In my opinion (gotta qualify here) you have nothing to lose and everything to gain. As long as it (or they in my case) is properly integrated it can be a wonderful addition. My JL Audio subs won't work the way you're going to do it but if you're missing bass you can complete your sound if you pull it off with the right sub(s). When it is done right you never know a sub is present but the bass isn't missing when it should be there. My subs get the full range signal and my NAD M3 integrated can send either a full range signal or use the built-in bass management feature to limit the bass signal sent to the mains if I choose. Works like a charm.
30 responses Add your response
I would go with the Wilson Benesch Torus Infrasonic Generator.. There is no better way to get those low notes at electrostatic speed.. It will cost you about $10,000, and it is much better then the Rel..
or you could ask yourself a couple of questions?
1. Does my system totally involve me with its realistic and natural sound.
2. Is the presentation relaxed enough to allow me to listen for long hours.
3. Does my system sound pure? Or is it a little blurry?
4. Am I looking for more bass because I really can't answer yes to any of the above and I don't know what to do next?
Something to think about..
I'm with Bizango1 - it'll probably be a big gain. I run my mains the way you plan to, with a high-pass filter (in my case 60Hz). I go further and bi-amp the mains - I run a full-spectrum direct signal to the midrange-tweeter sections, and high-pass only to the woofers. I use a y-cable to split the preamp signal. The improvement in my system overall is amazing. I'm very happy with the result.
IME, you will lose nothing - if you get it right - and you will likely gain a ton.
I've used a lot of combinations successfully, but the common denominator has been a sub controller. I think that getting seamless integration of your sub(s) is exponentially more difficult if you try to do it by ear. I also think that digital room EQ (at least for the subwoofer signal) is tremendously beneficial.
Until recently, I used a stand-alone Velodyne SMS-1 sub controller. Last week, I replaced my outboard x-over (NHT X-2), preamp (ARC LS 25), dac (Benchmark DAC-1), and sub-controller (Velo SMS-1) with an Audyssey XT32 enabled pre-pro (Onkyo 5508).
While I have not completely tendered my "official audiophile merit badge and ID card" (I still have the ARC gear and a bunch of Joule, Cary, PL, etc), I will say this system sounds awfully good to me.
I have agree you should only gain when set up properly. The KEY word is setup properly. When I first got my Dyn C1's I thought my Rel B3 was slow sounding compared to the C1's without the sub. Turns out I had it set up all wrong. Crossed over way to high and the volume was too high. I could always hear where the sub was. Once I had that set properly it still sounded slow to me. That was because I had the timing wrong (physical position). The low bass sound wave lagged the original signal. Had to move it about forward 4". Then I was a happy camper until I started reading these posts regarding subs and many stated the cable made a huge difference. So on a whim I ordered a Nordost baseline sub cable made to Rel spec. Holy Moly what a difference I heard. The attack and decay didn't exist in the stock Rel cable.
That being said it was all for the lowest bass that my speakers couldn't reproduce. I will add as a bonus that the mids actually sound better too (especially acoustic guitars - sorry Learsfool but I strongly disagree now that mine is set up right and I would agree if not set up properly). Nothing earth shattering like the low bass but just a hair better. On occasion I will disconnect the sub and it amazes me how few recordings have a real low bass.
Tghooper I don't have an external crossover so I can't comment on crossing them over that high. I have mixed feelings about doing that. My speakers are rated 45 - 22khz and my crossover is set at 41hz on the sub. Besides my sub cable is 5 1/2m and my tube integrated only has pre out but no amp in. Either way I wouldn't want 11m of cable between my pre out and amp in.
So in the end if your are considering a sub I say the 'nike' slogan - JUST DO IT !!! Just remember if it doesn't sound right you don't have it set up properly.
What might you lose?
You might lose your mind if you have too much trouble blending the sub with your mains and the sub doesn't have the controls you need to do the blend. Otherwise, adding a sub or two can be very rewarding, especially with a system based on a low-powered tube amp.
