Lawrence Phelps & Associates, A Corporation for Organbuilding

Organ and Sanctuary, One Musical Instrument

by Lawrence Phelps

Second in a series of lectures from a symposium on acoustics held at North Shore Congregation Israel, Glencoe, Ill.

Because Dr. Beranek's book refers specifically to concert halls, it seems to have escaped the attention of many that the subjective attributes it sets forth are indeed characteristic attributes of any room in which music is to be made, and churches are in no way an exception. The double standard that sets one group of criteria for evaluating musical performance and acoustical effect for a concert hall and a completely different set of criteria on a completely different level for music in the church has existed long enough. Certainly, there may have been a reason for accepting a somewhat lower standard of musical performance in the church a few years ago when the church was largely dependent upon voluntary, amateur personnel, but this is not the case any longer, since a large supply of well-trained professional musicians is available and growing in number every year.

It is not my purpose here to discuss the standard of musical performance which is continuously improving, but it is time that the conditions for musical performance in the church were brought into line with the standards required by professional musicians elsewhere. This is not to say that the same numerical values or ratings established for the various components of the subjective effect for concert halls must necessarily prevail in the church, but there is certainly no validity whatever for a church continuing to be exempt from the critical requirements of good musical listening.

Since taste is a large component in evaluating musical matters, let us set aside for a moment the purely musical aspects of the problem and consider the strange dichotomy that has existed in the church for years with respect to words spoken, on the one hand, and sung, on the other.

Does it make any sense at all to insist on a high rate of intelligibility for the spoken word while at the same time, in the very same room, no one seems to notice that the intelligibility of the words sung by the choir (blissfully singing to each other from one side of the divided chancel to the other, while their well shaped tones are lost in the cavernous chambers of the divided organ) is often very close to zero. Somehow, it just seems not to occur to people that the choir is sometimes singing words that are at least every bit as important as the text read from the Scriptures. I am sure that there are some places where the choir could go on Sunday after Sunday replacing the words of all sung texts with euphonious vowel sounds, without many who frequent the listening area being any the wiser. Yet, if the same choir were moved to the local concert hall and sang the same anthems, everyone would expect to hear the words clearly and indeed, would probably be quite critical if they did not. Is there something peculiar about the church that makes it an impossible place to apply the same discerning standards we insist on elsewhere? The only obstacle to applying these same values to the church is the much cherished tradition of indifference. The standards in the concert halls are maintained by the professionals who insist on them, while in the church, even if no one notices that things are not as they should be, there is no one to insist upon their being improved except the listener himself, and this means a lot of hard work.

Fortunately, every now and then, a valiant soul rises to the challenge, and then some great thing happens. It is for us who have to do with the planning and building of new churches and with building organs and making music in them, to be ready to rise to these occasions, when we are asked - to give the very best possible guidance, and this of course is why we are here today.

When Dr. Beranek's book first appeared, it was my hope that someone would soon show how the essence of its message was in fact a group of universal principles which were equally as applicable to the church as the concert hall, but, to date, so far as I am aware, no one has. Therefore, I would like to take a few minutes to examine some of the concepts set forth in its pages to see what they might mean in designing acoustics for the church.

Well, it only takes 22 pages for Dr. Beranek to get right down to that fundamental point that organ builders have been shouting about for so many years, but for a long time in vain so far as architects and acousticians were concerned. Right in the middle of page 22, he says: "A live room is acoustically superior to a dead room." Now every organ builder knows this is true because it is much easier to sell an organ if you demonstrate one in a lively church than in a dead church.

Dr. Beranek goes on to give us some very definite and more noble reasons, even though in subjective terms, why a live room is better for music.

First of all, he says that a live room has a "more uniform loudness," a more uniform distribution of sound. This is because the direct sound is heard combined with the reverberant sound, producing the impression on the listener that the sound is better distributed even at the rear of the hall where the volume level of the direct sound would actually have fallen off considerably. This is not to say, of course, that the sound would be the same in every seat in the room, but simply that it would be more uniform in a live room than in a dead room where one would hear only the direct sound and thus be much more aware of the decrease in volume level as one moved further away from the source.

