US2493638A

US2493638A - Synthetic reverberation system

Info

Publication number: US2493638A
Application number: US678707A
Authority: US
Inventors: Harry F Olson
Original assignee: RCA Corp
Current assignee: RCA Corp
Priority date: 1946-06-22
Filing date: 1946-06-22
Publication date: 1950-01-03
Anticipated expiration: 1967-01-03

Description

3 Sheets-Sheet l www Y reien/cfr;

H. F. OLSON SYNTHETIC REVERBERATION SYSTEM @fiers/ML .ww/N6 zam/muses Jan. 3, 1950 Filed June 22, 1946 finca zada lNvENToR Yzrr/yF/fm B C@ AT oRNEY Jan. 3, 1950 H. F. oLsoN 29493538 SYNTHETIC REVERBERATION SYSTEM Filed June 22. 1946 3 Sheets-Sheet 2 INVENTOR.

Jan. 3, 195o H. F. @LSON 2,493,638

SYNTHETIC REVERBERATION SYSTEM Filed June `22,2. 1946 3 Sheets-Sheet 5 Z5 #fir Fir [A frrg E 02mm Y l y Patented Jan. 3, 195@ UNITED '-srArEs Aram orties yThis invention relates V130"Synthetic reverbera- `Y:chambersarerooms. about the size ofthe average `tion systems. 'Sindrome-particularlyt0 an aCOUS- x. liv-ing room. with. a..r.everberation"itime oabout "ticelectronicreverberation simulator forintro- 31..sec0ndsv.Thlgmmd is fedl intofthechamber M'ducing--synthetic-reverberation- -in yreproduced v4by smeans-o1, aloudspeaker and is pickedv upiby a ,ffsound whereby-rendition of the sound may be 5;-n1icrophor1e. ..The loudspeaker-room'fmicrophone vj greatlyfenhanced. VAcombination constitutes a reverberation simu- ^When asourceof-"sound-is ystarted-in af room, lator. rAmonglthe .disadvantages of such a sysltheenelgyidOeS11013'buildup instantly due te JChe A1 .temis that the.reverberationtimeiis fixed; .Usuf--iinite' velocity of propagation of a sound Wave. .,ally, also,. .the space required for such a studio is Each;pencilof-`-sound sentrout by the source is 1Q aiiotherdisadvantage,particularly in areas where reflected many times fromithefabsorbing Walls of therents .arehign There are .also a number" of the -room-beforeeitf is ultimately dissipated to a otherdisadvantages.which. characterize reverlevelwfbelow thewambentfor' threshold level .of ...ulcerationsimulatorsfofjthis.type As substitutes fthe room. A-steadystatercondition obtains when --for such reverberation systemsyorgchambers; a

the= energy @absorbedl byrlthey -walls equals the lnumber otiothensystems havej been' proposed. -"-ene1jgy-'f delivered by-the-- soundV source. -In the The.v lattenincludeldisc recordings,'electrical` de- --sameway;when theisource is stopped, some time lay circuits, magnetic tape recordings with delay, -'-fisfrequired before thefenergyisabsorbed to alevel mechanical or spring wire delays, phosphor below-'the'ambient-or' threshold level. Thev -recoated,.discs, 4and lso Ton.; These systems all v--verbera-tion time VVhas Lbeen arbitrarily denedvas ltlpossessoneiorlmore'disadvantageswhich are Ytoo `-fthe time-required fori-'the sound to decrease to wellknoyvn torequirefurther comment. '-one-niillionth of the-original intensity aiterthe .Ehe primarygobiectjomy.present invention is source of the sound Ahasbeen stopped. to provide an improved'synthetic reverberation Ithas been Aestablishedthat there existsv Ian r,systemv which. isfreevfrom .the above mentioned optimum reverberation time for a particular-'room 25.,and Lother `objections which` arer found in synf and for a partlcular-typefof sound rendition.V This thetic reverberation Vsystems proposed heretofore. means"that; -for argiven room,` such as a'broad- More particularly, it isan object of my present casting'or a recording studio,v for example;rd-ifer invention to., provide an improvedk system for vent-reverberation'times'arerequired depending producing,.syntheticreverberation by means of upon the type of sound rendition. One of the 30 which theamountandjtypeof reverberation.v can major-problems@ sound broadcastingandsound f-be coiitrolled'over asubstantial range to provide 1recording-- hasbeen `to control the amount and optimum reverberation.characteristics either for type ofl reverberation'finthe -pi'ckuprpartof ythe a, AloeilriCillelI. IOOm-.Ol Studi@ 0I' fOr 2 particular System .or mitnevrecordingpart-offthe system,4 as rendition which'is to be recordedV or broadcast and thecasemayibegiinrorderrto providesound--Whicn 35' later, repIOdllCed.

