Speaker-room suckout and other tidbits.

I've mentioned this before, but I will mention it again: in typical home-listening rooms, spaced-apart woofer (or subwoofer) systems will generate a cancellation notch at some bass frequency that is dependent upon both the distance between woofer (or subwoofer) driver centers and the frequency. Depending upon the spacing between the systems (between woofer or subwoofer driver centers), at some bass frequency the rarefaction wave from one woofer or subwoofer will reach the other woofer or subwoofer just as it is generating a pressure wave. (More on this up ahead.) The two cancel out and you get a power-response notch. There is no way to get away from this with spaced woofer/subwoofer systems generating identical or nearly identical bass signals.

A similar thing happens with single woofers and subwoofers interacting with stiff, large-area wall, floor, and ceiling surfaces. The large surface area will reflect back the signal to the radiating driver as if it were being radiated by a second woofer or subwoofer at twice the distance from the single driver's center to the boundary. The boundary creates a mirror-image situation that mimics a second woofer or subwoofer driver.

For example, a situation where you have two spaced woofers or subwoofers 12 feet apart or another situation where you have one woofer or subwoofer 6 feet from a large boundary will each generate a suckout notch centered at 56.5 Hz. With one system you have a boundary and with two systems you have a faux boundary exactly between the two sound sources.

Note that this phenomenon is unrelated to standing waves, which involve boundary/ boundary interactions. The suckout effect is quite different and involves either woofer/boundary interactions or woofer/woofer interactions. There is a formula to calculate this notching as it relates to woofer/boundary interactions: 1130/d x 0.3

Here, "d" is the distance in feet from the woofer center (measured by the shortest route possible) to the closest part of the boundary, and 0.3 is the multiplier that calculates the frequency of the dip. Actually, the true quarter-wavelength multiplier should be 0.25 and not 0.3. However, because the boundary surface is not equidistant over its entire surface from the driver center, it has been found that 0.3 works better.

When calculating the suckout notch between woofer (or subwoofer) centers you would use half the distance (1/2d) between them as d. You would still use the 0.3 multiplier, because the spaced woofers are generating a faux flat boundary between them.

The big problem occurs when you have multiple boundary or inter-woofer interactions. For example, if the woofer (or subwoofer) centers are 10 feet apart and one or more of them are also 5 feet from a large room boundary the suckout notch will be augmented, in this case centered at 67.8 Hz. Note that the distances do not have to be exact. Woofers ten feet apart will still have additional attenuation applied if one or more of them are, say, 4' 10" or 5' 2" from a room boundary. The notching is not so abrupt that slightly different distances do not count. The suckout slope will be gradual enough for close fairly distances to still add to the effect. Obviously, it is a good idea to get as much asymmetry as possible when it comes to dealing with bass-range cancellations.

Actually, at least with full-range systems placed in typical locations, this suckout phenomenon is more likely to be a problem in the middle bass, instead of in the low bass, because the woofers in such systems tend to be fairly close to room boundaries. "Fairly close" in this case means less than, say, three feet. However, when woofer/ subwoofer systems are placed large distances apart (more than eight feet) or large distances from room boundaries (more than four feet) it can happen fairly far down in frequency, too.

With single subwoofers placed in comers, the issue does not exist, because at such close distances any boundary-related notching would be generated well above the operating range of the system. Indeed, one of the advantages of subwoofer/satellite systems that use only one subwoofer (at least as it relates to suckout notches in the range below middle bass) is that one can position the satellites so that any potential suckout effects they would generate would be below their crossover-controlled operating range. And as noted, any that the corner-located subwoofer might generate would be above its crossover-controlled operating range.

For example, if you have a sub/sat system with the sub located in the corner it is likely that any three-boundary suckout notches will be between 200 and 600 Hz. Obviously, if you have the sub/sat crossover set at 80 Hz. these artifacts will not be reproduced by the subwoofer. At the same time, the potential inter-woofer and some (but not all) of the woofer/boundary artifacts that would be generated by the satellites will be in the 50 to 70 Hz range, which is below the 80-Hz crossover point. This situation still does not solve any middle-bass, closer-boundary suckout problems with the satellites, but it does eliminate any for them that would involve longer-distance inter-woofer or woofer/boundary artifacts.

