Home-theater enthusiasts have long extolled the virtues of using two or more subwoofers. The advantages over a single sub are many, and include increased bass output with less compression, and better blend of the subs’ sound with the sound of the main speakers (that is, it’s more difficult to identify the subwoofers as the sources of the lowest frequencies). However, the biggest advantage of using two subs in a home-theater system is the mitigation of bass peaks and nulls, to produce a smoother, more consistent overall bass response audible at a greater number of listening positions.
Are multiple subs also advantageous for the lonely enthusiast of two-channel listening? I can optimize the position of a single sub for the best bass output at the only listening position that matters, and the only place in my listening room occupied by a chair: the sweet spot. Is it worth my while to get and position a second sub for a room optimized for a single person listening only to two-channel stereo?
Before I answer that question, based on my recent experience of adding a second high-quality sub to my two-channel system, here’s a brief account of the journey that led me to ask it.
Diego Estan
For most of my 30-odd years as an audio enthusiast, I’ve used three-way tower speakers in a dedicated 15’L x 12’W listening room with a single listening chair and no subwoofer. A good three- or more-way tower has two advantages over a two-way minimonitor: greater bass extension, slam, and overall output, due to the floorstander’s larger internal volume and dedicated woofers. Midrange clarity and detail can also be better, as a three-way’s midrange driver isn’t burdened with having to produce the extreme excursions needed to reproduce the lowest frequencies of the audioband, as in a two-way. However, I’ve found that, in general, good two-way bookshelf designs can outclass their bigger brothers in imaging and coherence, probably due to their slimmer cabinets, and their smaller number of drivers.
Once I got into towers, I always felt I’d achieved satisfying bass, but I didn’t know if that bass was accurate -- in other words, that the speakers’ overall frequency response was relatively smooth or flat. Was there a significant peak at 50Hz and a dip at 40Hz? I didn’t know. Ignorance was bliss.
Only a short time ago, I decided to experiment with a 2.1-channel satellite-subwoofer system. I fell in love with and bought a pair of Bowers & Wilkins 705 S2 bookshelf speakers and coupled them to an SVS SB-4000 subwoofer. I integrated the outputs of B&Ws and sub as best I could, but the results weren’t perfect. I tried the sub in four different positions, took rudimentary measurements with an SPL meter and bass test tracks at each position, and settled on a spot next to the left speaker and exactly a third of the way along the long wall. Following SVS’s recommendations, I set the SB-4000’s low-pass filter (LPF), or crossover frequency, to 50Hz, with a 24dB/octave (fourth-order) slope. This meant that above 50Hz, the subwoofer’s contribution would be significantly reduced (at a rate of 24dB per octave), and below 50Hz, it would reproduce all frequencies down to its low-frequency limit of about 10Hz. The B&Ws, however, would continue to operate full range, reproducing the entire audioband down to about 33Hz, where their output rapidly rolls off in my room. I tweaked the SVS’s built-in three-band parametric equalizer and volume setting to flatten the bass response as best I could. The results were pretty good -- I was hearing the B&W 705 S2s’ full output unaltered, with additional LF output down to below 16Hz courtesy the SVS SB-4000.
But I felt I could do better. My two main complaints were the B&Ws’ hot tweeters, which occasionally made themselves known with hot recordings, and a lack of punch in the bass despite the addition of the sub. My next step was to implement electronic room correction with Dirac Live software, built into a miniDSP DDRC-22D ($725 USD) I’d inserted in the signal chain between my digital sources and DAC. The miniDSP had come with a calibrated UMIK-1 microphone. I now began to get serious about measuring.
miniDSP DDRC-22D
My first calibration with Dirac Live was a revelation. Using the same 50Hz crossover frequency for the SVS SB-4000 and playing the B&Ws full-range, I chose a Harman International frequency-response target curve and loaded that into Dirac Live’s calibration tool. I used the Harman curve because their research has shown that it’s what most listeners prefer (the relative amount of bass boost in the curve is more a matter of taste). My target curve can be described as +5dB from 16 to 50Hz, then a steep slope down to +1dB at 200Hz, then 0dB at 1kHz, then a gentle slope down to -2dB at 16kHz (see graph).
