CORNELL UNIVERSITY LIBRARY 
 
\1 
 
 The original of this book is in 
 the Cornell University Library. 
 
 There are no known copyright restrictions in 
 the United States on the use of the text. 
 
 http://www.archive.org/details/cu31924101414914 
 

 SOUND 
 
 REFLECTIONS 
 
 
 
 . 
 
 
 
 
 3^^^ 
 
 9||v 
 
 Ki»^ 
 
 1 
 
 • PUBLISHED BY 
 
 JUNIUS H. STONE CORPORATION, 1400 BROADWAY, New York 
 
 VOL.1 
 
 No. 1 
 
FOREWORD 
 
 WE are undertaking in "Sound Reflec- 
 tions," to present the accumulated 
 knowledge gained by our Acoustical Engi- 
 neers in extensive practical experience, to- 
 gether with such other available information 
 on the subject as may be interesting. 
 
 Certainly architects are entitled to this 
 information, since it is to their vision and 
 courage that credit is largely due for the 
 development of the Science of Architectural 
 Acoustics to its present usefulness. 
 
 We hope it will be helpful; to that end 
 and the encouragement of our purpose, com- 
 ments, criticisms, suggestions or experiences 
 will find a cordial welcome at all times. 
 
 \ President 
 
 [3] 
 

 ACOUSTICS 
 
 5y George C. Hannam, M. E. 
 
 Manager, Dept. of Architectural Acoustics, Junius H. Stone Corp. 
 
 New York City 
 
 lis position of importance in every-day life; acoustics 
 formerly a gamble; attacking the subject scientifically; 
 a definite scientific basis established; factors governing 
 acoustics;echo; interference; reverberation; a reterberation 
 problem worked out; wire stretching; other fallacies. 
 
 ^t'O the average engineer, the 
 %hJ/ subject of acoustics does 
 not loom large in the every-day 
 matters of a practical world, 
 but when it is realized that it is 
 an important consideration in the 
 design of practically every court 
 room, church, music hall, theatre, 
 auditorium, legislative chamber, 
 bank and even office building, the 
 importance of a knowledge of the 
 subject will be appreciated. Faul- 
 ty acoustics might be found to be 
 the real reason for the failures of 
 some of the most capable actors, 
 divines and politicians, the world 
 has never had a chance to know. 
 Many a speaker and singer has 
 been severely criticised, thereby 
 suffering considerable loss of pres- 
 tige, when the fault has been en- 
 tirely due to the faulty acoustical 
 conditions. 
 
 There are thousands of audi- 
 toriums in the world that are 
 acoustic nightmares. The late 
 Professor Wallace C. Sabine, of 
 Harvard University, said that 
 the reverberation which accom- 
 panies lofty and magnificent 
 
 architecture increased until even 
 the spoken service became in- 
 toned as a Gregorian chant, and 
 that it is not going beyond the 
 bounds of reason to say that in 
 those magnificent churches in 
 Europe, which are housed in 
 magnificent cathedrals, the Cath- 
 olic, the Lutheran and the Pro- 
 testant Episcopalian, the form of 
 service is in part determined by, 
 the acoustical condition. 
 
 Acoustics Formerly a Gamble 
 
 In the past, it has been cus- 
 tomary in the vast majority of 
 cases to design theatres, churches 
 and auditoriums in general, build- 
 ing them at considerable expefase 
 to secure strength, desired seating 
 capacity, noble architectural lines 
 asthetic illumination, wholesome 
 ventilation, etc., and then gamble 
 as to whether the acoustics in the 
 resulting auditorium would be 
 good or bad. 
 » 
 
 * Luncheon Address, August 31, 1920, 
 Engineers' Club of Philadelphia. 
 