A lot of people are still confused about just what the suckout effect (often called the Allison Effect, after Roy Allison, who documented its existence years ago) is all about. Many people will mention "floor bounce" when discussing cancellation effects and speaker measurements, but the phenomenon happens with all large room boundaries and not just the floor.

This cancellation artifact impacts the power response of the system, whereas your typical floor-bounce artifact (where a second, reflected signal arrives later than the original after hitting the floor between the listener and the speaker) involves first-arrival signals. While the frequency of a floor-bounce notch will be effected somewhat by the listener's location, the much more important power-response suckout will be the same anywhere in the listening room. Below is an explanation of why the effect happens at all with woofer/boundary interactions, with my example primarily dealing with the effect in the middle-bass region. As noted above, at greater distances the suckout will happen at lower frequencies.

Let's look at a typical box loudspeaker system positioned in a room so that its woofer cone center is about two feet from each of the three nearest room surfaces (floor and two intersecting walls). When the speaker is radiating a very low frequency the cone moves relatively slowly and over a relatively long distance. If the radiated frequency is 40 Hz, for example, it takes 1/40 second (25 milliseconds) for the cone to execute one complete forward-backward cycle. Each half cycle takes 12.5 milliseconds (ms).

As the cone begins a forward movement it generates the start of a compression wave. This impulse travels at the speed of sound (approximately 1130 feet per second at sea level) to those nearby room boundaries and is reflected back toward the woofer cone, arriving there some 3.5 ms after it left, while the woofer is still generating the compression half of the sound cycle. The reflected waves increase the instantaneous pressure seen by the woofer and enable it to radiate more power than it could in free space. This is why placing woofers (or subwoofers) close to boundaries augments their outputs.

However, as the woofer tries to radiate at higher, middle-bass frequencies, it must reverse its motion more quickly. For example, at 140 Hz. (the middle bass, for sure), the cone reverses direction every 3.5 ms. It begins its half-cycle of motion (attempting to create a rarefaction) just as the compression-wave reflections from those two-foot distant room boundaries begin to arrive back at the woofer. In this case, the reflected signal is out of phase with the cone motion, decreasing its radiation efficiency. The result is a suckout notch in what would otherwise be a flat woofer-output signal.

As I indicated before, this phenomenon will exist in all parts of the room, since it deals with the actual power input of the speaker to the room. That sets it apart from a standing-wave artifact, as well as from your standard floor-bounce anomaly. It is also much more influential than the latter, because power response is a much larger percentage of the total output than the direct response.

The "Impedance Switch." No doubt, some of you have purchased Yamaha receivers or integrated amplifiers, and some have probably done so on my recommendation. They are superb devices for what they cost, and should be able to work well for years with no problems.

However, the Yamaha units, as well as some receivers made by a number of other outfits, have an impedance-matching switch on the backside that may be causing some confusion. If most of your speakers are 8-ohms or higher you would set the switch to one position, while if they are 6-ohms or lower you would set it to the other position. The problem is that some of you may have a mix of 4-, 6- and 8-ohm speakers in your system. Where to set the switch, and will setting it wrong cause problems?

With all the systems I know about the switch exists for only one purpose: to pass Underwriter's Laboratory heat-related safety tests when driving lower impedance loads. Some receivers don't include an impedance switch, but in that case you will often discover that the owner's manual mandates a speaker impedance of no less than 6 ohms. However, this does not mean that they cannot drive lower impedance loads. It simply means that they may not pass UL certification in such situations. A similarly designed amp with a switch option and with the switch set to the 4-ohm position will pass UL tests with a lower impedance load. However, the downside is that the unit will not be able to generate its full-rated power.

Basically, if a switch is included it will probably control the voltage to the amplifier's output devices power supply. At the 8-ohm setting, the power supply will run at its full design specification, for full-rated power into 8 ohms. This is the setting that is used when the amp is rated for its best power delivery, and most reviewers leave it at this position when testing the unit for maximum power output, even when measuring with 4-ohm loads.

When the switch is moved to the 4-ohm setting, the voltage to the power supply will be reduced (the actual amount will vary from manufacturer to manufacturer, and it might be anywhere from 10 to 25 percent). As a result, the amplifier's available power is reduced by a similar amount. As long as the power limits at the lower setting are not exceeded, the switch position has no impact on the sound. It just allows the amp to run cooler with low-impedance loads. The downside is that maximum power output is less, which might be a problem in some rooms or with some speakers.

