-------------
DRAW DRAPES (HELP CONTROL TOTAL ABSORPTION) ROUGH-TEXTURED ACOUSTIC-TILE WALL SOUND ABSORBENT MATERIAL (UPOLSTERED CHAIR, RUGS, THROW CUSHIONS)
----------------
Fig. 901. A large drape can be very effective as a sound absorber
when placed opposite the speaker. Placed behind or alongside the speaker,
the drape will damp reflected sound waves, thus helping to deaden a
room that is too "live". (Photo Klipsch and Associates, Inc.)
WHEN acoustic problems arise in connection with a hi-fi installation, their solution may require control measures ranging from the trivial to the monumental, depending on the nature of the difficulty. What you do in desperate cases will depend a good deal on how far the client is able or willing to go, for almost any acoustic difficulty can be overcome if expense and trouble are not obstacles. Unfortunately, the authors have never yet encountered an installation where these factors did not enter the picture in some form, so everything discussed here must be weighed against these twin influences and explained fairly to the client, so he understands just what he's being asked to decide.
In all cases where a really serious acoustic problem must be solved, it is best to evaluate it carefully, dope out a proposed remedy and submit a cost estimate to the client along with your statement of the trouble and your proposed solution. After that, the decision is up to him. Don't go ahead and do the job and then bill him for it unless you have a specific contractual agreement stating that you are to proceed in that way, or you will very likely wind up not getting paid, and with a lost client to boot.
The reason for this caution is partly psychological. If the client hears the system only after you've cured the trouble, he may very well take the attitude that "that's the way it should sound, so why should I pay extra?" It is far better to demonstrate just what the problem is and get his okay before attempting corrective measures.
Another reason for the suggested procedure is that most professional installers are not fully trained professional acoustic engineers, to say nothing of the average "do-it-yourselfer" making his own installation. And if the problem is a really nasty one, say a resonating floor or a tissue-paper apartment-house wall leaking sound into the next tenant's quarters, it may take an acoustic engineer and the architect to solve it to everyone's satisfaction.
In the examples discussed in this section, we have tried to indicate about how far a competent installer should go in dealing with acoustic troubles. However, individual cases can pose unforeseeable problems and, in cases where the solution looks doubtful, you should not hesitate to suggest to the client that an acoustical engineer be called in. This recourse is likely to be justified only in fairly large and elaborate installations, such as in public places, institutions or the mansion of a very well-to-do client.
Room acoustics
Once the sound leaves the speaker and travels out into the air, it is at the mercy of the environment. This environment may consist of anything from a very modest living room or bedroom to a large auditorium, or even an outdoor stadium or amusement park, for many businesses and public gathering places have begun to recognize the appeal of high-fidelity reproduced music. For our purposes, we will limit the discussion to acoustical problems arising in small- to medium-sized rooms, up to perhaps the size of the dining room of a modest restaurant.
As we saw in Section 8, room reverberation time can cause trouble in sound reproduction it it's either too long or too short. If you are a professional installer, what you can do about it will depend entirely on what the client is willing to let you do. If you're doing your own personal installation, you are perhaps freer. But then, your wife may give you less latitude than a client would.
However, regardless of what is done or by whom, it must be considered in the light of room appearance and functionality, as well as acoustics. If the best acoustic treatment for a particular situation would impair the usefulness or appearance of the room, it's out and that's that. Try the next best thing. Don't ever lose sight of the room as a whole, and the living that goes on in it in favor of a single aspect. Hi-fi may be your full-time business, but to your client it is a part-time recreation.
Excessive reverberation
A room that is excessively reverberant (too live) is a room containing too much surface area that is highly reflective to sound.
In its simplest terms, the solution to this problem is to provide enough area of high absorption to increase substantially the rate at which sound energy will be used up in the room.
A number of materials will help to do this. Some, such as car pets or drapes, are commonly used in interiors for their decorative effect, and incidentally, happen to be helpful acoustically. Others are acoustic materials designed solely to provide certain sound controlling properties.
When trying to increase absorption in residential or commercial spaces where appearance counts heavily, do as much as possible with commonly-used decorative materials. Laying a substantially wall-to-wall carpet, particularly with under-padding, will add a large amount of absorption to a room that has a bare hardwood or tile floor. Often, this in itself is enough to bring reverberation time back within tolerable limits. A number of small throw rugs are not as effective as one large carpet. The larger the carpet and the deeper the pile, the more effective it will be. A carpet with an under-pad will be more effective than one without, and the under-pad is a good thing anyway as it increases the service life of the carpet. Once a heavy carpet with under-padding is down, there is very little else you can do with the floor to control sound in the room.