 [4] 
 
 -* 
 
A Definite Scientific Basis 
 Established 
 
 Before 1895 but little definite 
 information was to be obtained 
 by an architect on tlie subject. 
 He was guided in large measure 
 by his own limited experience or 
 by studies made in existing build- 
 ings. The late Professor Wallace 
 C. Sabine, of Harvard University, 
 is to a considerable extent respon- 
 sible for most of our knowledge on 
 this subject to-day. He began 
 his researches in 1895 and in 1900 
 published in The Engineering 
 Record the results of his five 
 years' work. This was the first 
 real contribution on the subject, 
 and due to the scope of the work 
 and the fact that a mathematical 
 basis was established, the former 
 necessity of chance and guess 
 work was done away with. For 
 eighteen years after the publica- 
 tion of this paper he . continued 
 his researches, following the pro- 
 gram of investigations there out- 
 lined. 
 
 The knowledge gained from 
 these researches and from their 
 practical application, augmented 
 by contributions from other phy- 
 sicists, has made it possible to 
 determine from the plans of an 
 auditorium, with a knowledge of 
 the materials of construction, 
 what the acoustical conditions in 
 the finished building will be and 
 to make such changes as might be 
 necessary to overcome faults. 
 
 The simplest type of audi- 
 torium imaginable is a level plane 
 with a single person as audience. 
 
 In this case the sound spreads in 
 a hemispherical wave, diminish- 
 ing in intensity as it increases in 
 size. If instead of one person, 
 there is a large audience, the 
 intensity decreases more rapidly, 
 due to the absorption occasioned 
 by the clothing of the people. 
 The upper part of the sound from 
 the speaker in such an audito- 
 rium is entirely lost. The first 
 improvement that suggests itself 
 is to elevate the speaker. The 
 next step is to have the ground or 
 floor sloped so that each row of 
 persons will be slightly elevated. 
 Then, a wall should be placed 
 back of the speaker to reflect 
 that part of the sound towards 
 the audience. We now have the 
 design- of the old Greek amphi- 
 theatres. With the addition of 
 a roof to prevent the loss of the 
 upper part of the sound waves 
 and the construction of galleries 
 to elevate and bring further front 
 part of the audience, we have the 
 typical form of our modern au- 
 ditorium. 
 
 Factors Governing Acoustics 
 
 The size and shape of an audit- 
 orium determine the distance 
 traveled by the sound between 
 reflections; while the materials 
 used in the construction deter- 
 mine the loss at each reflection, 
 for which reason it is possible for 
 two rooms designed exactly alike 
 to have entirely different acous- 
 tics. An illustration of this is the 
 well-known case of Sanders Thea- 
 tre in Cambridge and the audit- 
 
 [Sl 
 
orium of the Fogg Art Museum 
 of Harvard College. Sanders 
 Theatre is an old building noted 
 for its good acoustical properties. 
 When the Fogg Art Museum was 
 designed, the architect was in- 
 structed to make the auditorium 
 a replica of Sanders Theatre. It 
 was taken for granted that similar 
 design assured similar acoustical 
 conditions. When the audit- 
 orium was completed, everyone 
 was greatly surprised to find the 
 acoustics almost impossible. As 
 a result of the development of 
 the subject since that time, the 
 reason for the difference is well 
 known. Sanders Theatre was 
 constructed of comparatively soft 
 materials. All seats were heavily 
 cushioned. The other audit- 
 orium was of fire-proof construc- 
 tion. The change produced in 
 the absorbing power of the vari- 
 ous surfaces was mainly respon- 
 sible for the difference in acous- 
 tical conditions. 
 
 Most people are acquainted 
 with the wonderful acoustical 
 conditions of the Mormon Taber- 
 nacle. From an ac6ustical stand- 
 point the design could not be 
 much worse. However, the build- 
 ing is constructed of wood, which 
 has the highest co-efficient of 
 sound absorption of any building 
 material, and it is due entirely to 
 the presence of so much wood 
 that the acoustics are satisfac- 
 tory. The shape of the ceiling is 
 responsible for the remarkable 
 echo. A replica of this building 
 with the use of hard materials 
 
 would result in an acoustical 
 horror. 
 
