I'm curious about something. In perusing the various a/v forums, I frequently see debates about large vs. small settings on a/v receivers/processors, variable crossover frequencies on those processors, difficulty in integrating subwoofers in a/v systems etc. I find myself wondering why people spend so much time fighting with their a/v processors, trying so hard to get things integrated properly. Why don't more people buy speakers that work with the 80Hz 12dB/oct high pass, 24dB/oct low pass crossovers built into most a/v gear. A sealed speaker with an F3 of 80Hz would seem to work perfectly with most receivers. Do people not buy them because there aren't many available? Or is there just a general preference for vented speakers that play lower? I'm genuinely curious about this and would be very interested in hearing your comments on the matter.

Good questions.

First, for the sake of discussion, let me define two categories of speaker: large (capable of full-range bass response) and small (less capable.) I know it's an oversimplification.

Personally I'd aim to put together a system comprised only of large speakers, where all of them had "flat" bass response down to below 30 Hz, run everything full range, and let true subs (4 in total) handle what is essentially only the LFE's. Unfortunately space constraints and budget prevent me, and probably most others, from accomplishing this.

However I am part way towards this goal, and in my main home theatre room am running large full-range arrays for all 3 front channels - and three subs. And while I hate to overuse the adjective, I can say that nothing beats the impact that you can achieve with full range speakers.

Having stated this own personal ideal you can guess that going to smaller and smaller speakers would be, for me, less than ideal. Where budget and space considerations come into play (as they always do) I would still aim for as extended bass as possible - coupled with electronics that would allow me to match up with each speaker's low-level ability.

I'm not saying that smaller 80 Hz speakers are in any way inferior. Or that you can't build an excellent system around them. It's just that having experienced both large and small set-ups, this is one of those rare occasions where size really does matter, in my opinion.

Besides, (and I know I'm generalizing) you can always get a large speaker to behave like a smaller one by slecting a higher crossover point. But the reverse is not true.

Regards.

Jeff Mackwood

Jeff,

Thanks for your comments. They raise another question for me. Is a large speaker with flat frequency response to 30Hz necessarily going to provide more impact in the, say, 200Hz-1kHz range, than would a smaller speaker with flat response to 80Hz?

E

Ellen,

If I remember correctly from HT article, human ear cannot locate the source of sound raging from 40 HZ to 90 HZ (someone correct me if I’m wrong) thus using one equipment (subwoofer) to take over that duty for 5 or 7 speakers makes it economical in general. There are some disadvantages to it if you get really picky like this thread here: http://ubb.outlawaudio.com/ubb/Forum4/HTML/000250.html

For speaker design fundamentals, try these two sites. They should give you a good explanation of why different crossover matters so much.
http://www.speakerbuilder.net
http://www.loudspeakers101.com/Sound.htm

Spiker, thanks for the info. I am familiar with the rudiments of loudspeaker design, having toyed at one point with the idea of building some speakers.

I am also aware that the folks at THX and Dolby have found thru their testing that bass frequencies under 80Hz are not localizable, hence the 80Hz crossover in most a/v and all THX certified processors. But all of that is more reason to use sealed speakers with F3 of 80Hz. Yet I get the impression that most people do not. I'm very curious to know why.

Ellen and Spiker,

Let me again start by saying that what I am about to say is a generality - which is always open to exceptions.

All else being equal, a large speaker with extended bass response should sound no different in the other (upper bass, mid, high) frequencies than a small speaker. One way to prove it might be to take two or three models from the same line of speaker company (Paradigm, Energy, psb, etc.) and compare them side-by-side with the same relatively high crossover point (80 Hz or what have you.) I suspect that they will all have just about the same sound and impact - especially if they share the same mid and high drivers, crossover design philosophy, etc.

What I meant by my earlier comment regarding "impact" is at the lower frequencies where they are present in large, and by my definition, not present in small speakers.