A rough-textured wall surfacing--rough plaster, textured paper, or fabric--will provide better absorption than smooth painted plaster, but not very much. The amount of improvement will vary with the coarseness of the texture and the softness of the surface, a coarse fabric being about the best thing short of acoustic plaster or acoustic tile.
Fig. 902. Installation of acoustic tile (a) by cementing directly
to ceiling. (b) by attaching to furring strips.
A large drape or lined tapestry can work wonders, particularly if it is on the wall directly opposite the speaker. (Fig. 901). The heavier and thicker the drapery material, the more effectively it will absorb sound. The efficiency of even a heavy material will be perceptibly increased if a lining is added, and further increased with interlining. As with a carpet, the larger the area of the drape, the more you can expect it to accomplish. Acoustically, a wall-to wall drape will do the best job and, fortunately, esthetically it is often easier to get a satisfactory effect this way than with partial draping. When a drape is not acceptable as a permanent feature of the room decor, it can be hung on a track that permits it to be run back to the corners when desired.
It is customary for some people to take carpets up and drapes down in the summer and put them back in the fall. If you hit a client like this in the summer, his reverberation problem will very possibly be solved by the normal fall replacement of these items.
Of course, what he can do in the summer is a nasty question, un less he has a reasonably-isolated private home and can leave the windows open so some sound energy can escape that way.
The old-style, large, overstuffed chairs and sofas were far better absorbers than the light, slim, trim modern ones, but there's nothing much you can do about that either, unless you revamp the listening room as a replica of grandma's parlor. This is not a solution that will appeal to many.
If normal decorative materials simply won't do an adequate job of reducing the reverberation time, you've no choice but to turn to specific acoustic materials and devices. There are, of course, a few people who would rather put acoustic tile on the ceiling than a carpet on the floor. They just don't like carpets.
An acoustic ceiling is fine as an absorber, but it involves much more labor than laying a carpet, even the wall-to-wall type. Quite a variety of patterns and textures is available in acoustic tiles.
Their coefficients of absorption also vary somewhat with different patterns, thicknesses and methods of mounting. But you needn't worry much about that unless you have a really fearful reverberation problem, in which case you probably need an acoustic engineer anyway.
The two most common ways of installing acoustic tile, shown in Fig. 902, are by cementing directly to the ceiling or by nailing or stapling to furring strips which have, in turn, been nailed to the ceiling. The furring-strip method is the better of the two acoustically since the coefficients of absorption are somewhat higher with this mounting than when tiles are applied directly to the ceiling. Also, the furring-strip method generally results in a better looking job, as the strips tend to bridge any irregularities in the ceiling surface, producing a smoother-looking surface upon completion.
If you are feeling energetic and would like to install an acoustic tile ceiling for yourself, either of the methods just mentioned are fine and will be within your capabilities. In addition, there are a half dozen other methods that we will not enumerate, as they should be handled by a contractor.
If, for some reason, an acoustic tile ceiling is undesirable, there is always acoustic plaster as an alternate possibility. While it will exhibit a rough surface texture compared with ordinary plaster, it will allow you to eliminate the patterns of acoustic tile, which may not be acceptable to your client. Acoustic plaster should be put up by a contractor familiar with the use of the material, as it must be handled differently from ordinary plaster.
Acoustic plaster can also be used to deaden walls as well as ceilings but, since it is nowhere nearly as hard as regular plaster, it can be readily chipped by kids, pets and carelessness. So you'll need to take extra precautions not to damage such a wall inadvertently, and the whole matter should be made clear to the client.
Echoes and standing waves
Seldom will any sound path in even the largest rooms considered in this guide be long enough to permit a genuine echo to show up, even if the room is fairly live. One of the authors one lived in a suite of rooms on the ground floor of an enormous old mansion that had, as a feature of the entrance hall, a huge circular stair case that began in a cavernous basement and rose through all four floors of the structure to an equally cavernous attic. When the double doors of the suite were open, a drummer's "rim shot" or other short transient sound resulted in three distinct echoes; one from the basement, one from the end of the hall, and one from the ceiling of the hall. A single loud chord of organ music gave practically the effect of a short scherzo movement, and an orchestral selection suggested a parade in which several bands at varying distances along the route were playing the same selection.