A live room enhances the bass and treble, and the balance of the bass and treble can be effected by the suitable choice of surface materials and by placing reflecting surfaces strategically with respect to the original source.

A live room has more "fullness of tone" caused by the blending of one tone with another, but, of course, an overly-live room would tend to obscure the clarity of the music.

A live room makes possible a wide "range of crescendo" because the reflected sound is less apparent when the sound source is at a low level, but as the intensity of sound emitted from the source is increased, the room responds more fully and intensifies the growth of the sound from the source.

In a live room there is good diffusion of sound, and the listener feels that he is more or less in the source rather than observing it from a distance as is the case in a dead room where the sound comes all from one direction.

Now since all of these things - more uniform loudness, enhancement of bass and treble, fullness of tone, range of crescendo and sound diffusion - are acoustically beneficial to the production of and listening to music, and all of them are virtues of a lively room, I suppose there is some justification for those who are only interested in a minimum crash course in acoustics in putting Dr. Beranek's book down at the end of the second chapter, saying: "See? we told you so all the time; all you have to do is make the room live enough and everything will be fine." Indeed, this is what organ builders have been saying as a class for years.

However, those who are a little more curious and patient enough to go on and read Chapter 3, will find that lurking in the shadows of that single, simple and generally virtuous attribute of a room which we call "reverberation," there are some untoward attributes that can cause problems for anyone interested in music rather than just sound effects. For example, there is an attribute of music called definition or clarity which is affected considerably by the acoustics of a room, and definition comes in two kinds: horizontal and vertical. Horizontal definition has to do with how well sounds that follow each other stand apart when heard by the listener and vertical definition is a matter of how well sounds that are played together can be heard separately.

That wonderful virtue of a reverberant place, fullness of tone, Dr. Beranek tells us, is primarily determined by the ratio of reverberant to direct sound energy that reaches the ear of the listener. But horizontal definition is also determined by the ratio of the loudness of direct sound to reverberant sound, but in the inverse relation. So, here already, we find a conflict and we see that too much of a good thing - fullness of tone - is no longer a good thing because it reduces horizontal definition. Vertical definition is also related to the ratio of direct to reverberant sound, but it is also related to several other factors, such as the relative response of the room at various frequencies and the degree to which tones are blended in the area immediately adjacent to the source, (the stage enclosure in a concert hall, for example).

There is a quality of music called intimacy or presence, which is also affected by the acoustics of the room. And Dr. Beranek tells us that acoustical intimacy is determined by something called the initial-time-delay gap which is defined as "the difference in time of the arrival of direct sound and the first reflected sound."

Timbre and tone colour, which we normally attribute to sound sources themselves, that is, instruments or voices, are much affected by the acoustics and are much dependent on the reflective or absorptive characteristics of the room at various frequencies.

Loudness and dynamic range are also much affected by the room acoustics, and while both of these are much affected by the reverberant nature of the room, they are also very dependent upon the reflective efficiency of those surfaces near the source and upon the nature of the main ceiling.

In Chapter 4, Dr. Beranek develops a list of 18 "subjective attributes of musical-acoustic quality to be used in the evaluation of concert halls." This list includes all of these attributes I have mentioned and adds: warmth, brilliance, balance, blend, ensemble, response, texture, no echo, quiet, no distortion.

Warmth is defined as liveness of bass, or the fullness of bass tone relative to that of mid-frequency range. A hall is said to be brittle rather than warm if the reverberation at mid-frequencies is longer than that at low frequencies. He points out that most concert halls that are well regarded have a warm sound, a rich, full bass.

