Qtc: system Q in a sealed box

Sealing a driver in a box adds the trapped air's springiness to the suspension's. Both the resonance and the damping rise together:

\alpha = \frac{V_{as}}{V_b}, \qquad f_c = f_s\sqrt{1+\alpha}, \qquad Q_{tc} = Q_{ts}\sqrt{1+\alpha}

Qtc is the single number that determines the shape of a sealed box's response. Once you know it, you know the curve:

|G(f)| = \frac{(f/f_c)^2}{\sqrt{\left(1-(f/f_c)^2\right)^2 + \left(\frac{f/f_c}{Q_{tc}}\right)^2}}

The landmark values

• 0.5: critically damped. No peak at all, the gentlest possible settling, and the earliest rolloff. See overdamped.
• 0.577: Bessel. Maximally flat group delay rather than maximally flat response: the best transient behavior of any alignment.
• 0.707: Butterworth. Maximally flat response, and exactly −3.01 dB at fc. The usual default, and a genuinely good one.
• 1.0 and above: a visible peak before the rolloff, and audible overhang. See peaking.

The trade-off you are actually making

Because both f_c and Q_{tc} scale with \sqrt{1+\alpha}, you cannot adjust one without the other. A smaller box raises Q and raises cutoff: it makes the bass both peakier and shallower. There is no box size that gives a low cutoff and a low Q from a driver that does not already have a low Qts.

That is the whole design decision for a sealed box, and it is settled before you cut any wood, by the driver.

Stuffing

Filling the box with fibre does two things. It slows the effective speed of sound, making the box behave as though it were 10–20% larger, and it absorbs internal reflections. The first effect lowers both fc and Qtc slightly; Baffle models it as a volume multiplier. It is a modest adjustment, not a way to rescue a box that is fundamentally too small.