Reference

It is an unfortunate fact that hysteresis problems are not restricted to the large-amplitude movements found in bass drivers. They can also affect subjective timing in the mid- and treble ranges, though not to the same degree. Analysis of many speaker reports suggests that tweeters with foam or fabric suspensions time better than those using vinyl or similar materials.

In addition, the potentially anomalous mechanical behavior of ferrofluid has to be balanced against its benefit in terms of reduced voice-coil heating. The ferocity of such vibrations---small in amplitude but very fast---in a tweeter can hardly be underestimated. Under strong drive, mounting screws can be shaken loose, leadout wires fractured due to metal fatigue, and the ferrofluid in the gap can look as if it is boiling, the agitation sufficient to turn some of the ferrofluid into a magnetic mist floating between the poles.

Experience has shown that any kind of random movement in loudspeakers conspires against dynamic expression and good rhythm. It is as if the noises produced by such randomness are so out of context that the musical message is disturbed. Under heavy drive, loose absorbent fillings can flop about more or less randomly. That very randomness disturbs the continuity of the bass rhythm. UK critic James Michael Hughes has even recommended removing all of a speaker's filling material as a matter of course.

My vote is still out concerning enclosure fillings. There is almost always a conflict between the need for good volume absorption to trap those internal standing waves which are capable of causing coloration, and the need for clean, fast bass. Ideally, the air trapped in the speaker chamber should be subject to adiabatic pressure changes, a rapid if ultimately non-linear phenomenon. Dense enclosure fillings, however, are often touted for their ability to provide an approximation of isothermal operation, in which the enclosure's effective volume is increased by 20-30%. Here the filling absorbs the natural temperature changes in the air volume despite its pressure drive from the bass unit. (The constant temperature which results gives rise to the term "isothermal.") The penalty is the slower transient response due to the small but significant delays present in the transition from adiabatic to isothermal operation. It takes just a little time for the air to give up its heat, then get it back.

The random position of the radiating unit's driving point also blurs the music's rhythmic content. From my experience of many examples, I have observed a clear association between good subjective dynamics and such fundamental aspects of speaker construction as the integrity of drive-unit frames, the strength of their fitting systems, and the strength and mass of the enclosures to which the drive-units are attached. These things alone will not make a good speaker, but they are essential if the other good aspects of design are to be heard to full effect.

Many reviewers and owners have heard hysteretic randomness in action when they experience the improvement wrought by judicious tightening of the fixing screws for loudspeaker drivers. This needs to be done with some care; it is only too easy to strip threads in wood-composition baffles where steel "T" nuts are not used. Overtightening can also fracture the driver frames. After manufacture, wood or wood-composition materials accept a degree of compression; speaker fixings can loosen with time, even in the interval between manufacture and delivery.

The sound-quality difference between a moderate tightness, which allows some random time-delayed vibration patterns between the driver frame and the cabinet, and that when fully tightened, can be quite dramatic in dynamic and rhythmic terms.

Loudspeaker sensitivity & dynamics
Another way of improving a loudspeaker's dynamics and rhythm, provided that care is taken to maintain a low level of energy storage, is to raise its sensitivity. This is achieved by using a larger, lighter cone which, being required to move less, needs much less electrical current in its motor coil to achieve the same spl. Less current leads to lower levels of distortion, as well as less compression and less heating---other enemies of good dynamics. One high-performance miniature I know well achieves quite good dynamics by virtue of a highly accurate acoustic energy response, in terms of both wave front and decay. However, it can never reach beyond this point due to its low efficiency and the corresponding need to drive it with comparatively high currents. With an alternative design with similar-sized working parts, the bass response has been deliberately sacrificed in favor of higher sensitivity (no less than 8dB more). In addition, the use of a much lighter cone enables the design to achieve its chosen performance characteristic: near-effortless dynamics.