Closing the doors solved the problem very neatly.
It's unlikely you will have to grapple with a real echo problem.
When you do, it's nine chances out of ten the trouble occurs be cause of a very high ceiling, a reflective wall at the far end of a very long room, or a long, hard-surfaced hall or corridor opening directly into the listening room. If the trouble is due to a very high reflective ceiling, making the walls and floor more absorptive will help some, particularly the wall directly opposite the speaker.
This is cheaper than treating the ceiling and is worth trying with some jury-rig drapes, such as blankets hung up. Also try relocating the speakers. If they are centered in a long wall of the room, re positioning them in a corner adjacent to the main listening area may make a considerable improvement. Hanging tapestries high up on the walls will also help, but this brings in a decorator problem which should be discussed with the client.
As a last resort, the ceiling may be covered with acoustic tile; but this not only alters the decor, it costs like original sin if scaffolding is required to do the job. One other solution, and in some cases the best, is a heavy drape cut to cover the ceiling, and hung from metal tracks at each end in such a way that it can be drawn back or pulled across to cover the reflecting area, at the client's convenience. Such an installation in a splendid old California mansion solved a severe echo problem without permanently hiding a very beautiful ceiling fresco.
Standing-wave trouble is somewhat more common, since it does not depend on a long sound path. It usually crops up in symmetrical rooms that are rather live, particularly when there are few door or window openings or other wall irregularities, and little absorptive furniture. Standing waves make themselves apparent by producing dead and live spots in the room, particularly at certain frequencies. If you find, when feeding the audio oscillator through the system, that certain frequencies in the lower middle range (say from 100 to 1,600 cycles or so) are heard very weakly at certain points in the room and very strongly in others, you've got standing waves. Even the deadest room will show some standing waves if you search for them hard enough, but usually only very live rooms show the effect strongly enough to require corrective measures.
Two technical expedients will solve almost all standing-wave difficulties. These are relocation of the speaker in the room, and the addition of some absorbing material in a suitable place. Placing a speaker dead center in any wall of a symmetrical room is particularly bad, and should be avoided, if at all possible. In or near a corner is much better on several counts, and somewhat off center of the narrowest wall of the room may be almost as good.
The absorbing material, if used, should go on the wall opposite the speaker, and on either adjacent wall, if more is required. Carpeting a bare floor will also help, and the addition of a large item of furniture or two in the room will at least alter the standing wave pattern, and will often reduce the strength of such waves substantially.
Insufficient reverberation Insufficient reverberation means excessive sound absorption; i.e., a dead room. Both cause and cure are just the opposite of the excessive-liveness condition previously discussed. Somehow, you have to get rid of some highly absorptive areas and replace them with highly reflective ones.
On floors, for example, the area of carpeting should be reduced as much as possible. Better yet, remove the carpet entirely. If the floor under the carpet is in such hopeless condition that sanding and refinishing won't produce a satisfactory result, a wide variety of flooring tiles in asphalt, linoleum, rubber and vinyl is avail able to produce a good-looking floor that is also highly reflective.
If you're the energetic type, this is also a job you can expect to do satisfactorily yourself, if you care to save a few dollars.
If the ceiling is smooth plaster, leave it alone. You won't be able to do anything better with it. If some misguided experimenter has tiled it with acoustic tile, you'd best not try to remove it. Getting the tiles down and repairing the ceiling afterward is a major project. Besides, it may be needed there to control sound leakage into the room above, or vice-versa. Some types of tile lose part of their absorptiveness with a heavy coat of paint, so you might see if that helps. But otherwise, look elsewhere to increase reflectivity.
If the ceiling has been treated with acoustic plaster, a thin coat of regular plaster on top will greatly increase reflectivity, but again, don't try to knock the whole business down. It becomes a very sloppy job.
When you get to the walls, the idea is to bare as much smooth plaster, paneling or what have you as possible. Where there are wall-to-wall drapes, try to cut them down to a narrow strip on either side of the windows. If madam won't hear of this, at least get rid of the linings and interlinings, if any, and leave the drapes open as much as possible when playing music. Sometimes, when madam won't go along with eliminating or reducing the area of drapes, she'll at least permit switching to a lighter material. Fine! Every little bit helps. Sell her on the lightest and gauziest material you can.