-fuwhenireproducedfl-will -fhave' t-hevrproper amount Another'obieoto'imy present invention is to .and `quality r off @reverberation The A--reverbemg ,prov-ideen improvedsynthetie reverberation Syszxtion;time1f0f-rstudi0s has', been @ontrouw-mm tem as aforesaidwhichwill have a substantially o viouslywithinicertainvlimits, Vby various means ,uniform reverberation" frequency characteristic ,xdesignedto vary-thezdegreegof absorption, 'It @and in whichthe reverberation characteristic .fsihaslalsobeen proposed tofco'ntrol the reverberacan loe 'readily COUIOlld 't0 entllate either :stionhtimeef: the conectad Sound byzthe use 0f Mthe high or' the. 10W vfrequency reverberation. .r directional?. microphones. 3 IfIovvever,v -conditions Still 'aIlOiJherObl'eCifOfcmy present invention iS :oftenlexistwhich necessitate a change in reverto'provide4 animplOVed 'Synthetic reverberation eherationama 4traction of a second. Obviously, to 45 system'which requiresno recordings and there- '1 change thelabsorptionY-'in a studio by meansof OIB S de VOdOf QXDGHSVG leCOI'dng and YepI'O- Slidingppanelsfopexamme; asx-,has been A@one ducing"n'iechanisms`v generally associated theres, heretofore; is a: 'rather cumbersome procedure. With. :'f Furthermore, very drastic changesl in the deg-ree A furtherobject of `my present invention is to of :reverberation Aare iften requiredl as. -for ex- 50 provide an improved synthetic reverberation sample', from-one=halfsecondto three seconds. systemofrtheiitypeeset'forthwhich can behoused At the present time, most broadcast Aandf-rein a compact cabinet :structure and which vcan be ffcordingf-'studios-employereverberation chambers placed in a room or studio at any suitable point V`V`Lto--varylthe-amountfof*-reverberatonin a'reproto provide the reverberationcharacteristics Ade- Jidu-'ced sound-program. ffThese---reverberation 55"sirednforrthe"particular"room and suitable -for the particular program to be either broadcast from or recorded in this room.

It is also an object of my present invention to provide a synthetic reverberation system as above set forth which is very compact in construction, which is readily portable and therefore can be easily transported from one room to another, which is highly efficient in use, and which is quite inexpensive in cost.

In accordance with my present invention, the synthetic reverberation system includes a plurality of delay channels for producing a plurality of echoes of the non-reverberant sound with different phases and mixing the same with the non-reverberant sound. One system constructed in accordance with my present invention included four loudspeaker-pipe-microphone delay systems which provided delays of approximately 23, 50, 68 and 100 milliseconds, the pipes providing the desired delays. The channel with the longest delay is used as part of a controlled, negative feedback network in which the amount of feedback controls the reverberation. The sound in this loop is fed into all four systems again and again. In this way, each second contains 40` components. The fact that my improved system supplies such a large number of components makes the introduced reverberation quite realistic. lIhe reverberation time may be varied from Zero to three seconds in a fraction of a second b-y merely varying the amount of negative feedback. A mixer makes it possible to adjust the ratio of direct to reverberant sound. The components of my improved reverberation simulator can be housed in a compact cabinet and can be placed in any desired location within a particular room.

