having maximally flat extension to a few Hz lower down (and that, rather than true bass quality, is unfortunately what sells loudspeakers).
      Conversely, the gentler the corner of the chosen filter, the better will be the transient response, with less overshoot and less prolonged ringing, which sonically yields better quality bass (tighter, better definition, better tunefulness, more transparent revelation of subsequent music, etc.). The gentler corner starts its rolloff at a slightly higher frequency, so it might not look as impressive on a frequency response graph, but it sounds a lot better.
      It's a truism among knowledgeable audio engineers that a more slowly changing (i.e. gently cornered) curve in the frequency domain means a quicker, more accurate response in the time domain - and, conversely, that a more quickly changing curve in the frequency domain means more sluggish, inaccurate response in the time domain. Since music signals actually occur in the time domain, and since we humans actually hear signals in the time domain, it is the time domain that should receive priority in being made as accurate as possible. We humans don't look at frequency response graphs while we listen to music. Indeed, the very concept of frequency is merely an artifice invented by us humans as an engineering convenience. And we humans invented this artifice as being the inverse of time (cycles per second), which of course is why faster change in one domain necessarily (by definition) implies more sluggish change in the other domain.
      These same guidelines and truisms apply to all kinds of filters. For example, a sharp high peak in the frequency response implies prolonged, high Q ringing in the time domain, whereas a gentle, broad hump in the frequency response implies merely some tonal balance effects, with no significant degradation of time domain accuracy.
      Which brings us back to that corner in the W1's response at 27 Hz. Since NHT can control the shape of that corner by purely electrical means in the X1 crossover, they have a wide choice of options. The best bass quality for the W1 could be achieved by making this electrical 27 Hz corner as a gentle corner, and having it gently transition from its progressive equalizing boost below 50 Hz into flat response below 27 Hz. This would have made the W1 into a pure second order bass system, with a gentle corner at 27 Hz, thereby giving it the best possible bass quality that a sealed bass system could achieve.
      Instead, NHT chose to have the X1's bass signal feed transition from boost (in the 50 Hz down to 27 Hz region) into a 12 dB per octave rolloff (rather than transitioning into flat response). This choice gives the W1 a subsonic filter, reducing signal drive below 27 Hz, which does have the benefit of reducing woofer excursion and stress and distortion for very low frequency signal information below 27 Hz. But this choice also introduces two new limitations. First, it reduces the W1's bass impact for bass transients, including music and film sound effects, since energy below 27 Hz is progressively reduced at the steeper rate of 24 dB per octave instead of just 12 dB per octave. Second, it degrades bass quality, since the corner at 27 Hz is inevitably sharper.
      Then, taking this NHT decision for added rolloff as a given, NHT still has a further choice as to just how sharp to make this corner at 27 Hz. The NHT decision for added rolloff below 27 Hz from the X1 makes the W1 into a fourth order bass system, instead of the second order bass system that it could be. But the question still remains of how sharp to make the corner at that 27 Hz point where the system response changes from flat (as equalized) into a fourth order rolloff slope. Here too NHT made a decision that does not afford the W1 all the bass quality it could have, for the corner shape chosen by NHT is quite sharp. This can be seen by simple inspection of the bass feed frequency response curve coming out of the X1 crossover and into your power amplifier that drives the W1. The overall shape of this X1 frequency response curve looks like a mountain peak. That's because the X1 progressively boosts the signal from 50 Hz going down to 27 Hz, and then below 27 Hz it changes to progressively rolling off the signal at progressively lower frequencies. So the X1's frequency response curve looks like a mountain with a peak at 27 Hz.
      But how sharp is this peak? Since the intrinsic response of the W1 at this 27 Hz point is already falling as a pretty straight line (at about 12 dB per octave), it does not contribute to the shape of the mountain peak at 27 Hz. Thus, the sharpness of that mountain peak at 27 Hz can easily be controlled by purely electrical means, in the electrical signal feed from the X1. NHT could therefore easily make this electrical circuit peak in the X1 as sharp or as gentle as they choose.