Everything sounded better. The bass I now heard was extended, tight, and detailed, never bloated or boomy. But it still wasn’t the last word in slam, impact, and punch -- the type of bass that thumps my chest with authority.
To get more slam, I shifted the SB-4000’s crossover higher in frequency, then recalibrated with Dirac Live, trusting the software to generate DSP filters that would blend the outputs of the SB-4000 and the 705 S2s, which now would significantly overlap. I kept the B&Ws running full-range because I had no component with high-pass bass management.
First, I set the crossover to 80Hz. Things were way better -- I could feel bass thumping my chest. I lived with that for a while, but couldn’t leave it there -- I love to tweak. Besides, my original measurements had told me that my main speakers’ biggest bass suckout at my listening position was at 130Hz -- a nasty, almost bottomless dip of 20dB. There wasn’t much I could do about this. In setting up a pair of speakers for two-channel listening, they first must be optimized for imaging and stereo separation -- only then can you try to mitigate bass-response anomalies, of which almost every room has some. I wasn’t about to move my 705 S2s -- that would compromise their imaging -- so I was stuck with a big bass null at 130Hz. Or was I?
I’d positioned the SVS SB-4000 to optimize its reproduction of the low frequencies, but not the B&Ws. What if I raised the LPF crossover frequency to 130Hz, to fill in the bass null? That seemed a very high crossover frequency, but as the sub sat next to my left speaker in the front soundstage, I wasn’t too concerned about the localization of bass. The consensus is that while bass below 80Hz is nondirectional, the direction from which bass frequencies higher than 80Hz are coming can be detected by a pair of human ears -- because, by definition, wavelengths of higher-frequency soundwaves are considerably shorter than those of the lowest-frequency waves, it’s tough to blend them with the even higher and thus more easily located frequencies output by the satellite speakers. Since my sub was relatively close to the left speaker, even if I could begin to detect the source of the bass, it would still sound as if it was coming from the general vicinity of the left speaker (compared, say, with placing the sub in a rear corner of the room but keeping the same high crossover frequency). Nine more Dirac Live measurements later, I was ready to hear the improvements. WOW! Now I had bass slam in spades!
A Bowers & Wilkins 705 S2 speaker and an SVS SB-4000 subwoofer
I wasn’t done tweaking. Still nagging at me was the lack of a high-pass filter (HPF) for the B&Ws, which I was still running full-range. If nothing else, I was wasting amplifier power by driving the speakers harder than I needed to -- and I was sure they were reproducing more bass than they or my room or my sub required. From Amazon, I bought a Behringer Super-X Pro CX2310 ($99) professional-grade, stereo, two-way, 24dB/octave active crossover. I kept the SB-4000’s LPF at 130Hz, and inserted the Behringer between preamp and amp to act as an HPF. Just as the SVS SB-4000’s built-in LPF would attenuate all incoming frequencies above 130Hz, the Behringer would attenuate all frequencies below the selected HPF before they were sent to the power amp driving the B&Ws. However, to make sure I cured my room of bass suckout, I wanted a bit of overlap -- the sub didn’t exhibit the 130Hz null to the same degree as the mains, but it did reveal a milder, 125Hz null. I set the Behringer to 120Hz.
After yet another Dirac Live calibration -- I was getting good at these -- I heard still more improvement in the bass. As the sub had now assumed all bass duties from 16 to 120Hz, there was even better slam and punch. As a bonus, my 705 S2s’ midrange detail and clarity had also improved, I assume because I was dumping less amplifier power into them, which in turn reduced each speaker’s cone excursions, cabinet vibrations, and port chuffing. (Note: The Behringer crossover, though an excellent tool for these experiments, was a stopgap. I’ve since replaced it with a more audiophile-approved, noiseless, passive, line-level, balanced HPF filter from Marchand Electronics, stuffed with high-quality inductors and polypropylene caps: the XM446XLR-A ($495), which I ordered with a fixed HPF of 120Hz.)