Brilliance is defined as a bright, clear, singing sound, rich in harmonics, and is affected by: the initial-time-delay gap, the ratio of reverberation time at high frequency to that at mid-frequencies, the distance of the listener from the performer, and the presence in the hall of suitable sound-reflecting surfaces.

Blend is defined as the mixing of sound from various sources in such a way that they seem harmonious to the listener. Dr. Beranek says, "Blend is partly a matter of the disposition of the orchestra which should he neither too wide nor to deep. Blend also depends on the design of the ceiling over the stage and on the presence of splayed surfaces that mix the sound before it emerges from the stage enclosure."

The attribute called ensemble is related to the ease of hearing among performers and affects their ability to play, or in the case of a choir, to sing well together. If the performance area is too wide and shallow, ensemble will be poor. The liveliness of the performance area makes an important contribution to ensemble.

The attribute called texture results from the pattern of reflections in the room and the sequence in which the reflected sounds arrive at the ear of the listener and are superimposed on the total effect of the other attributes.

My reason for presenting all of this material here is to be able to ask all of you just one question, which is this: Are any of the attributes mentioned here exclusively characteristic of a concert hall or do all of them refer equally well to churches? If they do, then I submit that the design of the acoustics of a church is not a separate thing requiring extensive studies and the establishment of a special scale of criteria heavily weighted with elements of mystery and magic, but simply a special case under the general subject of concert hall design. This will, of course, apply only to those churches in which music is performed and thought to be worthwhile. All other cases should be classified as a subtitle under lecture hall design, and it should be agreed from the start that if music is performed in them at all, it should be of the intimate, chamber type, and that the only keyboard instruments used should be the harpsichord or piano, or possibly a small positive organ.

The criteria for judging concert halls as presented in Dr. Beranek's book give us all that we need to work out a system for the acoustical rating of churches. Certainly the factors affecting the intimacy and definition will be somewhat difficult than in a concert hall, not just because we are dealing with churches but because of the nature of the musical literature used in churches and the conditions that were presumed in its composition. I have long advocated that the tonal design of organs should be based on the requirements of the existing literature, and I propose to be quite consistent and suggest that the acoustical design of churches should also be derived from the requirements of the literature to be performed. But just as I am no antiquarian where the organ is concerned, neither do I propose a purely antiquarian approach that would have us continue to produce churches that keep repeating the mistakes of our elders just because there happens to exist some music that was first performed under a particular set of conditions. Common sense and good judgment must always prevail if we are to learn the true lessons offered by history.

There are certainly other values to be considered in the building of a church than those affecting the acoustics. It is not just another room for listening or for displaying the virtuosity of musicians and preachers, but, on the other hand, it is not consistent with our age to consider the church as just a very special container for God and other mysteries. If music is to happen in it at all, then those acoustically functional considerations that will make it work well for music must be given much more attention than in the past. This may mean introducing into the church some of the paraphernalia now commonly associated with concert halls, such as stage enclosures and clouds. This has to date been done very rarely, and the only way to make these accessories acceptable to those who pay for them is to persuade them that they have virtues which surpass their strange appearance.

But deciding what should be done and actually getting it done are two different things, and perhaps we should have a special session sometime on the problem of the acoustical education of the general public. However, it should not be too difficult to successfully point out to the discerning members of any congregation that the intelligibility of their choir standing under a 40' high ceiling in some deep recessed area leaves something to be desired. If this ceiling happens to be covered with acoustical tile, the task is of course infinitely easier. However, a more basic and immediate problem is that the acceptance of the basic concepts involving the effective placement of the choir and organ in a church is far from unanimous among the professionals involved, and there is still a strong divergence of opinion among organ builders and organists as to just what manner of building an organ produces the most effective musical results.