 If a sound be produced in an 
 empty room, having exposed sur- 
 faces that are absolutely rigid, it 
 will last almost indefinitely — 
 that is, if we disregard the loss 
 due to the viscosity of the air, 
 which in practical cases is neg- 
 ligible. The sound will travel 
 back and forth from one surface 
 to another, and if the surfaces are 
 absolutely rigid, the original in- 
 tensity of the sound would be 
 maintained. Of course, no build- 
 ing materials are absolutely rigid. 
 Taking an open window as being 
 totally absorbent, and calling its 
 co-efficient 1.00, Professor Sabine 
 in his researches determined the 
 co-efficient of sound absorption 
 of most of the materials employed 
 in building construction. 
 
 The co-efficients of some of the 
 more common materials are given 
 below : 
 
 1 square foot each — 
 
 Open Window 1.000 
 
 Wood sheathing (hard pine) 061 
 
 Plaster on wood lath 034 
 
 Plaster on wire lath . .033 
 
 Glass 027 
 
 Plaster on tile .025 
 
 Brick 02S 
 
 Concrete 015 
 
 Glazed tile 01 
 
 Following are some miscella- 
 neous co-efficients which are of 
 interest: 
 
 1 sijuare foot each — 
 
 Carpet 20 
 
 Cheese-cloth 019 
 
 Cork 2.5 centimetres thick, loose 
 on floor 16 
 
 [6] 
 
Audience per person 4.7 
 
 Hair Felt 1" thick S3 
 
 It is frequently of assistance to 
 compare the reflection of sound 
 waves with the reflection of Hght 
 waves. A white ceiling will re- 
 flect about ninety-five per cent of 
 the light striking it. The same 
 surface painted a dull black will 
 reflect at the most but five per 
 cent of the light. A glazed tile 
 ceiling will reflect pinety-nine per 
 cent of the sound striking it. The 
 same surface covered with one 
 inch of hair felt (the best sound 
 absorber) will only reflect forty- 
 seven per cent of the sound. 
 Increasing the thickness of the 
 felt will further reduce the per- 
 centage of sound reflected. 
 
 Echo 
 
 Echo is a special case of rever- 
 beration. Ordinarily the pro- 
 longation of a sound after the 
 source has ceased vibrating is 
 continuous until it becomes in- 
 audible. When the sound, in- 
 stead of being maintained, is 
 short and sharp with a noticeable 
 interval between the direct sound 
 and the reflection, an echo is pro- 
 duced. It is necessary in the 
 case of the average ear for the 
 time interval to be greater than 
 one- twelfth of a second. Pro- 
 fessor Watson, of the University 
 of Illinois, in writing about the 
 acoustics of the University Audit- 
 orium, mentions an interesting 
 condition that developed during 
 a concert being given by the 
 University orchestra. The num- 
 
 ber was a xylophone solo with 
 orchestral accompaniment. The 
 echo was so pronounced that at 
 one time the leader of the or- 
 chestra beat time to the echo, 
 whereas the players who were 
 immediately around the xylo- 
 phone kept time to it. Those 
 farther away kept time with the 
 leader, and the resulting con- 
 fusion can well be imagined. 
 
 Curved surfaces should be 
 avoided in the design of an audit- 
 orium as far as possible. When 
 used, the center of curvature 
 should be located a considerable 
 distance from the location of 
 seats. The harmful efltects pro- 
 duced by a dome, curved ceiling, 
 pendentives, etc., cannot be en- 
 tirely overcome by acoustical 
 treatment because such surfaces 
 cannot be made totally absor- 
 bent. Surfaces covered with felt 
 one inch thick will absorb fifty- 
 three per cent of the sound energy 
 striking it. Forty-seven per cent 
 will still be reflected, which is 
 sometimes sufficient to be a source 
 of complaint. 
 