There is a lot of advice on this thread on how to do it, and I think most contributors are advocating for what worked for *them*, but it may not be the best for *you*. I've blended seven different sub configurations in four different systems and they've all succeeded but my methodologies have varied according to several variables:
--The room acoustics
--The bass extension of the main or satellite speakers. on full range I cross over at 40 Hz; on some small satellites you may have to cross over at 70/80 or even 120 Hz.
--The I/Os on the sub--single LFE line level, stereo line level, stereo line level I/O, speaker level, speaker level I/O
--How much phase control the sub has; some have none, many have just a 0-180deg. switch, some have a 3- or 4-way switch (0, 45, 90, 180 deg.) or continuously variable (my latest have 0-360 continuous)
--Crossover control--hopefully continuously variable from around 40 Hz up to 80, 150, maybe 200
Conventional wisdom says that you get more dynamic range and more amplifier efficiency by using a high pass filter to the stereo pair (mains). No doubt this can be true in some cases with careful crossover matching and placement and/or phase setting, but in my own experience I get a better blend by running the mains at full range and blending in the subwoofer at the lowest possible crossover point. If the crossover point is too high it starts making vocals and some lower midrange/upper bass sound thick and murky.
This also flies in the face of conventional wisdom, but I've had a couple of subs that actually sounded better using the speaker connections than the RCAs. Don't know why; on these certain subs (2004-era Mirage LF-100 and LF-150) the subs sounded faster, fuller, more musical and better level-matched with the speaker cables than with the RCA connection.
A month ago I added a single tiny Mirage MM8 sub to my 2-channel stereo with Mirage OMD-15 tower speakers. The MM8 has no speaker level inputs and no outputs at all, so there's no option to sent a high pass signal to the mains. The towers have high handling capacity and are rated down to about 36 Hz, but I wasn't really getting that in-room. The sub, crossed over at about 50 Hz, helped deepen bass extension and kept the sub from murking up the upper bass handled by the towers.
These subs are a good match for the speakers. The subs are sealed, very powerful (1200 watts peak each), and extremely quick, so they easily match up with the mains. As 9" cubes they're made to blend with music systems or add bass to small HT systems, but not to rattle your teeth like a JL Gotham.
Another deal came my way and just yesterday I added a second MM8 to this system. It's just way better with two. The second bass smooths out the bass response (less peaky) and it adds about 6 dB dynamic range.
The funny thing is that somehow the mains (even though I'm *not* running them with a high-pass filter) sound more spacious, more dynamic, more 3-dimensional, more clear.
Right now I'm playing solo acoustic guitar and somehow the sound is simply opened up. Not overly bassy, not murky, but rather airy, more real, more dimensional, more fully fleshed out throughout. I'm a believer now.
But also, where this was previously a somewhat modest system best suited to semi-acoustic pop, rock, small group jazz, vocalists, and chamber music, it's now ready to boogie with big band and full-scale orchestral pieces such as Pictures at an Exhibition, Pines of Rome, and The Planets.
Those who have far bigger budgets who have implemented subs that are linear down to 14 Hz or so testify that the infrasonic information energizes the listening room more like what you sense and feel in a live concert.
I'm well on my way to thinking that if you want a truly fully convincing system that draws you into whatever kind of music you play on it, it could start with the subwoofer(s). In other words, it may actually be beneficial to allocate more money for the subs than the mains. I have no doubt that a pair of JLs costing around $6K could make a $1K pair of speakers sing as never imagined.
Anyway, back to your system, I suspect you'd get your best results by running your main speakers full range and sending simultaneous signal to the sub either via a second set of speaker cables from the amp or a second set of RCAs from the preamp via Y-adaptors. Then use the sub's volume, phase, and crossover controls to blend it (them) with your main speakers.
For blending sub(s) with a 2-channel music system, if you can't afford (or lack the space or floor strength) for a do-it-all sub such as a JL or Velodyne DD+ series, then go for the light'n'lively type of sub with a sealed enclosure, perhaps with passive radiator(s), lots of power, and fully continuous crossover, volume, and phase settings for the best blend with your main speakers.
"This also flies in the face of conventional wisdom, but I've had a couple of subs that actually sounded better using the speaker connections than the RCAs."
