Hi efficient speaker, bass problems

"Because all high efficiency speakers have an uneven bandwidth when being driven..."

This depends of the thermal and mechanical power handling capacities of the drivers used. For instance a 4" fullrange driver with 1/2 mm of linear excursion and 30 watts power handling is going to go non-linear long before a 12" prosound woofer with 9 mm of linear excursion and 900 watts power handling. The latter will probably remain linear to a far higher SPL than most similar-sized low-efficiency home audio woofers.

Duke
dealer/manufacturer

In my opinion a fairly high output impedance (low damping factor) amplifier actually works extremely well with proper loudspeaker matching. In the bass region, a high output impedance has the effect of raising the woofer's electrical Q, which reduces electromagnetic damping. But if we start out with a woofer + box that would "normally" be overdamped, the net result with the low-damping-factor amplifier is a properly damped system... but in addition, you get a "free lunch": You end up with deeper bass extension than you would have gotten with a high-damping-factor solid state amp! If done right, you can get as much as 1/3 octave deeper extension. To put this in perspective, this is more than you'd normally get from doubling the box size while maintaining the same efficiency. I exploit this "free lunch" routinely.

Duke

Apparently in reply to my post above, Timlub wrote: "Extension, dampning and efficiency are the tradeoffs. The vectors go in different directions. The op questions grip vs eff. The extension is not. as I understand. part of the query. But I'm all in for a free lunch of any kind. Elucidate."

Sorry for the late resopnse; I didn't check back in on this thread until just now.

First let me say that high damping is not necessarily desirable in and of itself; it is the in-room frequency response that we want to optimize because that is what correlates best with subjective preference. An overdamped speaker sounds... overdamped.

Suppose we have a fairly low-Q woofer in a considerably larger-then-optimum vented box. Let's say we're tuning to the upper 30's, giving us a -3 dB point in the upper 30's with a steep rolloff below that. With typical room gain, we're getting down to the mid 30's.

Now let's drive this same speaker with a low-damping-factor tube amp, but without changeing the tuning frequency. Boom! Quite literally - we now have a broad 2 dB hump across the middle of the bass region, and we've managed to push the -3 dB point south a little bit. So far this isn't a net improvement, but we're not done yet...

Let's drop the tuning frequency to the upper 20's, which elminates any peaking and gives us extension into the lower 30's before room gain. We're on the right track.

Now let's factor in room gain, and push the tuning down into the mid 20's. We get a nice gentle rolloff across the bass region that is approximately the inverse of room gain, and real-world extension into the upper 20's. In contrast, before our low-damping-factor amp came into play, we could have expected extension to the mid 30's after room gain.

This isn't hypothetical. I manufacture this speaker, and chose my woofer and box size and port system with all of this in mind. The scenario I've described doesn't necessarily play out with all high-efficiency speakers.

Duke

You're absolutely right Timlub - my bad! It was indeed Lewhite I was responding to! Thus dissolves my illusion of infallibility...

Okay to answer the questions which I think actually did come from you this time, the subject of room interaction dips (and peaks) is almost a different topic from the one I addressed. One could argue that a speaker which goes deeper is likely to have more peaks and dips simply because its wider bandwidth gives the room more opportunity to screw things up. But beyond that, I don't see any reason why a low-tuned box with a low-damping-factor amp would result in a room-interaction dip. If you're talking about a sag in the bass response above the tuning frequency, actually that's more likely to happen using a conventional amplifier with a low-Q woofer that's in a somewhat oversized vented box. If you have a modelling program, the effect of low damping factor is to raise the apparent Qes of the speaker (without correspondingly lowering its efficiency... hence, a "free lunch").

Oversimplifying a bit, you can get bass boost from boundary reinforcement or from a low output impedance amplifier. The two can work in your favor, or work against you... but if you can adjust the tuning frequency of the box, that can help a great deal.