This OCD audiophile was still not entirely satisfied. I sometimes felt I could hear the source of the bass frequencies -- the bass pulled slightly toward the left. In fact, when Doug Schneider visited for a listen, he commented on this slight asymmetry in my system’s bass output, likely a result of that high 130Hz crossover frequency.
This was when I ordered a second SVS SB-4000 subwoofer.
I placed the second SB-4000 in a position that mirrored that of the first: next to the right-channel speaker, a third of the way along the long wall as measured from the opposite wall. Now I had two subwoofers positioned symmetrically in relation to the room and the main speakers.
Two SVS SB-4000 subwoofers flanking a McIntosh Laboratory MC302 amplifier and a Behringer Super-X Pro CX2310 crossover
I listened with one and with two subs, and with the crossover frequency set to 130Hz, as well as to the more typical 80 and 50Hz. In each case, an identical HPF crossover frequency was used with a matching 24dB/octave slope, using the Behringer active crossover. (I couldn’t use the Marchand XM446XLR-A at 80 and 50Hz, as its HPF is fixed at 120Hz.) I performed a full Dirac Live calibration for each of my six listening tests, and matched the volume levels for each session, in which I listened to three tracks, all 16-bit/44.1kHz FLACs ripped from CDs: “Hotel California,” from the Eagles’ Hell Freezes Over (Geffen); “Like a Stone,” from Audioslave’s Audioslave (Epic); and “I Feel It Coming,” from The Weeknd’s Starboy (Republic).
First, with the crossover at 130Hz, although the bass volume and extension sounded and felt the same, I heard/felt a definite difference in bass quality with two subs: it was punchier. However, the most important improvement provided by two subs was not in bass punch, but in the sense that the bass now seemed to come from everywhere. It seemed more even, more omnipresent.
With the crossover set to 80Hz, the difference was far more subtle. There was no difference in bass extension or punch, but I felt that two subs again created a more even bass response. A single sub very subtly pulled everything to the left.
Still, I’m aware that very subtle perceived differences in sound can be due as much to listener bias as to the sound itself. It was time to recruit my wife, who by now has gotten used to this. She was a good candidate -- a fan of hip-hop, she loves bass. More important, a glance at the tangle of wires snaking to and from the subs, crossover, and amp didn’t tell her if one or two subs was playing. She made a pretty good “single-blind” test possible.
I cued up The Weeknd’s “I Feel It Coming” with one sub playing, using the 80Hz crossover, asked her to focus on the bass, then asked her to keep her eyes closed while I connected the second sub and changed the Dirac Live filter. Her impression was the same as mine had been: She felt that the bass was slightly more even with two subs, but that the difference was extremely subtle. This wasn’t a double-blind test with ten trials, but it did support my own initial observations. With the 50Hz crossover, I heard no difference in bass quality or bass directionality -- and since I’d performed the test sighted and still heard nothing, I didn’t pester my wife to sit and listen again.
Conclusions
Is it worth buying a second subwoofer for a single listener? It depends. Even at the high 130Hz crossover frequency, the differences, though appreciable, weren’t significant. And because a single SVS SB-4000 sub delivered enough output for my room, the second SB-4000 didn’t provide any more headroom than I actually needed. (The situation may be different for smaller, less powerful subs in bigger rooms.) With the 80Hz crossover the difference was barely noticeable -- although, to an audiophile, “barely noticeable” can still be significant -- consider how many audiophiles labor over the differences they hear between cables. And with a crossover frequency of 50Hz, the advantages were nonexistent.
My advice: If you have a two-channel system with a single-seat sweet spot, spend all your subwoofer budget on a single sub of the best quality that your budget will permit, rather than two subs of lesser quality. If the crossover frequency is 80Hz or higher, then, down the road, if you’re looking for the ultimate in low-end performance, you might consider buying a second sub, as I did. The improvement wasn’t night-and-day, but I don’t regret the expense.
. . . Diego Estan
diego@soundstagenetwork.com