Room Acoustics for Stereo, January 1960 Electronics World

In the beginning, man created monophonic (mono) radios and phonographs that had sound with no spatial separation (left and right) in the source(s) and featured a single speaker. As such, except for being sure to not locate your radio or phono behind the sofa, sound perception at any point the room was fairly consistent - except maybe for volume level. Still, there was ample opportunity for the time of arrival due to multipath effects to distort the sound. Up until the 1950s or so, most homes had hardwood floors (with a few rugs) and rock-hard plaster walls to reflect sound waves, and rooms were relatively sparsely populated with furniture and wall hangings (look at photos in vintage magazines for proof), all of which provided means for distorted sound at a distance. And man said, "Let there be stereophonic (stereo) sound," which, while more faithfully reproducing sound to simulate a live setting, it also more than doubled the opportunity for multipath to distort the perceived sound. In 1960 when this article appeared in Electronics World magazine, a growing awareness of how a closed environment affected sound distribution helped to mitigate deleterious effects of multi-multipath (a term I just created) due to two or more speakers pumping out sound. More ardent audiophiles installed carpeting, ceiling-to-floor drapes over windows, acoustic ceiling tiles, tall leafy plants, and padded-fabric-covered furniture in order to minimize reflections. Stereo was huge in 1960s and 70s as documented by the plethora of articles about it in nearly every type of magazine.

Fig. 1 - An acoustically unbalanced room may cause the speaker system to sound unbalanced to the listener. Stereo speakers are at A and B, listener is at X.

There is an intimate relation between the stereo speaker system and the listening room. Here are the general principles to be followed for best results.

Good sound reproduction requires controlled acoustics in the listening room. For stereo listening, the room acoustic problem becomes even more critical because the reverberant room conditions and reflective wall surfaces may either greatly enhance the stereo effect or completely destroy it.

In achieving optimum stereo effect, there is an intimate relationship between the type of stereo speaker system used and the acoustic conditions of the room. Because of the many different types of speaker systems and the innumerable types of listening rooms, a simplified approach to the speaker-room combination will be covered. This approach may then be extended to diverse types of loudspeaker systems.

Proper speaker placement for high-fidelity stereo reproduction is considerably more complex than for high-fidelity monophonic program material. In single-channel reproduction we have but one problem as far as the loudspeaker is concerned, i.e., at which spot in the room will the loudspeaker perform best? With stereo we have a compound problem which involves: (a) how far apart shall the speakers be? and (b) how shall they be angled toward the listening area? This leads to the next point (c) which involves the best place to sit, stand, or recline to get the best stereo effect. Unfortunately, many articles in the stereo literature treat only lightly one additional factor that influences all of these considerations. This factor is the effect of the acoustical condition of the room in which the "stereorama" is displayed.

Fig. 2 - An arrangement, whereby improved acoustic symmetry among listener, speakers, and room is obtained by a rearrangement of room furniture of Fig. 1.

The conditions which characterize a good listening room, whether for monophonic or stereo service, involve two simple factors. The first is that the room be reverberant to the proper degree to give the reproduced sound "liveness." The room should not be too reverberant for then the reproduced sound would appear too cavernous, hollow, and indistinct. Nor should the room be overstuffed for then it would sound dead and somber. To a considerable degree, stereo systems themselves produce a psycho-acoustical revision of the absolute acoustic properties of the listening room. This, in effect, may make a room seem larger than it really is simply because of the apparent discrete localization of different sections of a large orchestra to separated areas within the room. The very essence of stereo sound reproduction thus "pushes" the walls of the room apart. Our minds automatically tend to convert this "spread out" reproduced orchestra into a "stage" for the orchestra.

Such psycho-acoustic enlargement, when listening to monophonic, or single-channel, program material radiated from one central speaker opening, would be virtually impossible. Some efforts are made to produce artificial enlargement of monochannel recordings at the recording studio by the judicious injection of controlled amounts of reverberation. There have been many "pop" selections recorded with this artificial reverberation added - but in many instances so poorly done that the music and singer are all but lost in the muddled "barrel of sound."