I suppose I was invited here today to expose my views on this subject specifically, since they are well-known to the organizers of this meeting even though they are not shared widely by my colleagues in the field. Since this is primarily an acoustical session, I will discuss the matter of organ tonal design simply by repeating again that the tonal design of any instrument should be dictated by the literature it is expected to perform. This will mean that smaller instruments will be oriented primarily toward the Classical repertoire under which heading we summarize everything up through and including the baroque era, and that the consideration of the Romantic repertoire is the only justification for building an instrument beyond a certain size. By the time the number of stops in an organ reaches 40, we can begin to think of it as a general-purpose instrument, the major Romantic works requiring an instrument of upwards of 55 stops.

A reform movement began in European organ building in the mid-l920s, motivated by the realization that the new instruments of that time were completely unsuitable for the Classical repertoire. A separate reform movement motivated by similar interests started a little later in North America and developed more or less independently of the movement in Europe until after World War II. Since then, there has been a gradual tendency for the American movement to embrace more closely the aesthetic principles which have produced remarkably successful instruments in Europe. Briefly, these are as follows

1) The organ should be constructed completely within the room it is to serve and should be placed in an open, free position, preferably elevated and on the central axis of the church.

2) Each division of the organ should be a complete tonal entity within itself and spatial differentiation among the divisions should be established by putting them one above the other, where possible.

3) The principal stop of each division should he placed in front of the other stops, and each division should be separately encased.

4) The voicing of the individual pipes should be done in such a way as to encourage free, natural speech on full wind and with a minimum of nicking.

5) The wind pressure should he no higher than is necessary to produce the desired result with this type of voicing.

6) The scaling of the pipes should be developed for each instrument according to the musical and acoustical requirements of each individual situation.

7) The chests on which the pipes stand should be of the key-chambered type generally known as slider chests. And,

8) The key action should be entirely mechanical.

It is interesting, I think, that each of these points has implications which are both acoustical and musical. However, aside from the matter of placement, it is probably the casework which is more interesting for our discussion here today because it has a strong acoustical function and because it makes the organ a very strong visual element in the church, and therefore develops important architectural considerations.

The necessity for casework is not something that was imposed upon the movement simply because it was a prominent feature of the outstanding instruments remaining from the old master builders of the past. Actually, the European movement began with completely encased instruments.

It was the goal of the early builders of the reform movement to produce instruments that were as effective as those of the old masters, but, of course, casework was extremely expensive so they began their work hopefully without it, feeling that the essential effect of the old work was probably due to the scaling and voicing of the pipes rather than the physical arrangement of the instrument. However, this was soon found not to be true. Instruments which were built with identical scaling and voicing to some of the old instruments on a pipe-by-pipe basis did not, in fact, produce anything like the sound of the old organs.

The search for a more effective technique went through several stages, involving extensive experimentation with pipe arrangements and partial encasements of various types. However, when organ building was resumed after World War II, complete encasement soon became the order of the day, at least in the work of the more successful builders.

Although I had become disenchanted with the effect of exposed pipework as early as 1949, I did not have an opportunity to build encased instruments until I became tonal director at Casavant in 1958. In 1959, I built several encased instruments, beginning with St. Paul's in Dedham, Mass., where only the Positiv was encased. This was followed rapidly by the large instruments for Eglise des Saints-Martyrs in Quebec City which was entirely encased, and the First Church of Christ Scientist in Denver, Colo., where only the Great, Positiv and part of the Pedal were encased.

In these instruments, I found that the large size of the pitman chests deprived me of some of the benefits I was expecting from the casework. Then, after seeing several new instruments by the European builders, I realized that it was not possible to get the effect I was looking for until I was able to take advantage of the more compact pipe arrangements possible on slider chests.