If you have inefficient 4-ohm speakers anywhere in your system (particularly if they are the left/right or center units) and/or are using them in a large room, should you take a chance and try the potentially more powerful 8-ohm setting? The answer is "yes," as long as the amp or receiver does not become excessively warm during high-output use. If it does get too warm, the 4-ohm setting should be used instead. While some warmth is no big deal, long-term excessive heat will shorten the life of your component.

If an amplifier or receiver is decently high in quality it should have no problems with low-impedance loads and if properly ventilated and the unit does not warm up to the point where the chassis seems "hot," the impedance switch can be safely left in the 8-ohm position. Just don't expect it to pass the rigorous UL thermal test while doing so.

The owner's manual that comes with any receiver that has an impedance switch will probably make a bold-print warning about cutting power before moving the switch, but I still want to emphasize right here that it is important to TURN OFF your receiver or amp before moving the switch. If you do not do this you could damage the unit.

Update Note: In a previous Skeptimania column I described how I biamped the Dunlavy Cantata speakers in my middle system. (I reviewed the Cantatas in issue 87.) I did this because I wanted to high-pass-filter the otherwise unfiltered midrange drivers. The Cantata is essentially a two-way system that employs an on-board woofer system to augment the low bass output of the two midrange/ woofer drivers. (Electrically and mechanically, the Cantata midrange and tweeter array behaves much like the earlier Dunlavy SC-II system that I reviewed in issue 70.) Biamping and employing an outboard high-pass filter would allow the midrange drivers to only do midrange work--a good thing.

Normally, you could do this by merely setting a typical surround processor or outboard crossover so that the main-channel speakers were high-pass filtered above a certain point and with the low bass going to a subwoofer. This is the easy way.

However, I also wanted to keep using the Cantata woofers, even though I was also going to employ a subwoofer. I did hookup in such a way that the separately amplified Cantata woofers and my SVS 16-46PC subwoofer were handling the low bass as a team, with the Cantata woofers given additional low-pass filtering by the same crossover that was high-pass filtering the midrange drivers. (The Yamaha DSP-A1 processor/amp that I use with my middle system allows for a "both" setting for subwoofer output.) This was overkill in terms of bass potency (it would be so even in the expanded-size room that I was having built), but I did like the idea of the Cantata woofers pulling some of the low-bass weight.

Setting up this package was rather tedious (complex RTA work would be mandatory for setting up anything this acoustically complex), but once I got the work done the result was terrific. Relieving the midrange drivers from bass work that was redundant with what the woofers and subwoofers could do well by themselves allowed them to cleanly do their jobs better than ever. The two Cantata woofers and the SVS sub did not exhibit any obnoxious cancellation effects, due the placement of the systems.

Unfortunately, I did run into one problem: the plethora of wires, amps, and equalizers managed to generate a small amount of ground-loop hum in the main channels. You could not hear it if you were more than five feet away from either speaker, but it was there nonetheless. No amount of diddling could shake that effect from the system.

Given my musical tastes, noise like that is not my bag, so I have reverted to a more conventional hookup. The Cantatas are now wired normally (not biamped), and they are treated as "small" speakers by the Yamaha processor. This rolls off the bass fed to them below 90 Hz. at 24 dB per octave, meaning that the Cantata's own woofers are borderline inoperative. (Their own internal 6-dB-per-octave low-pass filter rolls them in at 80 Hz., but of course by then the Yamaha is cutting the feed to them considerably.) The result is midrange sound and low bass as good as what I got with the more complex hookup.

As I noted above and in the previous article, the Cantata midrange drivers normally run into the low-bass range (no high-pass filtering), having them doing more work than I would like. I prefer that my midrange drivers not also handle the low bass, so I do want the signals to them high-pass-filtered.

The existing set up does just that, and so the Cantata midrange drivers do their job more effectively and the very potent SVS subwoofer (itself equalized down low by the Rane THX 44 equalizer) easily handles the low-bass requirements. (The system continues to use a Hsu Optimizer equalized SVS 25-31PCi for the center-channel bass.) The downside of all of this is that the Cantata woofers have essentially become dead weight, and I hate dead weight in any system.

I'll live with the situation, however, and at least the slight but still obnoxious noise is gone.