Now since you brought up the topic of dips, I'll bring up the topic of peaks, so now we're talking about peaks and dips, which arise from speaker/room interaction. They are inevitable. My preferred technique for addressing them is to use four small subwoofers scattered asymmetrically around the room. This way each will produce a different peak-and-dip pattern at any location in the room, because each is a different distance from each of the room boundaries (in the horizontal plane anyway), and the sum of these multiple dissimilar peak-and-dip patterns ends up being pretty darn smooth. Now if we're talking about two speakers and no subs, one way to introduce some of this staggering/smoothing effect is to place the port on the rear of the speaker, at a different height than the woofer. Toe this speaker in fairly aggressively, and the woofer and port are each a different distance from the room boundaries in all three dimensions! Of course the port is fairly close to the wall in this scenario so it's getting a lot of boundary reinforcement, but that's been taken into account in the design stage (and/or the port's tuning frequency is user-adjustable).

In my experience paying all this extra attention to room interaction makes more of a difference in small rooms than it does in large ones.

Since I've been referring to free lunches here and there, let me mention one more that comes along with low damping factor tube amps (like those atmasphere makes): Their characteristics help to partially counteract thermal compression. Thermal compression arises when the voice coil heats up, and when that happens its resistance rises. Taking an extreme example here, at high power levels your 8-ohm woofer may become, in effect, a 10-ohm woofer. Not only does it take more power to push your woofer, but a high-damping-factor (constant voltage) amplifier will actually put out less wattage into that increased impedance, making a bad situation even worse. On the other hand a low-damping-factor amplifier will put out the same or perhaps even somewhat increased wattage into the higher impedance! So if you plan on pushing your speakers hard, and you hate dynamic compression, you might look for speakers that work well with that type of amplifier. Not that this is the biggest of free lunches out there, but it's at least on the menu.

Imho, and as we all now know, I'm hardly infallible...

Duke

"Duke, some of us are not so enamored with the sound of most loudspeakers with ported bass." - Unsound

Me neither! But remember that unless you're outdoors you never hear the bass independent of the room. The bass of most ported-box speakers isn't voiced to synergize well with gain from boundary reinforcement. However if done right, I believe that a good ported box has qualitative advantages over a good sealed box... mainly because it can be more closely tailored to what's going on in the room.

"Duke , it does seem to me that you are discussing speakers designed for corner placement, if not how would this apply when speakers are setup and voiced away from the walls." - Weseixas

Corner placement would be the extreme example, but wouldn't it make sense to approximately match the speaker's inherent bass rolloff characteristics with its acoustic environment regardless of where it is? If the speaker is up on a stand out in the room away from walls, maybe even in an open-floorplan home, then you're going to need a great deal more bass energy from the speaker in order to get acceptable tonal balance. In this scenario we'd shorten the ports and raise the tuning frequency, and once again a ported box gives us options that a sealed box does not.

I don't know why more manufacturers don't incorporate user-adjustable ports. Most of my customers have ended up happiest with a tuning frequency different from the one I shipped the speakers with. One customer is even tuning the two speakers differently based on where they're located (one's in a corner, the other isn't). I use Precision Ports, available from Parts Express and Madisound and probably others, in case anyone wants to retrofit them to their speakers. They screw into the cabinet so they can be removed, and I just use a few wraps of electrical tape to secure the sections together.

Duke

What I've been talking about above might be called "room gain complementary tuning". Let's look at a couple of well-documented examples.

First up, the little Guru QM10, which defies expectations with its bass extension into the lower 30's. When placed up against the wall as recommended, the QM10 gets boundary reinforcement off the wall probably up to 200 Hz or so. As we go progressively lower in frequency, other room surfaces start to become a small enough fraction of a wavelength distant that their reflections are also in-phase, thus reinforcing the original sound. First the floor, then the ceiling, and somewhere along the way the near sidewall, and then the far sidewall. The designers apparently anticipated that this room gain would be about 3 dB per octave, and so they chose a woofer, box size, and tuning frequency that would give an approximately 3 dB per octave anechoic rolloff down to about 35 Hz. The rolloff accelerates after that, and after anticipated room gain we're -3 dB in the lower 30's.

Here are SoundStage's measurements of the Guru QM10. Notice that from about 140 Hz down to about 35 Hz (two octaves), the bass rolls off at about 3 dB per octave anechoic:

Guru QM10 measurements

If anyone has a modelling program, I challenge you to find a practical sealed box alignment that will result in a 3 dB per octave rolloff anechoic like that. A Qts = .50 sealed box still rolls off at 6 dB per octave anechoic.