The novel features of my invention, as well as additional objects and advantages thereof, will better be understood from the following description of two embodiments thereof, when read in connection with the accompanying drawings, in which Figure 1 is a diagrammatic view illustrating the basic elements of a system for introducing synthetic reverberation in reproduced sound;

Figure 2 is a diagrammatic view of one form of synthetic reverberation producing system in accordance with my present invention;

Figure 3 is an enlarged, partly sectional View of one of the loudspeaker-microphone units with the associated pipe of Fig. 1;

Figure 4 shows, graphically, for a fraction of a second, the decay pattern of the various sound output components of the system of Fig. 2;

Figure 5 is a set of curves showing the average attenuation characteristics of the several pipes shown in the system of Fig. 1;

Figure 6 is a set of curves showing the reverberation time vs. frequency characteristics of the reverberation simulator of Fig. 2 for various settings of the reverberation control; and

Figure 7 shows, diagrammatically, a second form of my improved reverberation simulator.

Before referring more particularly to the drawings, it may be pointed out that, when a source of sound operates in a room, the sound pressure at the microphone or pickup point may be expressed as follows:

where A1=amplitude of the direct sound at the microphone,

r1=distance between the source and the micro microphone.

In the above equation, the rst term of the right hand member represents the direct sound and all the other terms represent the reflected sound. The amplitude A of the reflected sound decreases with increasing distance r between the sound source and any particular microphone. From the above equation, it will be apparent that the conditions represented thereby may be simulated by a system of the type shown in Figure l in which the input provides direct signal or non-reverberant sound representing electrical waves along a channel I, and delayed signal or reverberant sound representing electrical signals along a plurality of

parallel channels

2, 3,

4, 5, G, 'I and 8. Each of the channels 2 to 3, inclusive, includes a sending transducer 9 coupled to a receiving transducer Il) through a suitable delay system Il, the delay systems Il each having a different degree of delay. The outputs of the transducers ID may be supplied through a plurality of ganged attenuators l2 in each of the channels 2 to 8, inclusive, to a mixer I3 which mixes the direct and reverberant sound representing signals and supplies them to a suitable output member. It will also be apparent from the system of Fig. 1 that, to obtain the required conditions in a room, would necessitate a tremendous number of transducers and would, therefore, be impractical. On the other hand, any reduction in the number of units must necessarily involve a compromise. It is desirable to include more components rather than less. particularly for the longer reverberation times.

In the simplied reverberation simulators of my invention as represented in Figures 2 and '7, there are forty components of sound per second. Such systems adequately ll most practical requirements.

Referring, now, to Figure 2, there is shown a microphone or other suitable input member 2l which is adapted to pick up a train of direct or non-reverberant sound waves and to translate them into a corresponding train of non-reverberant sound representing electrical waves. These electrical waves may be fed first through a master gain control 23 and then along two parallel paths 25 and 2l, the direct sound representing signals in the path 25 being fed to a variable or adjustable mixer 29. Non-reverberant sound representing signals supplied by the microphone or the like 2l are fed along the path 21 through one or more attenuators 3l and thence to an amplifier 33. The output of the ampliiier 33 may be fed to one or more amplifiers 35 and thence along parallel branches 3l, 39 and 4| through

power amplifiers

43, 45 and 41 to four electro-acoustical transmitters or

loudspeaker units

5l, 53, 55 and 51 each forming part of a channel for producing reverberant sound representing electrical waves or signals.

In the system of Figure 2, four such channels C1, C2, C, C4 are shown arranged in parallel. In the channel C1, the loudspeaker 5I is connected by an acoustic line 6I to an acoustc-electrical ""1 ediles-,63e