 Interference 
 
 So-called "dead" spots in an 
 auditorium are caused by inter- 
 ference. A body in vibration 
 sends off a wave of condensation, 
 which is immediately followed by 
 a wave of rarefaction. As long as 
 the vibration continues, these 
 waves follow each other. Should 
 sound waves traveling by differ- 
 ent paths come together again, 
 the condensation wave of one 
 
 [7] 
 
will meet the condensation wave 
 of another, causing the sound at 
 this place to be re-enforced; or, 
 should one path be slightly 
 shorter than the other so that the 
 condensation wave of one meets 
 the rarefaction wave of the other, 
 there will at this point be no sound. 
 An interesting example of this 
 phenomenon is the tidal inter- 
 ference observed on the Tongking 
 Peninsula in Asia. The tide of 
 the Pacific Ocean enters the 
 Chinese Sea through two chan- 
 nels. One of these channels is 
 broken up by many small islands, 
 which force the tide in going 
 through to follow a tortuous 
 course, greatly retarding it. The 
 other channel is short, deep and 
 practically unobstructed, causing 
 but little retardation. The tide 
 coming through these two chan- 
 nels produces an effect which 
 varies from place to place. At 
 one port on this peninsula the 
 tide coming from one channel is 
 just six hours later than the tide 
 from the other channel. It is, 
 therefore, high tide from one when 
 it is low tide from the other. As the 
 two tides are equal in size, they 
 neutralize each other so that there 
 is no change in the water level. 
 
 Reverberation 
 
 Of acoustical defects, the most 
 common is excessive reverbera- 
 tion. A sound produced in a 
 confined space will travel from 
 surface to surface until it is either 
 transmitted by the walls, or is 
 transformed into some other 
 
 form of energy, generally heat. 
 This process is called absorption. 
 The duration of a sound after the 
 source has ceased vibrating is 
 called reverberation. When the 
 duration of audibility of the 
 standard sound used exceeds the 
 ideal amount, the reverberation 
 is excessive. By using the form- 
 ula developed by Professor Sabine 
 it is possible readily to determine 
 the reverberation for any room 
 from the plans. From experience 
 I know the desirable reverbera- 
 tion in various sized rooms used 
 for different purposes, and can 
 readily figure by means of the 
 same formula the number of 
 sound absorbing units it is neces- 
 sary to introduce to obtain this 
 condition. 
 
 In the average room used only 
 for speaking, when the volume is 
 approximately 150,000 cubic feet 
 the reverberation under average 
 audience conditions should be not 
 more than 1.9 seconds nor less 
 than 1.3 seconds. For most 
 forms of music it is desirable to 
 have the reverberation exceed 
 2.1 seconds. When a room is to 
 be used for both speaking and 
 music, it is usual to compromise, 
 having the reverberation slightly 
 excessive for ideal speaking con- 
 ditions, and slightly less than 
 that demanded for ideal musical 
 conditions. The solution can be 
 varied, of course, to suit the 
 special conditions presented by 
 each case. 
 
 As the volume increases it is 
 necessary to increase the dura- 
 
 [8] 
 
tion of reverberation. Unfortu- 
 nately, a reduction in the rever- 
 beration produces a correspond- 
 ing reduction in the intensity. 
 For this reason, in a room having 
 a volume of say 400,000 cubic 
 feet, it would not be advisable to 
 reduce the reverberation below 
 2.7 seconds. This duration of 
 reverberation is slightly exces- 
 sive for an untrained speaker but 
 necessary to insure sufficient in- 
 tensity in the furthermost parts 
 of the room. 
 
 A Reverberation Problem 
 Worked Out 
 
 To illustrate the method of 
 calculating the reverberation and 
 the determining of the area of 
 acoustical treatment required, I 
 will take the case of a court-room, 
 60' wide, 100' long, and 20' high. 
 The following assumptions are 
 made: 
 
 Floor, Concrete. 
 Side-walls, Plaster on Tile. 
 Ceiling, Plaster on metal lath. 
 Area of glass windows, 500 square 
 
 feet. 
 450 seats. 
 Correct conditions desired, when 
 
 70 persons are present. 
 