Doesn't that imply that the main amp is superior to the amp in the sub, even within the limited frequency domain of the sub. I'm not sure what's meant by fast in the LF range where wave lengths are long. Perhaps it means the sub is farther from the listener than the mains, but most processors let you account for speaker distances during setup.
12-02-11: DbphdWith a powered sub, the amplifier in the sub is being used regardless of whether the signal is sent into the sub via speaker-level or line-level inputs. If the speaker-level inputs are used, conceptually what is happening is that the voltage range that is present at the main amp's output is divided down by the sub to a line-level voltage range, and then sent into the sub's power amp. Almost no current and power is drawn from the main amp by the sub.
The main difference between the two configurations is that the sonic effects of the main amp, whatever they may be, will be seen by the sub as well as by the main speakers, if speaker-level inputs are used.
Also, if line-level inputs are used, and the preamp's output is split between main amp and sub via a y-adapter, or if the two destinations are driven from separate output jacks on the preamp that are not individually buffered (as is usually the case when two sets of output jacks are provided), adverse effects can occur under certain circumstances.
First, the preamp will see a considerably reduced load impedance, which can result in frequency response issues such as deep bass rolloff, if its own output impedance is not low enough in relation to that reduced overall load impedance. The overall load impedance seen by the preamp will be equal to the product (multiplication) of the two input impedances, divided by the sum of those numbers. For example, if the sub has an input impedance of 20K, and the main amp has an input impedance of 50K, the combined impedance is (20 x 50)/(20 + 50) = 14.3K. For best results, the preamp's output impedance should be at least a factor of 10 less than that (i.e., 1.43K) at all frequencies. If the worst case (maximum) output impedance of the preamp across the audible frequency range is not known, and just a nominal output impedance is specified, it would be safer to use a factor of 50 or so.
Second, the upper treble, as reproduced by the main speakers, may be affected by the capacitance of the cable that connects preamp to sub, as well as by the capacitance of the cable that connects preamp to main amp. The total of the two cable capacitances will interact with the output impedance of the preamp to form a low pass filter. Under extreme circumstances (high preamp output impedance at high frequencies, long cable lengths, high cable capacitance per unit length), the bandwidth of that low pass filter can be low enough to result in a small but audible upper treble rolloff, and consequent softening of transients.
There is no universally applicable answer to the speaker-level vs. line-level connection question.
Wow! Lots of great info here...except for SoundsReal and his $10K sub. LOL
Doak - you bring up a good point. I'll have to check with the speaker manufacturer.
My plan is to use two small 8" subs. Found a good deal on Sunfire HRS-8 that's within my budget. My listening room is 12x16 with 13' ceilings.
I'll report back.
12-02-11: TghooperI use a pair of Mirage MM8's, which are 8" drivers and passive radiators in a 9" cube. However, they really don't add much below 35 Hz. Still, they're a good match for my system at this point. And I got them for less than $280 each.
If 14" cubes aren't too big and you want strong performance down into the 20's, you might want to consider these. They're sealed and tight, and the price is right for what they do. I have a local audiobuddy who uses them with his Magnepan 1.7's.
"Doesn't that imply that the main amp is superior to the amp in the sub, even within the limited frequency domain of the sub."
No. The sub driver is always driven by the sub's amp. The speaker-level inputs use a relatively high input impedance to reduce the main amp's output to just a couple of volts, and uses the low-voltage signal to drive the sub's internal amp. More likely there's some sort of frequency response difference caused by the interaction of the amp's output stage and the sub that is fortuitous.
Fast in the sense of rise time, which--regardless of frequency range--has to do with how quickly the cone accelerates, and perhaps more importantly, how quickly it stops. When you pluck a string bass for example, the fundamental plus its overtones are spread across the subwoofer, woofer, midrange were applicable, and tweeter. For ideal cohesion of just this single note, the cones of all four drivers should accelerate at the same rate; otherwise the overtones arrive at your ears before the fundamental. It stands to reason that a carbon fiber midrange or a beryllium dome tweeter with neodymium magnet is going to reach its excursion point faster than a 15" driver made of conventional materials with an average strength magnet seated in a less-than-rigid stamped frame.