I built several more encased electro-pneumatic instruments and, in 196l, produced my first completely encased mechanical-action organ. In 1962, I produced my last electro-pneumatic organ with complete encasement - the 72-stop, three-manual instrument for Notre-Dame du Cap in Cap-de-la-Madeleine, Québec. There can be no doubt that a close-fitting case greatly improves even an electro-pneumatic instrument on typical pitman chests, even though the cubic volume inside the case must be a great deal more than the more compact and more efficiently constructed mechanical-action instrument of the same number of stops. A large electro-pneumatic instrument, however, can entail case depths of 10' to 15', and we begin to experience some of the same problems of those encountered when working in a chamber: the rows of pipes near the front mask those further back, and there is not enough reflected sound from the case to enhance the first direct sound from the front pipes. This is partly because the walls are so far away that the energy level of the reflected sound is considerably reduced. Furthermore, the resonances of such deep cases with large cubic content are not so sharply defined and are perhaps too low in pitch to establish the sharp coloration that helps to provide an additional element of tonal differentiation between the divisions in a well-worked-out encased mechanical instrument in which the depth of the case need seldom exceed 3 1/2 ft.

I am often asked just what the difference is in tonal effect between an encased and an unencased instrument, and what causes it. It is one that is easy to hear; many people hear it immediately. The encased sounds win immediate converts. But, at first, some find it difficult to believe that the case I can make such a big difference, for after all, a well-designed exposed instrument is built very close to the walls, especially if using compact electric slider chests, and the reflective patterns from the wall around it seem to be not all that much different from those one would expect from a ease. Therefore, there is a tendency to feel that the real difference results from some sort of special technique used by the builder on instruments of this sort - perhaps more tender loving care in the voicing or perhaps in the selection and preparation of the pipe material, or perhaps almost anything except the case. And, is an encased instrument really all that efficient? After all, in an exposed organ, the sound seems to radiate from all around it, whereas in an encased organ, the sound comes out only through the front, or so it seems on superficial inspection. But, what really happens?

Well, first of all, let us take a look at what happens in an exposed instrument. What about the direct sound? Let us stop kidding ourselves about this. An organ has virtually no direct sound except from those few pipes which are in the very front of the instrument. Those at the rear are always masked by those in front and the sound we hear quickly enough to be called direct is what has managed to filter through after surviving an elaborate interference course involving numerous detours and short reflective paths. If the little pipes happen to be at the front and facing directly toward the listening area, they are virtually the only pipes from which we hear any really direct sound. If the organ is built close to the walls, naturally there follows not very far behind the direct and the not-so-direct-but-reflected-from-adjacent-pipes sound, the sound reflected from the walls, in turn followed by the sound from the ceiling, and much later by the reverberant sound from the more remote room surfaces. If the church is large and the organ is far from the listening area, the room considerably enriches the sound and the directness of the little pipes is not particularly apparent - or at least, not disturbing.

In this type of instrument, the organ and the church are literally one instrument, because all of the coloration, resonance, fullness of sound and projection must be provided by the walls and the room itself. The sound of the organ will have no resonant format other than that of the room. Any reverberant room will usually have a few frequencies at which there are definite peaks, and even in quite large rooms, these will usually lie in the mid-range of the keyboard. They can be detected often simply by playing tone clusters and listening to the decaying sound. Those pitches which die last are usually among the pitches at which the room is resonant. Some rooms have extremely sharp resonances at a particular frequency. It is clear that all of this room response is a very important contributor to the general effect of an unencased organ. And since no major stop in this type of instrument projects direct sound of a substantial nature, it is usually quite difficult to detect a difference between the direct and the indirect sound.

With respect to Dr. Beranek's vocabulary of subjective attributes, what does this mean? Well, obviously, it means that those attributes in which the ratio of direct to reverberant sound is important will be quite considerably affected. For example, if the room is quite reverberant fullness of tone should be quite well developed. But linear definition will not be particularly good because the reverberant sound is high with respect to the direct sound.