 Sabine formula for reverbera- 
 tion is: ^j 
 
 r = k- 
 a 
 
 where "r" equals reverberation in 
 seconds; V equals volume in cubic 
 feet; "a" equals open window 
 units of sound absorption; "k" is 
 a constant and is .05 when foot- 
 second units are used. A sound 
 is used having an intensity of 
 
 1,000,000 times the minimum 
 audible intensity and of pitch 
 one octave above Middle "C" on 
 the piano. 
 
 Area concrete floor 60' x 100' = 6,000 
 
 sq. ft. 
 Co-efficient of absorption of concrete 
 
 is .015. 
 6,000 sq. ft. X .015 =P(? units. 
 Area side-wall surfaces (60-1-60-1- lOO-f 
 100) 20 = 6400 sq. ft. 
 Area of glass windows = 500 sq. ft. 
 
 Net area plaster on hollow 
 tile =5900 sq. ft. 
 
 Co-efBcient of plaster on hollow tile is 
 .025. 
 
 5900 sq. ft. X .025 = 147.5 units. 
 
 Area of glass 500 sq. ft.; co-efficient 
 .027. 
 
 500 sq. ft. X .027 = 13.5 units. 
 
 Area of ceiUng 60' x 100' = 6,000 sq. ft. 
 
 Co-efficient for plaster on wire lath is 
 .033. 
 
 6,000 sq. ft. X .033=198 units. 
 
 450 seats, co-efficient .1. 
 
 450 X .1=45 units. 
 
 70 persons, co-efficient 4.7. 
 
 The co-efficient for one person is taken 
 at 4.6 due to the fact that the chair 
 occupied by each person has already 
 been considered at .1 and when occu- 
 pied its value as a sound absorber is 
 almost nothing. 
 
 70 persons x 4.6=322 units. 
 
 Total number of units, 90-|-147.S4- 
 13.5-M98-|-4S-|-322 = W(5. 
 
 Volume of room 60' x 100' x 20' = 
 120,000 cubic feet. 
 
 Substituting in formula r = k 
 V 
 — , the following results are 
 
 Si 
 
 obtained : 
 
 r = .05 X — pi , = 7.34 seconds, 
 olo 
 
 Experience has shown that the 
 ideal reverberation for a court- 
 room of this size is 1.5 seconds. 
 A reverberation of 7 seconds is 
 
 [9] 
 
very excessive, and it would be 
 difficult, if not impossible, to 
 hold" court in such a room. An 
 increase in the number of persons 
 present will accomplish a reduc- 
 tion in the reverberation due to 
 the additional clothing intro- 
 duced. The reverberation can 
 readily be determined for all 
 audience conditions. 
 
 Substituting l.S seconds — the 
 ideal reverberation — ^for "r" in 
 the formula and solving for "a", 
 
 '^^^^^\ 120,000 
 
 1.5 = .05 X — - — 
 
 "a" = 4,000. 
 
 This shows that 4,000 absorptive 
 units are required in the room to 
 have a reverberation of l.S sec- 
 onds. There are 816 units al- 
 ready in the room, so that 4,000 
 minus 816 equals 3184 additional 
 units which are required. These 
 additional units can be supplied 
 by means of acoustical treat- 
 ment, which consists of felt, 
 usually one inch in thickness, 
 fastened to certain of the avail- 
 able surfaces of the room._ The 
 felt is concealed by tightly 
 stretching a cloth membrane 
 one inch away from it. The 
 cloth is painted or dyed the 
 desired color. Only special paints 
 can be used for this purpose. 
 Each square foot of treatnient 
 has a co-efficient of .53. Divid- 
 ing the number of additional 
 units required, 3184, by the co- 
 efficient of absorption for each 
 square foot of felt, .53, we get 
 the number of square feet of 
 
 treatment required, namely, 6,007 
 sq. ft. The ceiling area is 6,000 
 sq. ft. so that treatment applied 
 to this surface will accomplish 
 the desired results. 
 