Thus, in monochannel reproduction, size of the room in "undirectional," that is, controlled in a "reverberation direction" but only by the reverberation inserted at the source. This is in contrast to stereo recording where spaciousness is not only injected at the source but is, moreover, under the control of the listener with respect to placement of the speakers and their acoustical environment.

The fact that there has been a fad for injecting reverberation into monophonic recordings, even though overdone, is indicative of the reeling that "spaciousness" does add psychological "depth" to our listening area. There are even electrical reverberation devices, intended for home use, which are designed to "liven up" (psycho-acoustically) the monochannel reproduction. In the case of monophonic reproduction, however, where the original source of the reproduced sound is a single cabinet, it- is more difficult to get a wide-stage psycho-acoustic effect without resorting to these highly accentuated effects incorporated in the original selection. It would be considerably better to make the recording under natural reverberation conditions and let the acoustic enlargement be obtained by means of the listening room environment. This, however, is a rather utopian ideal. Perhaps someday the standard living room will come with adjustable sound-conditioning facilities similar to the air-conditioning systems now in widespread use.

In the early days of stereophony, it was frequently noted that, when a switch was made from monophonic to stereophonic playback, to get the same "loudness" effect, the total gain of the amplifiers had to be reduced. Perhaps these were the first straws in the (acoustic) wind that tended to show that with monophonic reproduction the ear needed more reinforcement from its surroundings to get the same loudness effect as for the stereo reproduction. From this we may possibly conclude that a "live-r" reverberation condition is desirable for monophonic reproduction than for stereo reproduction. This leads us to the thought that it would be desirable not only to be able to correct our listening rooms in general so as to present the proper acoustic environment but to make this acoustic adjustment variable to accommodate changing conditions of playback methods, types of programs, or number of people in the room.

Stereo has the latent power to really put us in the concert hall, spatially, without monophonically overdone artificial effects if the speaker system is properly balanced in performance and is judiciously installed in a manner consistent with good concert hall practice. This means that we must look to our living room for the last link in the "realism" chain, with the aim of bringing it as close to good musical acoustic practice as possible.

In bringing the concert hall into the stereo room there is nothing we can do about the actual physical dimensions - in terms of feet and inches - of the listening room to approximate the concert hall. However, there is much we can do to change the psycho-acoustic dimensions of the room through adjusting its reverberation characteristics to a satisfactory "concert hall" level. Once this has been done, we can place the speakers so that they function best with the adjusted room conditions to heighten the illusion of expanded acoustic space.

The general principles of room acoustics are easily grasped. We sound very stentorian and robust when we sing in the bathroom because the hard, smooth surfaces bounce the sound around and around with little absorption so that the sound seems to persist loud (but not necessarily clear) as we continue to bellow to our heart's content. On the other hand, if we were enmeshed deep in a well-stocked clothes closet, we could shout until we turned blue with exertion and would still sound weak and feeble. The hard reflective surfaces of the bathroom make the room "live." The soft, absorbent "stuffing" of the clothes closet makes the closet "dead." In between these acoustical extremes lies an acceptable mean for good concert hall - or for that matter - listening room liveness. Our acoustic goal is to achieve a room condition which is live enough to give fullness and body to all the musical resonances yet not so live that recurrent echoes will blur the original rich sound. There have been many determinations as to the optimum degree of liveness for the concert hall, and these same principles may be readily applied to your own stereo listening area. However, before we illustrate the application of "acoustic room conditioning," with emphasis on stereo speaker placement, it would be helpful to examine the reverberation units with which we will be concerned.

Fig. 3 - Room acoustics are controlled by absorption properties of material surfaces.