There are various means of improving this. For example, if the ceiling is very high above the organ, a reflective panel can be installed that will help to increase the number of early reflections, which will give the impression of increasing the direct sound if the panels can be fitted closely enough, but here we are getting into something very much like casework Also, we can attempt to improve the definition by increasing the strength of attack by leaving perhaps a little more of the attack transient, called chiff, in the sound of the pipes, but then this chiff is reflected by the closest but still rather distant wall surfaces and the attack effect is somewhat prolonged. Thus the definition again suffers. Or, worse still, the chiff, which is high pitched with non-harmonic components, is so much prolonged that it is heard to clash against harmonics or fundamental notes of higher pitched stops at the same approximate pitch. Sometimes the chiff from bass pipes is prolonged so badly that the beating against higher pitches is actually heard as an out-of tuneness which disappears only if the chord is held longer than a second. It can also give a shaky effect to the tone, somewhat like the effect of unsteady wind.

What about some of the other subjective attributes, such as blend and balance? If the room is quite lively, very likely the unencased instrument will have good blend. I however, many lively rooms, (particularly churches), tend to be more reverberant at the low and mid range frequencies than in the treble, and the room can still be considered to be quite reverberant even if there is a rather large amount of material such as carpet with a rather high rate of high frequency absorption. Such a room might be considered to be warm, but what happens to the treble line? In some churches where reverberation is high, the balance required for good vertical definition may not be present. In a four-part fugue, for example, the bass and tenor may be very clear, the alto also, but the treble can sound somewhat recessed, as if it were being played on a separate, rather distant instrument. This is one of the most common shortcomings in organs everywhere, and it is by no means only a condition to be observed in unencased instruments. However, it is generally a more serious problem in unencased instruments because one is completely dependent upon the frequency response of the room.

How does encasing an organ improve all of this?

First of all, and most obviously, the principal stop of each division is placed in the facade. Thus, even if there were no other pipes in the organ at all, the instrument at least has some direct sound throughout the complete range of each division. Of course, not all of the pipes of every principal can be placed directly in the facade, but the smaller pipes of the treble must be placed in an advantageous position, preferably between the feet of the larger pipes, so they also speak directly toward the listening area.

Secondly, it is clear that the ease being very close to the pipes, produces a first reflected sound in a much shorter time than is possible with an exposed instrument where the closest reflecting surfaces are two to three times further from the pipes. Thus, the instrument itself tends to improve the matter of linear definition because the time delay between the direct sound and the first reflected sound is very short. Also, the chiff, that is the attack sound of the pipes, is reflected by very close surfaces rather than by distant surfaces, and thus the attack transients are much shorter than with pipes voiced in a similar manner in a more open position. This shortening of the attack sounds tends to produce better articulation, which in itself improves linear definition.

Also, the closeness of the case tends to reinforce treble sound, thus helping to improve the balance of parts and vertical definition.

Then, there is the blending effect caused by the very strong group of early reflections from the case and the resonant format imposed upon the sound by the resonant frequencies of the case itself. The effect of this resonance is to give each section of the organ a distinguishing sound of its own, superimposed upon the sound of the pipes. In this respect, the pipes are only a means for exciting the resonance of the chamber, which then responds with a tone which is the characteristic of its size and shape. It is of course an oversimplification to compare this effect with the effect of the strings on the sounding-board of a piano or violin, or the reed of an oboe or clarinet on the resonant tubes of those instruments. Nevertheless there is an effect which is quite similar and which tends to reduce the "pipy" effect of the organ, tends to make each division an independent instrument with a better integrated sound of its own rather than simply a collection of related but highly individual pipe sounds.

There is a further advantage to be gained from the case in those situations which make it physically possible. Normally, we think of making the case of fairly substantial material so that its reflective qualities at rather low frequencies will not be disturbed. That is, we do not want the sound to be dissipated in vibrating the case and producing heat instead of sound energy. However, if the case panels are made of extremely thin materials - solid oak of say half a centimeter thickness - the sound of the pipes at low and mid frequencies virtually passes right through the case with very little loss in energy and travels on to the walls of the church to be reflected back to the listening area at a rather substantial delay. On the other hand, the panels are thick enough to reflect high-frequency sound directly so that the brilliance and clarity of the direct and early reflected sound remains unchanged. However, the richness of the low-frequency sound passing through the case and returning somewhat delayed to combine with the direct sound at these frequencies, gives an added warmth which is not possible in any other way and which is particularly characteristic of instruments built in this manner.