 Wire Stretching 
 
 I am frequently asked regard- 
 ing the value of stretching wires 
 to overcome acoustical . difficul- 
 ties. The fact that they have 
 been used so extensively mis- 
 leads most persons into believing 
 that beneficial results can be 
 thereby obtained. There is no 
 theoretical basis for their use and 
 to my knowledge there is no place 
 where any improvement in the 
 acoustics of any auditorium has 
 been brought about by stretching 
 wires in any shape, form or 
 manner. 
 
 It is claimed, e.g., that wires 
 radiated from above a speaker's 
 head to parts of the room will 
 assist in carrying the voice to 
 locations where it would other- 
 wise be difficult to hear. A 
 knowledge of the fundamentals 
 of sound enables one to see that 
 this is an impossibility. A wire 
 can only be stretched so as to be 
 in tune with one note. _ It is, 
 moreover, only possible to impart 
 a microscopic vibration to the 
 wire which is not audible. If the 
 wire were connected in the proper 
 manner to a relatively large 
 sounding board, it would be 
 audible. Even if it were pos- 
 sible to obtain the desired vibra- 
 tion, it, would be a decided dis- 
 advantage because an undue 
 
 110] 
 
prolongation of each sound would 
 be produced. It is desirable in 
 both speech and music to have 
 each sound absorbed quite rapid- 
 ly so as not to interfere with the 
 proper, hearing of succeeding 
 sounds. 
 
 Sometimes it is claimed that 
 an echo can be broken up by 
 sitretching wires in front of the 
 dffending surface. Experience 
 pjroves this cannot be accom- 
 plished. The sound wave, due 
 t!o its 'large wave front, passes 
 tftirough the wires with practi- 
 cally no change. The action can 
 b/e studied by watching an ocean 
 wave pass through a row of piles 
 stpaced at various intervals. There 
 is some disturbance noticeable at 
 tlie piles but a few feet beyond 
 the wave presents a united un- 
 broken front. 
 
 j The reason for the misunder- 
 standing regarding the use of 
 vii'ires is probably due to the fact 
 tjiat they have been used for 
 ytears as a means for support and 
 hunting for decorative purposes. 
 An improvement in the acoustics 
 has frequently been noticed and 
 the credit for the change has been 
 given to the wires instead of to 
 the flags and bunting. It is 
 simply the presence of additional 
 st)und absorptive materials, placed 
 ift usually a particularly desirable 
 Icjjcation, that causes all sounds to 
 be absorbed more rapidly, thus 
 reducing the period of reverbera- 
 tion or making a surface almost 
 totally absorbent, which other- 
 wise Would produce an echo. 
 
 I know a court-room where 
 wires were stretched the length 
 of the room and connected with 
 cross wires that were attached to 
 ventilator grills. The clerk of 
 the court told me they thought 
 the undesirable reverberation 
 would be carried by the wires 
 to the ventilators and then out 
 of the room by the current of air. 
 "It seems plausible," he added, 
 "but unfortunately it doesn't 
 work." 
 
 Other Fallacies 
 
 There are many features of the 
 design of a room that are erro- 
 neously blamed for the faulty 
 acoustical conditions. A sus- 
 pended ceiling, e.g., is thought by 
 some to be set in vibration by 
 the sound waves, thus producing 
 the undesirable reverberation. 
 While making an examination of 
 a council chamber, I was told by 
 the Board in charge that they 
 thought the trouble was produced 
 by the barrel vault ceiling, which 
 due to the suspended construc- 
 tion, was set into vibration by 
 the sound waves. They showed 
 me bags filled with sand which 
 they had placed on top of the 
 ceiling to stop the supposed 
 vibrations. No benefit was ob- 
 tained. The force of the sound 
 waves produced in this room was 
 actually too feeble to produce 
 even a microscopic vibration. 
 The trouble was due to the use 
 of hard, fire-proof building ma- 
 terials, without the introduction 
 of sufficient sound absorbing ma- 
 terials to neutralize their effect. 
 
 11