If a room is "live," its boundary surfaces are very reflective. If it is "dead," they are highly absorptive. The common factor is, then, absorption. The great body of research that has laid the foundation for dealing with this factor of absorption (and reverberation) was done by Sabine. He proposed that the very logical, and very understandable, unit of absorption be, literally, an open window. Obviously, if sound is generated in a room and some of that sound reaches an open window, that portion of the sound that falls on the open window will leave the room entirely, as shown in Fig. 3. As far as the interior of the room is concerned, the open window completely "absorbed" the sound that hit it - for none of that sound found its way back into the room. A square foot of open space in a wall thus becomes an "absorption unit."

Nothing, but nothing, can quite equal the absorption of a hole in a wall looking out into the world. Whatever the material may be, if it is physical, if it can be weighed, calipered, touched - it will absorb some sound, reflect some sound, and transmit some sound. Substantive materials thus have individualistic absorption characteristics defined by their "absorption coefficient" in relation to "open space." Whatever the material being considered, its absorption is always less than "open space," a square foot of which is considered to have "unit absorption." Thus, the absorption coefficient of any material is always less than 1.

Referring to Fig. 3, the sound from the speaker system that finds its way out of the window will be completely lost as far as further usefulness to the room is concerned. It has, in effect, all been absorbed by the open window. On the other hand, the sound from the loudspeaker which hits the carpet is partly absorbed and partly reflected back into the room. Likewise, the sound hitting the plastered walls and ceiling is absorbed to a lesser degree and is reflected back into the room with more intensity than is the case with sound striking the carpet. As far as the listener is concerned, the aggregate sound that reaches him is composed of the direct ray A, the highly reflected ray B, and the somewhat reflected ray C. It is obvious, then, that the degree of liveness, or "concert-hallism," presented to the listener by his room is controlled by the absorptive properties of his surroundings.

Fig. 4 - There are many stereo speaker systems in use today as is shown above.

This problem of the room condition is important for any type of sound reproduction but for stereo it is doubly important (no pun intended). For the sake of preliminary illustration of the effect of the acoustics of a room on the resultant sound field of a stereo speaker system, we will use as the stereo sound source two speakers, 6 to 8 feet apart, oriented toward the central listening area by approximately 15 degrees each and we will assume that the speakers are as nearly balanced in their performance as is possible.

Now, even though we accept this "paired speaker" condition as being most suitable for our illustration, it should not be inferred that it is the only acceptable stereo speaker configuration. Much definitive work still needs to be done on speaker pairs (or trios), balanced or unbalanced, and we shall have considerable comment to offer concerning this problem after we have explored the preliminary example of the effect of room acoustics on the reproduced sound from the balanced and separated speaker system.

Suppose, now, that we place our perfectly balanced speakers, at the previously specified distance and angular relationships, in a room typical of many new home constructions where the living room and the dining area form a sort of an "L," as shown in Fig. 1. The speakers are neatly balanced on either side of the equipment cabinet. At the opposite wall is a foam rubber settee backed up against a hard plasterboard wall (decoratively papered, of course) with probably an arrangement of picture miniatures hung above the couch (a large, heavy picture would probably rip its supporting hooks right out of the plasterboard) . In any event, small pictures or large, the backing wall will be rather "hard" acoustically - it will reflect quite a bit of the sound that hits it.

We turn on our stereo system, lower ourselves onto the comfortable settee and listen. But although our reproducing speakers are balanced, phased, separated, and angled properly, the listening system is no longer balanced. Not even the channel-balancing control on the amplifiers or "stereo control center" will correct this acoustically unbalanced room.