At times, when the lady insists on keeping the drapes, you can at least get her to part with the valences. Even this is a small step Fig. 903. Loose display pieces of china, ashtrays and other knicknacks tend to rattle, but the noise will stop when a piece of felt padding is inserted under the offending member. (Harmon-Kardon, Inc.) forward. And don't overlook the possibility of getting wall drapes re-hung on tracks, so they can be drawn back to expose wall area when music is to be played.
You will also make a bit of progress if you can get some of the larger overstuffed upholstered pieces cut down or removed. This isn't easy, but perhaps you can at least get the aprons off the ones on legs, so a little air can pass under them.
Generally, you'll have your troubles in getting very far toward livening a room that is too dead because such a room is likely to be furnished in a fairly plush period style. By the time you've re moved enough absorptive material to help matters appreciably, you've probably ruined the decorative effect. If that's the problem, all you can do is take out what little you're allowed, crank up the gain and treble controls, and hope for the best.
Object resonances
As indicated in Section 8, these are elusive little devils and, what's worse, they are often difficult to stop even after you've located them.
Buzzes and rattles caused by ashtrays, china, vases and bric-a brac are comparatively simple to control. A felt pad or a doily under the offending piece will usually stop the trouble. If you don't want the pad to be visible, simply cut a felt pad to the shape of the bottom of the piece to be padded and cement it on with Duco or something similar (Fig. 903).
A rattle in a casement window is generally pretty simple too. A strip of self-adhesive sponge neoprene gasketing will do the trick nicely, and will also serve as a very effective dust and weather seal at the same time. The same trick will often work on hinged closet and cabinet doors when the door is rattling against an inside stop strip.
Where a hinged door is rattling on its catch, moving the strike plate in, sometimes as little as 1/16 inch, will stop the noise. If a cabinet has bullet catches on its doors and they rattle, there's no point fooling with the bullet catch. Add a spring-clip type catch inside and you can get a good positive hold on the door.
cause, even if you can manage to fill up the grooves enough to stop the rattle, you'll find that now the doors bind. Nylon glides are now being made for such doors that will prevent rattles. Remember them if you're planning to build some sliding-door cabinets, but they won't help you with an existing one that's giving trouble.
One scheme that will help is to glue a piece of fairly soft felt to the front surface of the rear door, in the middle of the edge that is hidden behind the front door when both doors are closed. If this felt is carefully trimmed with a razor blade to just the right thickness, it will usually stop the rattle without causing an intolerable amount of friction when the doors are opened.
Double-hung windows that rattle are also a nasty problem, for the same reason as sliding doors. If you get them tight enough to stop the rattle, you're likely to need a hydraulic jack to open and close them. Sometimes, weatherstripping consisting of a wood or metal molding with a felt edge, if properly installed, will stop these rattles and still let you open and close the windows (Fig. 904). You may have to keep renewing the weatherstripping as the felt wears down.
So many kinds of objects can resonate and rattle we couldn't possibly mention them all if we spent the rest of our lives at it.
In a good many cases, you're just going to have to swear and sweat until you can locate them, and then do the best you can to think of something that will shut them up. If it's any encouragement, about 90% are easy. It's the last 10% that starts you talking to yourself.
Structural resonance
Fig. 904. The inhalation of felt-edged weather strips around a double-hung
window will often help control resonance rattles.
Aha! We now come to the most persistent villain in the sound picture, particularly in buildings of certain modern low-cost types of construction. Many of them compare to older structures as a cracker box compares to a bank vault and, when a really good hi-fi installation is cranked up a bit in one of them, weird things can happen. One of the most common and most aggravating is for an entire wall, floor or ceiling to be highly resonant at some frequency in the middle or low range of audio frequencies. When ever this note comes along in the music, it booms out like the trumpet of doom, and maybe sets the dishes rattling in the kitchen, or the dog howling in the yard. Even notes an octave or two below it or adjacent semitones will excite it, and the sound will be heard on the other side of such a wall just about as well as in the listening room. Since few musical compositions consist of a single tone and harmonics thereof, the effect in the next room is even more aggravating than would be the case if the whole program leaked through on an equal footing.