I agree that nothing much below 80 Hz is localizable (if that's a word) - and do know that this has been the subject of debate in other threads regarding "stereo subs" - however having multiple full-range speakers PLUS subs (handling mostly the LFE's) impacting on you, is simply a completely different experience than with small speakers and the same subs. I can prove it by simply by dialing up the crossover point on my full-range front and centre front speaker arrays and redirecting that bass to my three subs. It's just not the same.

(One sidebar comment: a single subwoofer will be more noticeable than multiple subs. Not necessarily localized. However it's more likely to set up all sorts of room response variations which you will notice. Multiple (mono) subs, properly positionned, will tame the peaks and troughs - and does, in my opinion, contribute, incorrectly, to the argument that stereo subs are the way to go. My own listening experience is that multiple mono subs are more effective, Again it's open to debate - but I would refer you to the excellent white paper on the subject on the Harmon International web site.)

But as I said before space and budget (economics) will almost always cause us to compromise. I personally believe that the THX standard is a compromise - even though it's a darn good one.

Regards.

Jeff Mackwood

Jeff,


Believe it not, Webster says it is


In what way is it not the same? Do you mean that the bass is louder? Does it have less audible distortion? I would expect problems in using the SMALL setting on the receiver with large full range speakers that do not roll off at crossover point. I would think that there would be a hump in the response because the speaker is not rolling off. There would be no 12dB/oct roll off at the crossover point to combine with the 12dB/oct highpass in the processor. So instead of combining with the 24dB/oct low pass to get flat response in the crossover region, you'd get something else that wasn't flat.


I am surprised to read this. I had always thought that multiple subs placed at varying positions throughout the room can cause cancellation. But perhaps that's only if both/all subs are receiving the exact same signal.


No doubt.

E

edit to fix myriad typos




[This message has been edited by Ellen (edited November 10, 2003).]

Jeff,
I’ve thought about getting a set up like yours at one point after reading about “Stereo Subs”. Of course it will cost more but for people just wanting to enjoy decent audio quality (not necessarily top notch) in music and movies at home, will it be really worth the investment? Is it night and day difference or less? I’m debating…

I would say that if you have two subs already, setting them up in stereo is worth a try. If you have to buy another sub, I would say it would probably not be worth it unless you already wanted to buy the extra sub. With two subs, you can always set them up conventionally if you find that running them in stereo doesn't work for you. The nice thing about the stereo sub setup is that if you already have the two subs, there is really no cost to do it other than some extra interconnects.

soundhound,


This requires that your subs have a built in electronic crossover, right? Those with passive subs powered by a couple of spare amp channels would have to purchase an external crossover, no?

Yes, I'm assuming that the subs are of the type that have built in line level crossovers. With other types you would need an external crossover, or something like the Outlaw ICBM.

Okay, just checking. I've got a couple of 12" subwoofer drivers sitting idle in the blackhole that is my hall closet. One of these days I'll get somebody to build a couple of enclosures for them. I do have an active crossover and power amp sitting around too, so maybe I'll try out stereo sub thing.

But I think that is a separate issue from the whole large/small vented/sealed speaker issue that made me post my initial question. I was hoping more people would respond. Guess I'll just continue to wonder why folks set their systems up as they do.

Ellen,

there are plenty of sub enclosure box options here: http://www.acoustic-visions.com/~acoustic/products/subwoofers/enclosures/

Where to start?...

Gotta agree with Soundhound regarding playing with two subs and seeing (err...hearing) if there's a difference. Personally I don't notice any "stereo" sub effect, but I'm sure others have since they swear by it. As I tried to say, I think that going from one to two subs is a big improvement not because of a stereo effect, but because two subs (properly placed of course) tends to smooth out the local peaks and dips. Additional subs give additional improvement but the old law of diminishing returns starts to kick in really heavy by the time you get to four. From what I recall of the Harmon paper, there's also the potential of some trade off with extreme bass extension with multiple subs as well - but I'd have to re-read it to be sure. All I know is that when I read it, and looked at my main home theatre's set-up, that I set my goal at four subs. One to go!

But back to the question of impact etc. and big vs small...