Speaker "A" sends a direct sound to the listener via ray 1 and considerable wall-reflected sound to the listener via rays 2 and 3. On the other hand, Speaker "B," while it may transmit direct sound to the listener, sends very little reflected sound to his ears for there is the "open space" of the dining ell in front of it. Thus the sound from Speaker "A" will be louder to the listener. A second effect is that there will be considerable high-frequency loss from Speaker "B" as heard by the listener. The high frequencies from Speaker "B" that continue on straight ahead are almost completely lost to the listener while the highs from Speaker "A" come to him directly and by reflection. Consequently, although the speakers may have been perfectly balanced acoustically, their reproduction at the ear's location is unbalanced, with the unbalance being preponderantly in the high-frequency range, due to the differential reflective conditions bridging two speakers and the listener. This is unfortunate because much of the stereo effect is contributed by the higher frequencies. If we lose the highs, we lose the directional sense of the system.

Fig. 5 - A stereo reproducing system should be able to "place" the instruments in a manner duplicating their original location.

We may overcome these deleterious effects of the distortion of the direction of the sound by one of two methods - or preferably - a combination of both. The first obvious method is to rearrange the furniture; the second method is to alter the reflective properties of the room. Either of these problems may prove difficult to overcome unless one finds some way of convincing the lady of the house that it is actually her idea to rearrange the furniture, or to hang a drape here or there, or to install acoustic tiles hither and yon. Having thus easily (!) surmounted this first "managerial" problem, let us see what might be done with the furniture arrangement shown in the room of Fig. 1 to provide a better acoustic balance among the speakers, room, and the listener.

Fig. 2 shows one alternate arrangement of the speakers where almost complete acoustic symmetry among the speakers, room, and listener is available. The listener is in the direct way of each speaker. Each speaker also faces almost equivalent open-ended spaces, that is, Speaker "B" radiates into the dining ell and Speaker "A" radiates into the living room. To the listener, then, there is presented a condition wherein there is a balance between the direct rays from each speaker and the reflected rays.

There is also a second-order effect, but a desirable one. If the rooms are moderately "live" and the program is played at good volume (who ever heard of low volume?) - then an additional acoustic enlargement of the room takes place. The sound components from the speakers that travel into the dining ell area and into the living-room area find themselves bouncing around within these somewhat live end spaces. A "liveness" effect - or reverberance - is thus produced which adds to the direct rays received at the listening area. This partly gives rise to a psycho-acoustical enlargement of the listening room in a way that somewhat simulates the larger chamber music room - provided, of course, that the sound is reproduced loud enough to energize these reflective areas. This arrangement of Fig. 2, though quite severely limiting the effective stereo listening area, does provide a solution (for illustrative purpose) of balanced stereo listening.

Some Different Approaches Before we opened the discussion of the above problem of the acoustics of an unbalanced room upon a balanced speaker system, we deferred our comments about other types of speaker systems and configurations for stereo application. It would now be well to give this matter some consideration before going into the specific problem of room acoustical analysis and adjustment where these different types of speaker systems are used.

There are in use today speaker systems of many types which, as shown in Fig. 4, although greatly different, yield discernible and psycho-acoustically acceptable stereo. In addition to the two separate speaker systems (Fig. 4A), there are systems which consist of two full-range end systems with a "phantom fill" as a third channel in the middle (Fig. 4B). There is a system where the major bass information is all blended in one central system and the higher frequency stereo-determining components are displayed by end-placed outboard speaker components (Fig. 4C). There is also a system which is enclosed in one cabinet where the stereo effect is produced by reflecting into the room - from each of its back hinged end doors - the full program content of each channel respectively (Fig. 4D). In addition, there is a system where the lows are blended in one woofer and dispersed by the rear wall, and the high-frequency components of each channel are guided by the front hinged end doors to be randomly reflected by the room walls (Fig. 4E). Finally (at this writing, at least), there is the system where the blended lows and highs from one channel radiate from one system and only the highs from the other channel have a separate speaker (Fig. 4F).

Obviously, not all these systems are 'best" systems. They each have their individual merits and drawbacks. And obviously. they cannot all produce exact replicas of the original sound coming out of the instruments. But one thing they all do. They give some impression of the stereo effect, whether it be a correct or an incorrect one, to which we then add our own personally devised psycho-acoustic image of what we think the stereo reproduction should sound like - and we then have "stereo."

