Last week, feeling relaxed from a Florida golfing trip I gazed over the five subwoofers in the middle of my HT room and pondered whether my wife was right in questioning why I needed to buy another 3 subs. After all I had just spent two months experimenting with room treatments and building basstraps which greatly improved acoustics so why more subs she asked. The 2' x 7 ' basstraps in particular, built with 4" 12lb density roxul mineral wool had yielded a coefficient of absorption of 0.63 at 125hz so there was reason to smile. However, listening to an Axiom ep600 and a SVS PC-Ultra at reference levels does much to reveal their strengths and weaknesses. My solution: add two more subs for a quad sub configuration to increase headroom, reduce power compression, and smooth freaquency response across all the seats. If I kept the Ultra it would also lessen the demands on that sub and eliminate the annoying bottoming I had been experiencing. So I ordered another ep600 and two ep500s with the promise to sell one of my five subs to my brother. And to make calibration easier I had a Velodyne SMS-1 delivered.
Phase issues were a paramount concern especially when mixing different subs. If you have phase problems which is likely with different subs you will not get the same FR and headroom advantage as 4 identical subs. A 180 degree phase differential will create total cancellation at that frequency, but even a 45-90 degree differential will really screw with the in-room FR and will decrease headroom (the whole goal of buying more subs). I placed all subwoofers together first, at the same gain and phase settings. One ep600 was the control sub (A). I then tested each subwoofer against the control unit. I had a 6 dB increase when I connected the second ep600. With the two ep500s also connected I measured a 12db (6db+3db+3db)) increase over reference as set at control sub A. All was good so far.
This procedure will immediately reveal if one of your subwoofers was wired reverse polarity (even if the phase control is set the same as the others) at the factory. Reverse polarity can come from three things:
1) the woofer VC was wired backwards
2) the amp polarity control was wired backwards
3) the woofer leads were connected backwards
Having obtained the desired increase in headroom and established that the four subwoofers had the same phase/polarity, I then proceeded to the next stage of actually placing the subs.
Here is where the SMS-1 started to prove its worth right away. Rather than go through the time honoured but undignified crawling method, I placed the control sub at the prime seat and moved the SMS microphone around the room quickly locating optimal spots on the left mid sidewall, front wall between the left main and the centre and righ mid sidewall. For my first configuration I decided to preserve as much of the maximum 12db incremental headroom as I could by placing two ep600s on the left sidewall , and the ep500s on either side of the centre speaker, all connected through XLR outputs. I also thought that frequency response would be smoothest here though phase issues needed to be checked. With a setting at -7 from reference 75db I was registering a 9.5-10.5 of the maximum 12db possible increase in headroom from the quad Axioms subs. Pretty good.
Early in the process I had wanted to ensure that the 500/600 combination would not cause unecessary phase problems. Those two subs are essentially identical except for size of the cabinet enclosure and a 600w digital amp in the ep600 and a 500w amp in the ep500. To address my concern, I asked Axiom to run FR sweeps of these two subs combined in their new anechoic chamber. Ian and Tom ran some sweeps which showed no problem with phase between those two excellent subs oweing to their similar design. So no apparent phase problems with blending the 500 and 600 subs in a quad setup.
Two challenges for any multiple sub setup consist of maintaining phase integrity firstly between the subs and secondly between the mains. In my HT where I run deep extension mains with output measured down to 11hz, the mains are set to large which eliminates the step of phase integration with the front speakers. The subs are essentially LFE only with some redirected bass from the surrounds. An added benefit of this setup should be less IM distortion compared to mixing redirected bass from the mains with LFE though this is debatable as several people I have spoken to suggest there should be no difference in distortion if the redirection is done digitally at the processor. In any event running the mains full range allowed me to concentrate on getting proper phasing betwen the subs thereby significantly simplifying problems with phase integration of multiple subs with the mains. My research indicates phase adjustments are best accomplished with the delay setting in the processor and setting phase on the subs at either 0 or 180 deg. Analog variable phase controls are less useful because bandwidth has continuously varying delay with frequency, so no single setting can phase align across the band.