      NHT in fact chose to make this purely electrical mountain peak quite sharp in the X1 (indeed, the Q of the X1's low pass subsonic filter at 27 Hz is a high 2.2, and the initial rolloff slope on the left side of the mountain peak is much steeper than 12 dB per octave, which makes the peak all the sharper as a frequency response corner). This means that the corner shape of the W1's transition at 27 Hz, from equalized flat to rolloff response, is very sharp. And this in turn means that the bass quality and bass transient response of the W1 is further compromised, beyond what it needs to be. In other words, the sharp corner produced by the X1's sharp peak itself causes overshoot and ringing in the electrical signal fed to the W1 (via your power amplifier driving the W1), and this degrades the W1's bass quality. The W1 itself could intrinsically produce better quality bass if the electrical mountain peak in the X1 were less sharp. Note that the W1 itself has faultless transient response at this 27 Hz system response corner, since the W1's intrinsic response is a straight (sloping) line at this point (the W1's intrinsic corner is up at 50 Hz, nearly a full octave away). So the W1's compromised bass transient response at 27 Hz is entirely the fault of the electrical signal furnished by the X1.
      It's natural for us to intuitively think that bass overshoot and ringing (the primary culprits producing poor bass quality) are always caused by mechanical and acoustical phenomena, such as a too heavy woofer overshooting its mark and then reactively continuing to oscillate after the music's bass transient has ceased. But this bass overshoot and ringing could as easily be caused by a purely electrical filter that overshoots and rings in its transient response. The transient misbehavior of the mechanical and acoustical system with that heavy woofer can also be represented or modeled as a filter. And a filter is a filter, whether its root causes are purely electrical or are mechanical and acoustical - in both cases the math is the same and the sonic result is essentially the same.
      This lesson applies to the W1 and the driving electrical signal from the X1 crossover. Admittedly, there is some effect upon overall bass quality from the mechanical/acoustical Q of the W1 woofers in the sealed cabinet, up at their 50 Hz intrinsic resonance. But the primary determinant of the W1 system's bass quality is the nature of its primary frequency response corner down at 27 Hz. And that is wholly determined by the shape of the purely electrical filter in the X1. The fact, that this purely electrical filter in the X1 is shaped like a sharp mountain peak, causes overshoot and ringing in the bass transient response of the electrical signal fed from the X1 to your power amplifier, even before it reaches the W1. Any filter (whether electrical or mechanical/acoustical) with this sharp mountain peak shape would cause this same overshoot and ringing. So this purely electrical filter in the X1 is the primary factor determining the W1's bass quality (and degrading that bass quality below what the W1's intrinsic mechanical/acoustical nature itself could be capable of).
      It might be difficult to change the mechanical/acoustical properties of a woofer driver in an enclosure, but it is very easy to change the properties of a purely electrical filter. Thus, NHT could very easily change the purely electrical filter in the X1, to more fully realize the potential of the W1's intrinsic bass quality performance, and thereby improve the W1's bass quality beyond its present very good rating.
      Indeed, in the limiting case, the W1's bass quality and bass transient response could be taken to its maximum intrinsic capability by making the X1's mountain peak so gentle that it would not be a peak at all, but instead would be a gentle (low Q) transition to flat low frequency response, thereby making the whole X1 cum W1 subwoofer system a true second order bass system, with a gently cornered transition at 27 Hz from flat to 12 dB per octave rolloff response. The X1 could provide this much better quality bass from the W1 if it would simply include a back panel switch to allow the user the option of bypassing the X1's subsonic filter, with its transient-degrading high Q sharp corner at 27 Hz. Many other electronic components with a subsonic filter allow the user the option of defeating this subsonic filter, and the X1 should do likewise.

Upper Frequency Capability of W1

      We've been discussing the bass performance of the W1 (cum X1) at the low frequency end of its pass band. What about its performance at the upper end of its pass band? Many other subwoofers have problems at the upper end of their pass bands, problems which make it difficult to sonically integrate them with satellites, and problems which cause undesirable sonic colorations in the upper bass and/or warmth regions. The common vented enclosure subwoofers have particularly bad problems in these regards. But the W1, although a fourth order bass system like these other vented enclosure subwoofers, has here again a distinct advantage over these other fourth order systems.