Sometimes, relocating the speaker will help, particularly if the offending surface is the wall opposite the speaker, but this is seldom the complete answer. On a floor, a heavy rug will help, and the addition of a heavy piece of furniture as near the middle of the floor as the lady of the house will permit will substantially alter the frequency of the annoying resonances. On a wall, a heavy drape may help some if the frequency is fairly high. At low frequencies, the drape will not absorb enough energy to do the job.
A large, heavy cabinet shifted back against the offending wall may be the most effective expedient. Don't forget that if this is possible, you can enhance the effect by cutting a large piece of builder's deadening felt to fit the back of the cabinet, and tacking it on so that it forms a deadening layer between the back of the cabinet and the surface of the wall. The felt must come into solid contact with the wall, or most of the effect will be lost.
In houses of the popular dry-wall type of construction, one or more of the gypsum board panels may be excited into lively vibrating when the music is loud. In an interior wall, filling the space between the two wall surfaces with the type of insulating material that is meant to be poured or blown in by an air hose may do the trick. This can often be put in through one small hole cut in the panel near the ceiling. Be very careful how and where you make the hole, and how you patch it and replace the flap of wall paper or repaint the seam. Often, it's better to work from the room adjacent on the other side of the wall, if it is a closet or storage room. Outside walls will usually already have insulation between the surfaces, so this treatment won't work there. Even resonant ceiling panels can be controlled in this fashion, if you can find a convenient way to introduce the insulation.
Another point to remember is that where structural resonances show up, it is particularly important to isolate the entire speaker enclosure carefully from solid contact with the structure. A thick foam-rubber pad under the whole bottom of the enclosure, and a similar pad on the back if it normally touches the wall, are strongly recommended.
Sound leakage
When the sound of the reproduced program leaks out of the listening area into places where it's not wanted (say the nursery or a neighbor's apartment), something must be done to confine it or soak it up. The method used depends greatly on the manner in which the leakage occurs.
Air transmission
Most people with hi-fi systems whom we've dealt with are rather decent sorts, but nevertheless, most of the air-transmitted sound that gets into areas where it is not wanted and inconveniences innocent bystanders results from the hi-fi owner being careless, thoughtless or just an inconsiderate pest. A fellow who'll simply close. his windows and doors when he wants to crank up his system fairly loud has stopped most of the noise that might disturb someone else.
But most is by no means all. There are various other ways by which air-transmitted sound can get out to go wandering. Heating, ventilating and air-conditioning ducts can provide dandy pass ages through which sounds may travel astonishing distances. The factors determining where and how far sounds will travel in such a duct system become rather complicated. The size of the main ducts and the number and location of branch ducts, bends, vanes, and dampers as well as the positions of dampers, will affect how far sound will travel and where it will come out. If your sound is bothering someone by traveling through a duct system, the best solution is to call in an engineer qualified to specify the type and location of baffles to control it. If your sound travels through the ducts, other sounds probably do, too. Possibly you and the fellow ...
Fig. 905. An oversize hole around piping through which sound diffraction
might occur.
... you're bothering can enlist the support of some of your neighbors to approach the owner of the building and have proper controls inserted for the entire system. After all, one day someone else in the building may put in a hi-fi, at which time you're liable to be on the receiving end.
On occasion, overly generous steam or water pipe holes through a floor or wall (Fig. 905) permit sound to pass through the excess area of the opening and diffract into the adjoining space. No sizable amount of sound energy can pass through this kind of opening, but often the pipe itself helps, and the total may be bother some. Check to see if there's a loose hole around a pipe.
Sound transmission through solids you'll refer to Section 8, Fig. 814, you'll see the sound energy being radiated into a room breaking up into three components. One stays in the room, a second is absorbed by the wall and the third goes through the wall into the adjoining space. The last is the component we're concerned with now.
This kind of transmission often means real trouble, because it indicates that the basic construction of the wall is too flimsy. If it's a wood or metal frame partition with inside air spaces, you will get some relief by having rock wool blown into the air spaces, but this is a sizable job too. Covering both wall surfaces with acoustic tile will also help, if other considerations make this practicable.
An extensive cabinet treatment that attaches securely to the wall (Fig. 906) will increase the wall mass and thus help matters, but often this is not feasible unless more cabinet space is useful and it is permissible to locate it on that particular wall.