verberant Ysound'representing electrical Waves: or i signals rare supplied'f -byltheo-mixing- Anetwork 'lll through an amplier'liflvto'the mixer 2l-9 Wherethe direct and the reverberantfwsound representing `r'electrical' xwaves `are mixed in the desired' pro portion. The --signalsi suppliedhby thee mixer 2 9 I Lfthe:several:'microphones "1 l 55:13; 515 `and 11 may "il#speakerv or a microphone? This unit is a dynamic' Wunit 1 having a magnetic yoke 9 I" which supports, v`centrally,A a'fmagneticfpolepieceliSS@and a magi-netic yokeV plate 95 which i has a-suitable open- ?=ing'atffthecenten 13m-provide ianair-:gap-f191- in =--we11 known manner. ff Within the air-'gap'l is :vafyoicescoililse car-riedibyy a diaphragm F-'flexiblyzsupportedzfor `'vibration by a centering` spider i S1203, also. 2 ini-:well llinownumanner. A The .unit inzi'cludes anthroat 11:05: .whichiis:` connected tolits `@associatedpipe?for .examplai'the pipeiiyby a .conical :connectorfl` 6l a fwh-ichim'ayi or may 'not be .n an :integrat '.part i' ofir v.they` pipe; sf r may .be found. mogtfsuitabla ii Ther-pipes f5 H1163, 65 and'l may o .betoneninchI i diameter# aluminum :pipe having a Anatali thickness of 1.1'approximatelyv one-:thirty- :sasecondi inchixAszmentionedaheretoforc; the pipe ff'rlzhas a length ofithe^order of -25:reet,the pipe 65 has a length? off the; order: of .f55 i1eet,:the.` pipe. 63 whasg. ai lengthi of;.fthe orden. of 75 feet, and the longest Loipe Bildhasaplength-Lof :the order of 110 feet." @Pipes-of :thesepl'eI-igths will introduce in the channels-(14, C3.;C2,@C1rtinie.delays of the order er1-23, 50, 68,y and'- :10U-milliseconds', respectively. The conicalorftaperedpipe portions, such as Athe portion ila; may :be approximately i8 inches in" length. The attenuation characteristics of the .iour pipesvElWSSGgIGl, as deSCribedabOVe, are shownf in lFigure 5.1, Irdesired, suitable .equalizers may beeniployedbetween the input and the oututottheamplier system`35f43; 45, 41 to com`l f pensate forthe greater attenuation at thehigher frequenciesin' theiseveralpipes Si', 63, 65 and'f'l.

It ivillbe apparent`,"fromV the foregoingdescription, that .the electrical waves or. signals The acoustic -lines- 6 I ,i

11-=representffreverberant `Asound of different 1@phases by ireasonfofvlthe different-delays. provided L15? l`arefcohnedto thefzseveralipipes high sound f pressures- `may be :obtainedi .with- V small inputs. 'i This meansfthattheratiofof signal to noise is `very largeimore'- than=160 db .if Furthermore, -the non-'linearidistortion-1isflow because .the loud- 1 speaker-units, -K whichr are usually the principal f source fot i: distortion; ioperate i at avery low level :and'f therefore, do 'not produce any appreciable snon-linear?distortiorrl The `reason for using four i2delayfchannelsfasshowniriiFigI 2, is to obtain L soundcomponentsfwhich are staggered withrrespectfto @time 1l impulse for. a train Fofvwavesi: enters the system.

'luppose;V for-example; that an 'if'lhe-outputv consistsl'of, first; the" direct` sound impulse; a reverberant soundfrepresenting"impulse Vor" aphase correspondingvto A feet-later due to -22 :the .125: lfoot Jpipef'l #another reverberant sound A.1representing impulsevof:- a phase which Scorrespends to.55..ieetlater.due.towtliepipeig another i l reverberant soundf L11-@presenting impulse .f f of; a

I phasewwhichi corresponds 'to-L75' feet later :by rea- -.son..off; thespipe: ,S331 and'- final-ly ai reverbera'nt The-f channel C1, a 'whichi includes: :the longest system by connecting.A arporton of; theoutput'of the mixingnetworki'l 9: 'z to' the a ampliers f l35 i; through a variablereverberatiorrztime::controlling 'or feedback4 controllingsdevice f fnl r; (preferably a step-'byf-step.control) and an amplifierfl; The negative.:feedbacknsystem' represented, by the -a asholine 1 l memo-Figure 2 feeds Asoundinto the so .that sound impulses:followfjofnphases :correspendi-ng. Eto 135 feetr afterrthefroriginalg direct y-xinnfiulsex-which enters theysystem, 165` feet later, --185 Y'feet zlaterg 220eteet later,- -245 feet later, z etc. A parte of the components fforyone second: reverberation. timeis shownfin Figurenfl. Itfwillibe system Without feedback,- whereas the improved systemy describedabove:- r `makes "this entirely feasibleandproducesa-train-ot sound which is .,suppliedby the output member l89 with precisely the desired, ,ty-pe: and amount of reverberation. .The .st-ep gain` control Hl'lzl which controls'the amount-of .feedback-may becadjusted to a first positionito causegthe-system ,tooscillate. The

succeeding positions of :the gain control. I Q1 may be suchasitoset .thereverberation time for three seconds, two seconds, one second, and then onehalf second. A.system such as described above can be housedveryfnicely .in arelatively small cabinet 'Withtheoperating knob of the gain control lM/,mounted atsome .convenient placezon ,one Wall of the cabinet. ,In one construction, ythe *entire systemincluding the. delay pipes, loudspeakers, microphones,. ampliers, mixers, regulated power supply, etc., was housed in a cabinet 36 inches Wide, 24 inches deep and 7 8 inches high.