Of course, there is a certain range of frequencies just between those which are directly reflected by the case and those which I pass through the case with very little loss, where attenuation takes place. This can, to a considerable extent, be controlled by the scaling of the pipes themselves. Here rests the argument for the encased organ.

This type of construction also make the organ an interesting piece of furniture with a functional design that reflects its tonal composition and which will therefore tell us something about how the organ ought to sound even before we hear it. I have provided a collection of examples of encased instruments so that the architects present may see the very large variety of design possible and of which these examples actually present a very small portion. Actually, with this type of instrument, it is possible for the architect to have some considerable influence in the overall appearance of the instrument, which of course is not possible with an exposed organ where only the woodwork at the base presents any possibility for design on the part of the a architect.

Another advantage of the mechanical, encased instrument is its considerably smaller space requirements, an advantage which many architects are finding most appealing.

Also, the encased organ, so to speak, carries its own shell around on its back like a turtle and is thus more independent of the room in matters of blend and balance, warmth and brilliance. And the case goes a long way in providing good horizontal and vertical definition. There can be no doubt that, even in fairly dry acoustics, an encased organ sounds a good deal better than an exposed one, but, nevertheless, it is still largely dependent upon the room to provide such important attributes as fullness of tone, enhancement of bass and treble, uniform distribution and diffusion; thus it is still necessary for best results, even with an encased instrument, to provide an acoustical environment that is well designed to complement the instrument in every way This means that the space must be lively and that its reverberant character must be carefully planned to provide those subjective attributes that the organ cannot furnish for itself. And although reverberation is not (as we have seen) the only important characteristic, it is nevertheless important to determine just how long the overall period ought to be.

Many charts and tables have been prepared to attempt to answer this question, based on the conditions prevailing in existing structures generally thought to be good examples. I suggest a much simpler rule of thumb that I think will work out quite well and should be able to be universally applied. I propose that, as an absolute minimum, a fully occupied church should have no less than double the reverberation period normally thought to be ideal for a concert hall of the same seating capacity. A longer period might be tolerable, especially if a high level of direct sound can be achieved, as is usual with an encased instrument. Of course, the high level of reverberation that has in the past been demanded by some organ builders is completely unrealistic, both because of its being impossible in small rooms and because of the lack of clarity, even for organ music, that such long periods normally produce. If we apply the criteria for good musical effectiveness, very few churches of cathedral proportions qualify. Music in such "rolling" acoustics is seldom anything more than sound effects. Except for fullness of tone and warmth, such rooms rate very low on the scale of the subjective attributes of musical-acoustic quality. Of course, to those who prefer the awe-inspiring rumble of the cathedral, no argument in favor of music will have any meaning.

The placing of an organ is, of course, a very critical thing, and it is also closely related to the placing of the choir. The principles governing this whole matter are so obvious that I am amazed at the strange proposals that we are often asked to consider. If we can find some good solution to the problem of projecting the sound and the words of the choir better into the listening area while at the same time placing the organ in such a way that it is easily heard by the choir as well as by the audience, we have accomplished our task. Unfortunately, this is the area where the biggest amount of compromise always seems to be necessary. This is unfortunate because much good that has been done in the design of the building and its acoustics can easily be undone by a poorly arranged performance area for the music. It is astonishing to see the things that are suggested as more important than the effective placement of the organ and choir. Chief among these, of course, is stained glass which for some strange reason is considered to be more holy than an organ case. Years have been spent in the collaborative efforts of architects and acousticians and organ builders and church musicians in attempting to work out co-operative compromises.

I suggest that our efforts might be more fruitfully employed in the working out of co-operative non-compromises. For, after all, what can a compromise produce in a first-class situation except a second-class one?