If you look at the evolution of subwoofers, the emphasis has (generally speaking) evolved from cone diameter to absolutely massive magnets, heavy deep excursion cast frames, high excursion surrounds, and way more powerful internal amps powering much lighter, smaller diameter cones.
Even if you argue that rise time of the sub cone is irrelevant (and I think it's *very* relevant), you must at least be able to grasp that the higher amp power coupled with a stronger magnet and lighter diaphragm better controls diaphragm motion and makes the diaphragm stop when it's supposed to with less after-ring. This would result in cleaner bass more in keeping with the acceleration and stopping characteristics of the lighter midrange and tweeter, helping to prevent the sub from muddying the sound.
Also, a sub with a faster rise time also has higher frequency response which can be handy in blending with some L-R speakers. If a sub has a 12 dB/octave slope in its crossover, then if the crossover is set to the Dolby surround standard of 80Hz, its -6dB point is about 120 Hz, so the idea that a sub only need make a 40 Hz wave is simply not realistic.
Well Al and Irv, I suppose I should be embarrassed by your uncovering what I didn't know about subs being driven by speaker-level inputs, but in fact I learned something. Thanks.
Johnny, I certainly agree with your point about needing tight control of cone excursion; I'm skeptical about your rise time argument.
On a personal note, I use a pair of Velodyne HGS-15s controlled by a VMS-1 bass manager. The subs blend seamlessly with my KEF Reference 104/2s. I was attracted by Velodyne's high-gain servo and powerful amp, which hold promise for matching cone excursion to electrical input.
Has anyone experimented with using standard RCA cables with Y-adapter vs. using a Y-cable? I only see the real cheap generic Y-cables available as most cable companies don't make the latter.
Seems like the using a cheap $5 Y-adaptor on any cable will pick up noise...no matter how expensive the cable.
While I'm generally in agreement with the authors, I'd offer a slightly different take on "bass speed" than the article Al so kindly linked (based on my own experience with subwoofers over the last 5 or 6 years).
The issues of acceleration/deceleration of the woofer cone is real, but a small (i.e. lightweight) cone is not the only answer. As in a car, there are 3 primary factors at work: weight, motor power, and brakes.
High quality subwoofers have oversized motor assemblies and accelerate more than quickly enough to make that issue moot regardless of driver size (per the article). However, the "brakes" are to me, the more interesting part of the equation.
The suspension system is the effective "brake" in a subwoofer, and different subs are tuned quite differently. Sealed boxes inherently provide a pretty stiff suspension, while ported subs can vary widely. The trick is to properly "damp" the cone so that it stops moving when the input signal stops. The result is "fast" bass. One measurement of this is "group delay" and, IME, it seems to correlate pretty well with the phenomenon that I'm describing.
Even though instantaneous start/stop cone response (zero group delay, if that were possible) might seem ideal, I've concluded that faster is not always better.. IME, subs can even be "overdamped" (too fast?), because you're really trying the match the sub's damping to the low frequency behavior of the the bass driver in your main speaker, rather than to some abstract ideal of "high speed". And, depending on the design of the main speaker in question, that behavior will vary considerably.
I use a pair of 12" Rythmik subs with adjustable damping and, at various times, I've mated them to Ohm 100s, Maggie MMGs, SF Minuettos, Monitor Audio Radius 270s, and a few others. I optimize the interface for flat, smooth response with a sub controller (most often Velodyne SMS-1), and then tweak the damping by ear, then re-optimize the frequency response (if necessary). I have ended up with different damping settings, depending on the speaker with which I've tried to mate the sub.
As a general rule, maximum displacement capability will dictate how low and loud a subwoofer can cleanly go. In this regard, large drivers are a much more efficient solution than small drivers. This is why the measured distortion of various subs at high(ish) SPL and low frequency almost always favor the large driver, and almost always by a very large margin.
My bottom line:Speed matters, but faster isn't always better. And, while speed matters, smaller (drivers) does not mean higher speed. As a general rule for high quality subwoofers, bigger is better.
In some ways I'm just re-stating what the authors of the linked article are saying, but doing so thru the lens of my own experience, FWIW.