Themeasured reverberationv timesof this particular 'construction'.for various settings of the gain control' l Q1, is 'shown inlFigure 6., It willbe noted; from 'thislgureY thatthe reverberation time for onefhalhone. and twoseconds is quite uniform with respect t frequency. The lower reverberation time at the higher frequencies is about the same as that obtained in a room. The reverberation time with respect to frequency is not quite as uniform for the three second setting of the gain control H11. This is due to the fact that, for this setting, small variations in amplifiers have a great effect. However, neither is the reverberation time of a room using a single loudspeaker and microphone uniform with respect to frequency. Hence, the variation in reverberation time with frequency at the three second setting is not serious. If desired, the reverberation time vs. frequency characteristic can be changed to give greater reverberation at either the lower or the higher frequencies. For the characteristics of Figure 6, the amplifiers were adjusted to give the sam'e reverberation time for the frequency range from about 100 to about 2000 cycles per second.

In Figure 7, there is shown a modified form of my improved synthetic reverberation producing system. It will be noted that, in this system, a single loudspeaker unit I is employed. This reduces the number of power amplifiers required.

, In place of the

pipes

53, 65 and 61,

small bleeder pipes

63a, 65a and 61a, which are tapped off from the pipe 6| at suitable points spaced along the length thereof, are used to feed the microphone units 13, and 11, respectively. The pipe 6I may again be a 110 foot, one-inch diameter, aluminum pipe, and the

pipes

63a, 65a and Bla, which may be one-quarter inch diameter, aluminum pipes, are tapped off from the 110 foot pipe 6I at such points as to provide a 75 foot path through the pipe 63a, a 55 foot path through the pipe 65a, and a 25-foot path through the pipe 61a. By selecting the proper tapping points on the pipe 6 l, the response frequency characteristic of all the microphone units 1|, 13, 15 and 11 will be substantially the same. A fixed attenuator III may be provided between the two one-stage ampliers represented by the blocks 35 for changing the shapes of the reverberation time vs. frequency characteristic curves for the purpose either of obtaining a uniform reverberation frequency characteristic, or of accentuating either lthe high or low frequency reverberation. The direct and reverberant sound mixer 29 of Figure 2 is omitted in the modification of Figure 7 since it is of little or no value in standard broadcast practice. The simplified synthetic reverberation producing system of Figure '7 can be used as an adjunct in the same way that reverberation chambers are now used. With the simplifications shown in Figure 7, the number of vacuum tubes required is greatly reduced, as is also the amount of one-inch pipe required for the delays. With these changes, the unit can be housed in a portable cabinet of about 40 inches by 34 inches by 18 inches.

From the foregoing description, it will be apparent that I have provided an improved synthetic reverberation simulator which is simple in construction and which can be used in any surroundings for the indicated purpose. Although I have shown two modifications of my invention, it will undoubtedly be apparent to those skilled in the art that many variations thereof, as well as changes in the particular forms herein described, may be made within the spirit of my invention. I therefore desire to have it understood that the forms of my invention described above are merely for the purpose of illustration and are not to be taken as limiting.

I claim as my invention:

1. In a synthetic sound reverberation producing system, means for translating non-reverberant sound waves into non-reverberant sound representing electrical signal Waves, a plurality of sound wave conveying channels for producing reverberant sound representing electrical signal Waves, each of said channels including an electroacoustical transmitter, an acousto-electrical receiver, and an acoustic delay line connecting said transmitter to said receiver for conducting to said receiver the sound waves emitted by said transmitter, signal wave conveying means coupling said first named means to each of said transmitters to apply said first named signal wave thereto, a signal wave mixing network connected to said receivers for combining the reverberant sound representing signal wave output thereof, electrical coupling means including said mixing network and a portion of said signal Wave conveying means providing a feedback path between said receivers and transmitters for a portion of the combined reverberant sound representing electrical signal waves from said receivers, and gain controlling means for said electrical coupling means for controlling the degree of feedback of said last named signal waves and the reverberation time of the system.