Transmission by direct conduction
When your speaker is mounted in a free-standing cabinet, this kind of trouble is usually easy to control. It usually happens be cause the enclosure is standing directly on a bare floor with no soft material intervening. Get some carpet under it and the situation will be markedly improved. Where there is carpet already under it and you've still got excessive transmission, simply isolate it further by using rubber mounts under the legs or base. And don't forget about padding between the back and the wall.
One stubborn case was not gotten under control until the entire enclosure was suspended on shock mounts from a bracket on the wall. But this was a particularly odd case of combined conduction transmission and structural resonance.
A speaker mounted directly in a wall can produce the most horrible combinations of conduction transmission and structural resonances on record. And the worst of it is that there is no way of effectively shock-mounting the speaker or the baffle board with out leaving some nice, short, direct paths for the back wave to come right around in front where it will ruin the bass response.
You may just have to pull the speaker out of the wall, completely enclose the back in an infinite-baffle arrangement, and sink the whole works in the wall, with some padding between enclosure and supports.
Fig. 906. An extensive cabinet treatment securely attached to a wall
will add considerable mass, enough to damp most of the wall resonance.
Enclosure and speaker factors In a book devoted primarily to installation, it should be possible to assume that all the equipment is in good order and working properly, including the speaker and enclosure. Alas! Would that it were true, but 'taint. Even the best speaker may be damaged in shipment or storage, and the same applies to enclosures.. Before we abandon consideration of acoustic problems, it's necessary to say a few words, none of them kind, about certain equipment troubles that may fool you into thinking they are really acoustic.
Speakers
A speaker with a torn cone or with an off-center or out-of-round voice coil will often give out with a ratty, scratchy sound that will make you think something in the room is vibrating, unless you scrunch down and jam your trusty analytical ear right into the grille cloth. Dirt or packing material in the voice-coil air gap (it can still happen in some speakers) will also produce a scratchy noise. It can be removed by putting some 60-cycle (for convenience) signal into the voice coil and blowing some air through the gap. If you use a compressed air hose, be very cautious. A slug of water or a metal particle in the stream will make a neat hole in the cone that few will appreciate. Even an excessive squish of air may tear something. If the speaker is actually defective, repair or re placement is the only course.
If you are mounting the speakers themselves in a built-in en closure or in a wall, don't even open the boxes until you are actually ready to fasten the speakers into place. That means you have not only finished the carpentry or whatever is required, but you've thoroughly cleaned up the shavings, plaster chips and other loose stuff that might conceivably get into the speaker mechanism.
And it should go without saying (but won't here) that you must phase speakers properly in multiple-speaker assemblies. The dry-cell test applied to the voice coil leads to see if they all jump for ward or backward together is tried and true. And to get tweeter and woofer to pump together at the crossover frequency, put in that frequency with your oscillator, and try the effect of reversing connections to either the woofer or the tweeter, but not both. The connection that gives the loudest response at the crossover frequency is the right one.
Enclosures
The best enclosure can occasionally let out a raspy buzz at certain frequencies. This means that something that should be held firmly in place, isn't. It could be the speaker itself (try the screws), the baffle board (try the screws), an internal partition or stiffening batten (try the screws and glue joints) or the back (try the . . . but you get the idea). Even screws that look and feel tight may stop the buzz if they're given another quarter-turn. In cabinets that seem to have too few screws too far apart, particularly those holding the back on, don't hesitate to add some more, if you trace the buzz to the offending member.
In horns and some other enclosures, internal structural members may be loose because of a poorly made or damaged glued joint.
Re-glue if possible, or add a glue block or two and maybe a couple of screws. And don't forget the possibility that the whole enclosure may be rattling against the wall or floor if there's no intervening padding.
One other odd-ball cabinet trouble is worth mention, not because it's very common, but because it may drive you out of your mind if you're unprepared for it. We've hit it three times in some 8 or 9 years ourselves. The first time, a cabinet buzzed atrociously at high volume, particularly on organ music. Every thing inside and out was gone over practically with a microscope.
No dice. Everything was sound and tight, but it buzzed. Switch speakers, it buzzed. Finally John Karlson suggested tapping the buzzer all over with a rubber-headed mallet. Eureka! There was the buzz, inside the wood of one panel forming the box. It was a loose part of a knot in one of the core plys of the fir plywood from which the enclosure was constructed. Sneaky, eh? We drilled a little bitty hole from the inside, injected some sticky stuff, let it set, and killed the buzz for keeps.