The question then arises how do we treat all these systems when integrating them to the acoustical condition of a particular room. Unfortunately, there is no definite and rigorously circumscribed "best" relation between the speakers of these various systems and any room at random. The number of solutions would obviously approach infinity, depending upon room size, shape, acoustic condition, speaker system, the listening habits of the auditor, and the type of music being played.

Fig. 6 - (A) Higher frequency radiation should overlap for best results. The overlap area is determined by speaker separation and angular orientation. (B) Graph of maximum stereo listening area as a function of speaker separation and orientation, based on radiation from 12" loudspeakers.

To briefly illustrate the complexity of the problem, take the last item just noted-the type of music being reproduced in relation to speaker system and room acoustics. If the stereo program material were that of a string quartet, we would expect, for stereo realism, that the stereo effect encompass a rather small area governed by the usual intimate geometrical relation among the two violins; viola, and cello, as shown in Fig. 5. If the speaker system were an adjustable one in the matter of speaker separation and speaker orientation - as would be possible with the units of Figs. 4A, 4B, 4C, and 4F - then after auditing a few programs the listener might be able to strike an arrangement where the stereo effect produced would be fairly close to the original sound distribution. The arrangement in Fig. 4D could also be acceptably adopted so that the doors, hung from the back edges, and properly angled, would project the quartet image into a centrally confined area with a minimum of deteriorating acoustic splash from the side walls which would otherwise tend to swell the "size" of the quartet. Fig. 4E would probably not be as adaptable to this type of music for, at best, even with the front-hung doors completely closed (onto the front) the stereo-determining middle - and high-frequency components would still be strongly projected towards the side walls of the room and would reach the listener with strong reflected level from these comparatively distant areas with the result that, again, the quartet's geometrical reproduction would be "swelled."

On the other hand, if the reproduced program were a full-bodied and large symphony orchestra, then this last sys-tem with its doors wide open, would direct the sound toward the outer reflecting wall to give a resultant enlarged acoustical image more in keeping with the original effect. The same effect could, of course, be obtained with the other systems through greater separation of the individual speakers, con-trolled center fill, and system angling. Perhaps this brief dissertation on the correlation of the reflecting properties of the room, the speaker system philosophy, and the program material will bring home the difficulty of achieving one perfect solution-obviously a single solution to suit all conditions is out of the question. But, despite the complexity of the problem, or rather be-cause of its complexity, we must make some simplifying assumptions in an effort to approach analytical results which may then be applied to more complex systems.

Even if we limit ourselves to analysis of the stereo effect of two identical systems, we may run into an imponderably large number of physical geometrical relationships between the two loudspeakers that do not necessarily bear any definite relation to the room characteristic. In order to resolve this matter of the placement of two identical systems in order to get the maximum stereo effect, the writer undertook to examine the matter in an analytical fashion in a paper which was presented before the Audio Engineering Society Convention in September 1958. The analysis endeavored to determine how far apart two identical speakers should be placed and how they should be angled toward each other in order to obtain the maximum area in which the stereo effect would be perceived when the individual speaker system consisted of a simple 12" speaker. One of the results of this analysis is shown in Fig. 6 from which it is seen that for such a system the maximum stereo listening area will exist when the speakers are placed about 7 feet apart and are both toed-in by about 15 degrees from the wall against which they are mounted. Now, obviously, this result cannot be accepted as gospel truth for all of the other types of speaker systems previously illustrated but it does at least give us a toe-hold on a premise that will permit us to discuss the acoustics of a room and its adjustment in terms of a frequency space pattern of which we can be reasonably sure.

After we have gone more intensively into the study of the room itself and its acoustics, we may then be able to convert these findings for the balanced system for application to those situations where the composite type of system is to be employed.

Room Acoustics for StereoJanuary 1960 Electronics World

Room Acoustics for Stereo
January 1960 Electronics World