2. In a synthetic sound reverberation producing system, means for translating non-reverberant sound waves into non-reverberant sound representing electrical signal waves, a plurality of sound wave conveying channels for producing reverberant sound representing electrical signal waves, each of said channels including an electro-acoustical transmitter, an acousto-electrical receiver, and an acoustic delay line connecting said transmitter to said receiver for conducting to said receiver the sound waves emitted by said transmitter, the acoustic lines of said channels being each of different length whereby to provide an acoustic delay of different magnitude in each of said channels, signal wave amplifier means coupling said first named means to each of said transmitters to apply said first named signal waves thereto, a signal Wave mixing network connected to said receivers for combining the reverberant sound representing signal Wave output thereof, electrical coupling means including said mixing network and a portion of said signal wave conveying means providing a feedback path between said receivers and transmitters for a portion of the combined reverberant sound representing electrical signal waves from said receivers, and `gain controlling means for said electrical coupling means for controlling the degree of feedback of said last named signal waves and the reverberation time of the system.

3. A synthetic reverberation producing system according to claim 2 characterized in that said feedback path is a negative feedback path.

4. A synthetic reverberation producing system according to claim 2 wherein there are foury of said channels, the acoustic lines of said channels being, respectively, of the order of 25 feet, 55 feet, '75 feet and 110 feet in length.

5. A synthetic reverberation producing system according to claim 2 wherein said longest acoustic line has portions common to the others of said lines, and wherein the remainder of each of said other lines is constituted by bleeder lines connected to said longest line at spaced points therealong.

6. In a synthetic reverberation producing system, means for translating non-reverberant sound waves into non-reverberant sound Waves representing electrical waves, a plurality of sound Wave conveying channels for producing reverberant sound representing electrical signal waves, each of said channels including an electroacoustical transmitter, an acousto-electrical receiver, and an acoustic delay line connecting said transmitter to said receiver for conducting to said receiver sound Waves emitted by said transmitter, signal wave conveying means for applying to each of said transmitters non-rever berant sound representing electrical signal waves derived from said first named means, said channels serving to convert the Waves applied to their respective transmitters into reverberant sound representing electrical signal waves, a mixing network connected to said receivers for combining the reverberant sound representing electrical signal waves supplied by said receivers, means including at least one of said channels arranged to provide a negative feedback path between said receivers and transmitters through said mixing circuit and at least a part of said signal wave conveying means, and output circuit means for combining non-reverberant sound representing electrical signal Waves derived from said rst named means with reverberant sound representing electrical signal Waves derived from said mixing circuit.

7. A synthetic reverberation producing system according to claim 6 characterized in that the acoustic lines of said channels are all of different length whereby to provide an acoustical delay of different magnitude in each of said channels to produce echo simulating eiects of different phases.

8. A synthetic reverberation producing system according to claim 6 characterized in that the acoustic lines of said channels are all of different length whereby to provide an acoustical delay of different magnitude in each of said channels to thereby produce echo simulating eiects of diierent phases, and characterized further in that at least the channel having the longest one of said acoustic lines is included in said feedback path.

9. A synthetic reverberation producing system according to claim 6 characterized by the addition in said feedback path of means Ior controlling the amount of feedback in said path and the reverberation time of the system.

10. A synthetic reverberation producing system according to claim 6 characterized in that said combining means comprises a variable mixer.

ll. In a synthetic reverberation producing system the combination according to claim 6, characterized by the addition of means for selectively controlling the relative amounts of reverberant and non-reverberant sound representing electrical signal waves delivered to said output circuit means.

HARRY F. OLSON.

REFERENCES CITED The following references are of record in the file of this patent:

UNITED STATES PATENTS Number Name Date 2,104,811 Pster Jan. 11, 1938 2,292,745 Davis Aug. 1l, 1942 2,318,417 Phelps May 4, 1943 2,403,232 Parisier July 2, 1946l 2,421,424 Kreuzer June 3, 1947