When I used to run dual subs and had an ICBM in the loop, I would play with the ICBM's crossover points. Starting at bypass I'd roll off the full range speakers and let the subs handle more and more of the load. It's hard to keep all else equal (like invidual volume levels) but I would pause after each step and recalibrate the levels for all speakers - including the subs. The sound definitely changed. Thinner? Somewhat. Louder? Not in absolute terms since I was recalibrating but yes there was obviously less bass coming from the full range speakers and more from the sub - but of course that's what you would expect. More or less distortion? Nothing noticeable. Was I hitting any crossover points that threw things out of whack? None that I could tell. More impact? Yes. Explosions like in the U-571 depth charge sequence were (and are) much more like external cardiac compression with the full range speakers running full range. You were "hit" from multiple points - rather than from just two. Perhaps that's the answer. There was even less cancellation and smoother response with all running full range. Maybe it's because my full range arrays are fed plenty of power, are relatively high efficiency, and are capable of considerable peak output.

Unfortunately I could probably go on at length - and Monday Night Football is about to begin.

I hope if not confused the issue too much.

Regards,

Jeff Mackwood

Jeff, is this the Harman white paper to which you referred?

Addressing the original post, you bring up the ultimate question, IMO.

So many posts regarding 'seamless' sub integration overlook the fact that unless your satellites are 12 dB/octave (sealed), you will not be able to acheive a 'seamless blend' using standard BM.

I think the answer to the question raised is that 'seamless' integration of sub/sat at 80 Hz requires the sat to be flat (+/- 3 dB) to 40 Hz, and, yes, it's because there are very few, if any sats available that can meet this requirement.

The problem with 'large' sats is that, as most people know, having, in effect, 5 or 6 subwoofers that have little, if any, placement options (for example, corner loading a full range speaker is a disaster), makes this setup less than optimal.

Stereo subs still requires that you match the high pass filter/sat speaker type to the low pass filter slope on the sub.

The road to the answer is a simple one: Use a low pass filter slope of 36 dB/octave on the sub when using ported sats with a 12 dB/octave high pass slope. Why this isn't a standard option just baffles me.

The next step is to have a dedicated LFE sub system to keep the LFE (mono) signal out of the sub/sat integration scheme.

Stereo has no LFE, MC music uses the LFE as a bass boost channel with little to some signal and movie soundtracks use huge amounts of LFE + 10 dB signal. The only way to control it properly, as to calibration and integration, is to keep it discrete through playback, with it's own LP point, slope and phase/time adjustment options.

Then again, I could be completely mistaken.

Bob,





I've read them over a couple of times and these two statements seem at odds to me. Do you mean to say that you need the second if you can't meet the first? That is, if you've got sealed speakers with an F3 of 80Hz that can reach Dolby reference levels you are in business. But if not, you need some speakers that are flat to 40Hz to make up for the fact that you don't have speakers that match up ideally with the standard crossover in most receivers? If that isn't what you are saying, can you please elaborate? This sounds like that whole "set the crossover to one octave above your speakers' F3 point" thing that I have never understood.

Ellen, like Ben Franklin was fond of saying, "In theory, practice and theory are the same. In practice they are not". BoB is basically saying the same thing in the 2 statements; an electrical filter is relatively easy to make, and is typically very close to specs. So a typical 12dB xover at 80 hz would be down 12dB at half and double the center freq, ie 40 and 160 Hz. A sealed box sub would have to have a B2 response, ie Qts=.707, to properly interact, which in reality doesn't always happen. But wait, there's more. The real problem lies in the room interaction; if you did a top-notch analysis of your rooms response, you'd likely be horrified. My room has dips of 8-10 dB, and peaks of 10 at 55, and 18dB at 120 Hz. Even after a bass trap, which causes my wife, and therefore myself no end of grief, and equalization, I still have +/- 5 dB through the bass range. So IMNSHO, the exact response of speaker/xover is less critical than theory might suggest.
Mike

Just as a side note - I've read some interesting stuff on bass line arrays, and my old system was basically able to be configured this way. It seemed to work pretty well, but that's a sample size of one.