The SMS-1 though features digital phase alignment by 15 deg increments So after setting phase at the subs to obtain the highest SPLs, I next looked to the SMS for further phase adjustments. Phase was further tweaked by running 15hz to 200hz sweeps and adjusting phase by 15deg increments while looking at the FR graph produced by the SMS. None were necessary and the phase setting at the SMS remained at 0. As a final phase adjustment the subs were set to the same delay as the centre speaker plus 3 ms to account for routing through the SMS . After this tweaking but before parametric equalization the following FR response was shown at the SMS display. Note that based on my measurements the visual graph appears + 3hz inaccuate at frequencies less than 35hz. In other words it should be shifted 3hz to the left. Hopefully the next firmware upgrade which will extend the range to 5hz will correct this anomaly.
Fifteen minutes of tweaking frequency, level and Q for each of the eight equalizers resulted in this FR graph for the quad Axiom subs. The SMS has a subsonic filter at 15 hz whose slope can be adjusted for increased output below that frequency. Rather than leave the slope at a steep -24db, I adjusted it to -6db per octave for more infrasonic response though this adjustment did little to improve extension beyond the DSP limited 16hz in the 600s and 19 hz in the 500s.
Next came the listening tests starting with Jack Bruce's bass lines on Spoonful from Cream Live at Royal Albert Hall DVD. His Gibson bass guitar sounded as clean and refined as one could hope. The usual clips from movies such as WOTW, FOTP and M&C sounded as good as I have heard. All was good...until my brother reminded me that I had promised him a sub.
Before the holidays I had promised him the Ultra but a single sub was not going to perform well in his 6000cf space especially because he like myself enjoys listening at close to reference levels. He had already blown up a Yamaha sub on WOTW and I did not look forward to hearing complaints about the Ultra bottoming so reluctantly and after much cajoling from him I agreed to part with one of my ep600s which I had co-located on the left sidewall.
So on Sunday, I recommenced the process again this time with placement of the ep600 on the left sidewall 5 feet from the left front corner , the 500s on either side of the centre speaker and the Ultra in the right sidewall 5 ft from the front corner all roughly 12ft equidistant from the prime listening position . My concern being a phase issue between the Axioms and the Ultra. My reasoning here was to try a suggestion by Bossobass where the subs should be on average equidistant to more seats to ensure better phasing and smoother FR before equalization. The variable phase on the Ultra was left at 0. Also in order to reduce phase differences between the Ultra and the Axioms and get more extension, I set the Ultra to 12 hz tune to ensure the flattest FR and set the low pass filter at 40hz essentially concentrating the sub to a range around 10hz to 50hz with gain again at -7db. In addition I hoped to minimize THD distortion in the audible range from the infrasonic bass. After going through the phase exercise again the initial SMS graph at the main listening position looked as follows:
A bit bumpier than the all Axiom quad configuration above but not too bad and fixable with the SMS parametric equalizers. I was also fortunate to preserve around 10db of the incremental headroom while extending lower. Here is where the Ultra's peak infrasonic response from room gain was put to good use.
Another 15 minutes of adjusting the SMS eight parametric equalizers and reducing gain on the four subs combined to 0 reference yielded this final graph.
Wonderful looking FR at the prime seating position and the seats on either side, but frequency response elsewhere still leaves much to be desired. No more bottoming subs and no power compression is evident at reference levels. HT bass is the best I have heard so far though I still shut off the subs for music listening. My brother loves the ep600 and my housefold is happier now that the sweeps have stopped.
Other than that alot of low bass vibrating the viewers. Actually the four subs together are calibrated to just 1 or 2 db over 75db reference so its not much louder than where you probably have your Hsu set. Its just very distinct and clean sounding, well articulated really is how I describe it without any noticable artifacts.
A frequency response to die for. 