The Guru is not alone in using a 3 dB per octave rolloff to take advantage of room gain. Let's look at the corner-placement-recommended Audio Note AN-E reviewed by Stereophile. From 200 Hz down to 25 Hz, the rolloff averages about 3 dB per octave:

Audio Note AN-E measurements

My point being, here are successful real-world examples of room-gain-complimentary tuning, including the all-important measurements that reveal what the designer is trying to do. If we were designing a speaker that would normally be getting less boundary reinforcement, we'd want to take that into account as well.

Now wait just a minute, you're probably thinking; all this theory is great, but sealed boxes sound just plain have more natural-sounding bass than vented boxes! Okay, here's why (imho of course):

Most sealed boxes roll off somewhere between 6 dB per octave and 12 dB per octave. Most (nearly all) vented boxes are tuned for deepest-loudest-possible bass, so they are "flat" as low as possible and then roll off very rapidly (18 to 24 dB per octave). Of the two, the sealed box's rolloff characteristics come closer to approximating the inverse of room gain, especially if it's a low-Q box, so sealed boxes will work better in a wider variety of rooms than "typical" vented boxes. And too much bass (more likely from a typical vented box) is a more heinous crime than too little bass.

But take both speakers outdoors, where boundary reinforcement is negligible, and the flat-tuned vented box will have better tonal balance than our sealed box, and much better than our room-gain-complementary tuned Guru or Audio Note. It's not just the speaker and it's not just the room (or where it is in the room) - it's both.

Duke
dealer/manufacturer/no affilication with Guru Pro Audio or Audio Note

Unsound, I don't think the ear can hear all of these lags in the bass region.

That's one of the reasons bass is perceived as omnidirectional inside a room: Our time-domain resolution is so poor down there that the ear/brain system cannot tell the difference between the first-arrival sound and the reflections.

In contrast (so I've been told by a psychoacoustics researcher; never tried this myself), outdoors the ear is able to discern the direction of low frequency energy down to a much lower frequency than in a room. This is because there are no reflections to confuse the ear-brain system by arriving before it has resolved the initial impulse. We can hear direction at very low frequencies, as long as it's not masked by reflections arriving before our ear/brain system can isolate the first-arrival sound. All of those lags you mention arrive before the ear/brain can isolate the first arrival sound, so perceptually they just get blended in with it. And the room's effects on the frequency response are what dominates the perceptual modification of what would otherwise have been a clean first-arrival impulse.

I used to believe that the superior time-domain performance of a good sealed system was the reason for its "tight, natural" sound, but after reading Earl Geddes and Floyd Toole on the subject, I now believe that the in-room frequency response is the dominant factor by far.

Come to think of it, for many years the most natural-sounding bass I'd ever heard was from an IMF transmission line. A transmission line would inherently have poor time-domain behavior but might well have not only a room-gain-complementary native frequency response, but also significant physical displacement of its two bass sources (woofer and line terminus), with ensuing dissimilar room-mode-interaction from the two along with its smoothing effect. This would have addressed the two main problems of in-room bass reproduction: Gain from boundary reinforcement, and peaks-and-dips imposed by room modes (actually just different manifestations of the same problem). But no one really thought in those terms back then.

Duke

"Duke, we have all read that bass is omni-directional, and yet very often it's very easy hear where the sub-woofer is (perhaps it's because the overtones appear to come elsewhere?)." - Unsound

"Bass localization is detectable indoors" - Weseixas

Hearing the location of a subwoofer is not necessarily the same thing as hearing the direction that low frequencies are coming from. Subs can be generating harmonic distortion that is more audible than the actual fundamental, and they can be passing lower midrange energy at a level that is still audible. We easily hear where a bass instrument is in the soundstage from the overtones and other higher frequency sounds (drumskin).

"Despite the theories, some of us clearly prefer the sound of sealed boxes over ported ones. As the old cliche' goes; "the proof of the pudding is in the eating"." - Unsound

If you re-read my theory, it is consistent with your observation.

"...get the math wrong and you will know, phase is absolutely critical." - Weseixas

Phase is audible in the bass region only to the extent that it impacts frequency response. So when you hear a difference after adjusting the relative phase of your subs, or their distance relative to the mains, what you are hearing is the effect on the frequency response.

Duke