      The chief problem, that other vented enclosure subwoofers have, is glaringly obvious if you simply stick a microphone into the vent or port, to measure the acoustic output of the vent itself. The output response from the vent itself has a large hump at the bottom end of the pass band, which is proper for its role as an extender of the woofer's bottom end. But, in virtually all vented enclosures, the output response from the vent also has a very high and very sharp peak way up around 300 Hz. That's sonically very bad news for all these other vented bass systems. This spurious output around 300 Hz intrudes upon the spectral domain of the satellites, making it difficult to properly integrate vented subwoofers with satellites. This spurious vent output adds too much energy to your overall system around 300 Hz.
      Moreover, because this vent response has the shape of such a sharp peak in the frequency domain, it necessarily has prolonged ringing in the time domain. This prolonged ringing will obscure the subsequent genuine music, and will make your system sound muddy, thick, and less transparent. This prolonged ringing also means that your ear/brain will hear this 300 Hz ringing as a foreign coloration, and that your ear/brain will be attracted to paying extra attention to it, because it lingers long after each triggering musical transient has fallen silent, and because its coloration (its one-note boom) is so obviously a foreign sound imposed upon and added to the musical content, an obnoxiously intrusive sound that is obviously not part of the music.
      Because of the extra attention that this foreign coloration commands, you can hear it and its degrading effects even if the amplitude of that 300 Hz peak is reduced until it is significantly below the peak or average music level. Indeed, even if its peak amplitude is reduced below the peak music level, and even if its average amplitude is reduced below the average music level, its lingering ringing might actually still be above the instantaneous music level in that time period of quiet music after the strong triggering music transient has fallen silent. Thus, with these vented subwoofers, you have to make sure that that nasty sounding 300 Hz sharp peak is reduced so far down that its obnoxious misbehavior is rendered passably inaudible. That's why vented subwoofers have to have their frequency response at the upper end of their pass band cut off so early (by 80 Hz or 100 Hz at the highest), and so very steeply.
      But that early and very steep cutoff creates further sonic problems with these vented subwoofers. First, you can't employ these subwoofers to deliver any warmth to your system. Many satellites do not have virtues noted above for the M6, and are too lean in the warmth region (100-300 Hz). To sonically integrate with these satellites, a subwoofer should be able to add some energy in this warmth region. But you can't use most vented subwoofers to do this, since their response must be cut off sharply before (below) the warmth region, in order to prevent their obnoxious peak at 300 Hz from intruding. Second, the subwoofer upper frequency cutoff (low pass filter) must be made very steep (to reduce the electrical drive to that nasty 300 Hz peak), and this steepness degrades your system's transient response, with a double whammy.
      The first whammy is that the electrical filter feeding your satellite should also therefore be made very steep, so that the two filter responses (the upper frequency cutoff for your subwoofer and the mating lower frequency cutoff for your satellite) dovetail properly. But this steep low frequency cutoff for your satellite will degrade its transient response at the bottom end of its warmth region reproduction, causing sonic muddiness from a one-note boom at around 100 Hz.
      This degraded transient response is not the satellite's fault, since the satellite intrinsically might have much better transient response around 100 Hz, but is instead the fault of the steep, sharp cornered electrical filter feeding the signal to the satellite. And that electrical filter fault is in turn the fault of the vented subwoofer having that nasty peak around 300 Hz in the output response of its vent.
      Here's another case where a prior electrical filter needlessly degrades the transient response of a loudspeaker below its intrinsic capability, in this case the transient response of your satellite around 100 Hz. But in this case the electrical filter must introduce this sonically degrading flaw, in order to properly dovetail with the subwoofer filter, which must be very steep in order to avoid the yet worse sonic flaws from the subwoofer vent's obnoxious peak around 300 Hz. One sonic evil in order to prevent a yet worse evil, but a sonic evil nevertheless, and a whammy caused ultimately by that subwoofer vent.
