice-coil inductance. | Higher Le rolls off the top end sooner. Matters in crossover design. |
| Sd (cm²) | Effective cone area. | With Xmax, tells you how much air it can move. |
| Xmax (mm) | How far the cone can move one way while staying (reasonably) linear. | Excursion limits low-frequency output, especially in sealed boxes. |
| Mms (g) | Moving mass, including the air load. | Heavier cones go lower but need more motor to control. |
| Bl (T·m) | Motor strength. | Strong Bl relative to mass gives control and a lower Qes. |
| Sensitivity (dB) | Loudness at 1 m for a set input. | How loud it plays for a given input, not how loud it can go. See the catch below. |
Vd (displacement) is Sd × Xmax. It's handy for comparing how much air two subwoofers can shift, and for sizing a passive radiator (more on that in Sealed vs ported vs passive radiator).
EBP: the quick sealed-or-ported test
EBP (Efficiency Bandwidth Product) = Fs ÷ Qes.
It's a rule of thumb, but a useful one:
- Under about 50: tends to suit a sealed box.
- About 50 to 100: could go either way. Model both.
- Over about 100: tends to suit a ported box (or a passive radiator).
Look at Qts alongside it. A driver with low Qts and high EBP is built to be ported. A driver with higher Qts and low EBP is happier sealed.
Three real drivers, worked through
All figures from the manufacturers' own datasheets.
SB Acoustics Satori MW16P-4 (6.5" mid-woofer) Fs 28 Hz, Qes 0.28, Qts 0.27, Vas 50.2 L, Xmax 6 mm. EBP = 28 ÷ 0.28 = 100, with a low Qts of 0.27. Textbook ported driver. Sure enough, SB's own Ara kit puts it in a bass-reflex box tuned to 36.5 Hz. MW16P-4 on our site
SB Acoustics SB34NRXL75-8 (12" woofer) Fs 22 Hz, Qes 0.29, Qts 0.28, Vas 205 L, Xmax 10 mm. EBP = 22 ÷ 0.29 = about 76. That's in the "either" zone, but the low Qts nudges it towards ported, and SB's datasheet lists "optimized for vented enclosure" among its features. SB's Gema 3-way runs it in a big box with a slot port tuned to 28 Hz. Note the Vas: 205 litres says this one wants a large cabinet whichever way you go. SB34NRXL75-8 on our site
SB Audience BIANCO-18SW500 (18" subwoofer) Fs 30 Hz, Qes 0.53, Qts 0.52, Vas 334.8 L, Xmax 9.21 mm. EBP = 30 ÷ 0.53 = about 57 (the datasheet rounds it to 57), with a much higher Qts. This one is comfortable sealed, which is exactly how our Sealed 18SW500 build uses it: a 90 L sealed box with some DSP EQ to extend the bottom end. BIANCO-18SW500 on our site
The sensitivity catch: 2.83 V vs 1 W
Sensitivity is quoted two ways, and they aren't the same thing for every driver.
- 1 W / 1 m: one watt in, measured at a metre.
- 2.83 V / 1 m: 2.83 volts in. Into 8 ohms that's 1 W. Into 4 ohms it's 2 W.
So a 4-ohm driver quoted at 2.83 V looks about 3 dB louder than it would at 1 W, because it's getting twice the power. The MW16P-4 is a 4-ohm driver quoted at 90.5 dB (2.83 V), which is about 87.5 dB at 1 W. Neither number is wrong; just compare like with like.
Other things the numbers won't tell you straight
- Xmax isn't measured the same way by everyone. Some makers use the voice-coil overhang alone, some add a bit (SB Audience's footnote adds a third of the gap depth), and some quote a distortion-based figure. Comparing Xmax across brands is rough at best.
- Power handling is about heat, not volume. A driver rated at 500 W can still bottom out at a fraction of that in a small box at low frequencies. Excursion runs out first.
- T/S parameters have tolerances. Datasheets are measured on run-in samples, and your pair won't match the sheet exactly. For subs and ported boxes it can be worth measuring your own: a tool like the Dayton Audio DATS V3 measures T/S parameters and impedance.
- The frequency response graph is measured on a test baffle, not in your box. Treat it as a guide to the driver's character, not to what your speaker will measure.
Putting it to work
- Check Fs, Qts, Qes and Vas. Work out EBP.
- Decide sealed, ported or passive radiator.
- Model it in free software (WinISD, VituixCAD's enclosure tool) to choose a volume and see the response and the excursion at your target power.
- Check Xmax at the bottom of the range you care about. That's usually where a design runs out of steam.
Quick questions
What's a good Qts for a subwoofer? It depends on the box. Around 0.4 to 0.6 is comfortable sealed; around 0.2 to 0.4 tends to suit ported. Model it rather than going by the number alone.
Is a lower Fs always better? Not on its own. Low Fs with a tiny Xmax won't play loud down low. Look at Fs, Qts, Vas and Xmax together.
Why does my multimeter read 3.4 ohms on a 4-ohm driver? That's Re, the DC resistance. Nominal impedance is a rating for the lowest part of the impedance curve above the bass resonance (under the IEC standard, the minimum shouldn't drop below 80% of it). Re is always below it.
Do I need to measure T/S parameters myself? For a proven design, no. For your own sub or ported design, it's worth it, especially if you're building near the edge of what the driver can do.