My opinion on the multiple sub issue:

 Yes, worth it for smoothing out response in the main listening area.
 Trade-off: Outside of my main listening area, such as adjacent rooms and beyond, there is a very noticeable cancellation effect that is not present with only one sub.
 Stereo vs. multiple mono: I choose stereo for the difference (as it seems to me in my listening environment) in psycho-acoustic effect in playback of as-it-happened, “live” microphone-to-track recordings with little or no “mix-down.” For heavily mixed productions where items are electronically “placed” in one or more channels from a single recorded track, I doubt I could tell the difference in an A-B comparison test. With differences in both preference and listening environment from person to person, if you have access to a more than one sub, spend some time listening with differing setups and find what works best for you.

Regarding crossover selection and the issue of which speaker will handle what range of frequencies, I take into account several factors. Some of these heavily reflect my personal preferences:

Do I wish for the same loudspeaker in all 7 positions? Yes. Is the goal of high accuracy speakers with true ‘full range’ at all 7 positions realistic for me? No.

Also, my listening expectations, since moving to a system with a subwoofer about eight years ago, have come to include a reduction of low-end inter-modulation distortion. This IM distortion occurs when, for example, medium-level signals in the 300 Hz to 600 Hz range are literally riding on large-level signals in the 40 Hz to 80 Hz range because both ranges are being reproduced by the same moving surface. IM cannot be totally eliminated, but ‘sharing the load,’ by letting a subwoofer handle the lowest frequencies, helps a great deal as I hear it. For years and years in listening to four-way full range speakers with 16”, 8”, horn and tweeter components, I was accustomed to this type of inter-modulation being present. Without realizing it at first, when I went with a good subwoofer about eight years ago, I missed the this effect and had my subwoofer gain too high trying to make up for the missing low-end IM effect. After some careful tweaking and ‘learning’ the new sound, I now find a reduction in low-end IM effect to have a more ‘clean’ sound.

If I crossover at 80 Hz, then my subs are handling the lowest two-and-a-quarter octaves, and the woofer in my three-way loudspeakers are handling the next three-and-a-quarter octaves. This, in my mind, is a reasonably even sharing. If I crossover at a higher frequency for technical reasons, I am increasing the likelihood of localization. If I crossover at a lower frequency, the sharing of the low end becomes more lopsided and increases low-end IM effect. So for both technical and listening preference reasons, I choose to crossover at 80 Hz.

One last point before I cease being so verbose. My powered subs can cruise in bass-heavy passages, without ill effect, at power levels two or three times that of my three-way speakers. By moving more of the largest amplitude signals to the separately powered subs, the rest of the system can cruise more easily at even loud levels, and is capable of more overall if I really want to push it.

Thanks for listening, everyone.

BB4TB

All great points.

The bottom line seems to be that there's no single right or wrong answer - but lots of fun to be had trying different approaches and finding what works best for you.

Jeff Mackwood

This is a cool piece:
http://www.nhthifi.com/2004/products/product_detail.asp?ProductLineID=2&ProductID=47
It allows for the stereo subs scheme AND a 3rd sub for discrete LFE (looking at the I/O and controls, that is. I haven't read the manual).

Ellen,

If your crossover point (the point at which the sat begins to drop going down and your sub begins to drop going up) is at 80 Hz, the sat has to continue to play as the frequency drops below that point while your sub has to continue to play as the frequency rises above that same point.

IOW, the crossover filters are doing the attenuation in both cases. If your speakers can't play in the crossover region without their own natural drop off interfering in this process, then the crossover (or, blend from the sat to the sub) cannot be predictable.

This is why a sat should be flat (+/- 3dB) to as much as 1 octave below the crossover point
for the filter to be able to do it's job correctly.

This is also why a ported sat (which has a natural roll off of 24 dB/octave, and when the standard 12 dB/octave filter is applied has a new roll off of 36 dB/octave) doesn't blend well with a sub that has a 24 dB/octave slope applied by the low pass filter.

Sealed sats that are flat to 40 Hz are much less available than ported sats that are flat to 40 Hz.

Sorry, I'm not that adept at the short version.

Charlie...it appears that your system was way ahead of it's time...and a great looking one, at that!