      The second whammy is that the required steepness, for the vented subwoofer's upper frequency cutoff, further degrades the bass transient response of the subwoofer itself. We know that the lower frequency rolloff slope of a vented subwoofer is already very steep (fourth order, 24 dB per octave). Now we find that the upper frequency cutoff slope for the vented subwoofer must likewise be made very steep. And this upper frequency cutoff must also be at a pretty low frequency (80 Hz, to 100 Hz at the highest), which is merely two octaves or less away from the low frequency rolloff corner (which is at 20 Hz or higher). This means that the subwoofer operates only over a narrow spectral region for its pass band.
      When we put these three factors together - a narrow pass band region, a steep slope at the bottom end, and a steep slope at the top end - we see that the total response of this vented subwoofer has the shape of a narrow, steep sided mountain. And that necessarily implies poor transient response, with overshoot and lingering ringing, for effectively the whole subwoofer range (not just for the corner frequency down at the bottom of its range). If the vented subwoofer could have a wider pass band, and if its upper frequency cutoff slope could be made more gentle, then it would have better transient response and better bass quality, with tighter bass, better definition, better tunefulness, less of the overhang that sounds slow and heavy and obscures subsequent music, etc. But the vented subwoofer can't have these desiderata, all because its vent puts out that nasty peak around 300 Hz that must at all costs be subdued.
      How does all this relate to the Evolution W1 subwoofer? The W1 has no vent. So it has none of these aforementioned, sonically distressing problems. The W1 can reach much higher than 80 Hz or 100 Hz, without spurious coloration, so it can supply some of the warmth region if you wish. Thus, you could use the W1 well up into the warmth region, to meet small satellite loudspeakers, or even to overlap with and supplement lean sounding satellites (not that the M6 has this problem). Even with the M6, you could offload some of the warmth region from the M6 to the W1 subwoofer, thereby lessening the excursions of the M6 woofers and lowering the modulation distortion of the M6 even further (although, as discussed above, the M6 is already very good in this regard, since its woofers cross over at a very low 350 Hz).
      Indeed, the manufacturer's measurements show that the Evolution subwoofer driver is extraordinarily well behaved past 500 Hz (though, having a metal cone, it doubtless has some peaky misbehavior at a yet higher frequency that you'd want to keep subdued). Of course, you wouldn't actually use the W1 up to 500 Hz, since satellites can easily reach below that, and also since the large W1 driver becomes directional at these upper reaches of its range, staying flat only on axis (and, for other reasons discussed below, you don't want to listen to the W1 on a woofer axis). The point is that the W1 is extraordinary flexible about the crossover frequency you choose for crossing it over to your satellites, and this flexibility can pay big sonic dividends in obtaining smooth, uncolored integration between subwoofer and satellite.
      Furthermore, since the W1 has no vent and no nasty peak around 300 Hz that must be reduced, the slope of its upper frequency cutoff could be gentle instead of steep. This gentler crossover slope could portend better transient response for both the subwoofer itself and for the satellite, as discussed above. Indeed, the W1 could theoretically be cut off at its upper frequency end with the gentlest possible first order (6 dB per octave) filter, since its response is so well behaved into the warmth region, and the mating low frequency rolloff for a satellite could likewise be rolled off at its low frequency end with this gentle first order slope as its dominant pole -- assuming that the satellite's own internal resonance were an octave or more below your chosen crossover frequency (which in turn could be quite high, given the W1's good upper frequency extension). This crossover tactic would yield the best possible transient response from both subwoofer and satellite.
      The particular upper frequency cutoff (low pass) filter chosen by NHT for the W1 is built into the X1 electronic crossover, and it happens to be a second order (12 dB per octave) filter, which is a middle ground between the very steep filter required for most vented subwoofers and the theoretical ideal of a first order filter. It's a good idea that NHT provides the second order cutoff slope in the X1, since you might want to choose a lower crossover frequency that is closer to the M6's internal bass resonance, but we'd like to see a switch option for choosing the more transient perfect first order cutoff slope as an alternative user choice.
      We see here that the W1 reaches higher in frequency, without adverse coloration, than most vented enclosure subwoofers, giving you extraordinary flexibility in mating it with various satellites. And we saw earlier that the M6 satellite monitor reaches lower with richer warmth and

(Continued on page 113)