Recording Binaural Sound on Discs

Use of lP discs having two recorded bands of­ fers a very compatible and practical binaural record system. Great new marlcet potentIal IndIcated for record player manufacturers.

arm containing two cartridges side by side as shown in Fig. 1.

C••patIJlUlt" The confusion of multiple stan­ dards has already shaken up the record industry once or twice and another 4pheaval would probably be impossible. Therefore, both the bin­ aural record and its playback means must be interchangeable with exist­ ing standards. Due to the halving of available elapsed time with two channels, the 12-in LP is a natural starting point. With a normal 12 minutes of binaural playing time and a possible maximum of 14-15 minutes, most musical require­ ments can be met. The method de­ scribed here produces a record which can be played (one "ear" at a time) on existing equipment, and the binaural reproducing system will likewise play regular records by simply blocking one of the Author viewing binaural cut In procell on spedal recording lathe cartridges up off the record.

By EMORY COOK Cook Laboratoriea In F ig. 5 is ~hown the schematic Stamford, Conn. arrangement for playing back the record of Fig. 4. Within the limits of travel in arc of the arm across F the advantage!:. and potential however is costly to record and to the narrow band of grooves a high I wide sales appeal of binaural playback and further, a great deal of degree of accuracy of radial rela­ sound are to be exercised, the realm special equipment would be neces­ tion of playback points is obtained. of disc-recording, where cost and sitated. In the practical case the positive compatibility factors with existing 3. A single sideband carrier sys­ error is made to equal to the nega­ standards are favorable, offers great tem has been proposed, but the limi­ tive error in terms of wavelengths, possibilities. Let us then review the tations of this method are such that - at 1000 cps for instance. The several .. ways in which a synchro­ the frequency range of each "ear" maximum error then measures to nized double track in disc can be is restricted to a low value that could be of the order of magnitude of produced: not be considered acceptable from a O.Ol-in. along the groove longitudi­ 1. A binaural recording can be fidelity standpoint. nally at the middle of the record made on opposite sides of a disc, 4. To date then, it appears that by corresponding to less than a wave­ playing back both sides ~imul tane­ fa r the most compatible and practi­ length at 1000 cps along the groove. ously. This method is not very prac­ cal system is that of placing two The purist might say here that we tical however, because stampers recorded bands on the same side of cannot tolerate a phase error of cannot possibly be centered and an LP disc as diagrammed in Fig. 5. 180· in the range of maximum aligned in the press in the necessary As with ordinary recording equip­ directional sensitivity of the human rotational accuracy, and a wholesale ment, the cutting styli move on a ear, but this is not true, and it is redesign of playback turntables radius,-from outside in. Therefore not a matter of "toleratioR" anyway would be necessary. existing recording equipment is unless we confine our remarks to 2. Interleaved grooves are another easily modified to perform the the use of earphones as a listening theoretically possible means of bin­ mastering function. The playback medium. As a matter of fact, ear­ aural disc recording. This method arms may be integrated into a single phones can be given no considera-

Reprinted from TELE-TECH • November 1952

Copyright by CALDWELL-CLEMENTS, Inc., 480 Lexington Avenue, New York 17, N.Y. getting into the correct grooves; it merely means instead that there is to be a "free" lateral motion of one cartridge with respect to the other. However, the business of "loose pivots," i.e., rattles, in vertically pivoted arms is well known to pro­ duce non-linear effects at some mid­ frequency. The answer to the prob­ blem is the packing of the pivot points with viscous damping, so that the compliance of the cartridge needle is two or three times higher than the viscous compliance for fre­ Fit. 1. a_·up of dual cartridge playback arm al _nufactured lIy Lly· quencies at and below the undamped l",doII Electronic Corp., lIYlngston, N. J. 111 and 121 are adjust....nt sueWl lateral resonance. In bringing binaural into focus as a practical medium, the cost factor is tion ,at all commercially. If we another and more insidious danger particularly important. A first re­ think in terms of spaced loud­ to watch against. In establishing a action might be that the cost would speakers in a room then a full 111l6-in. dimension between points be almost doubled but this is not the wavelength error at 1000 cycles of pickup we must allow a tolerance. case. Binaurally, power output per can be thought of as corresponding Not only is it quite unthinkable to channel can be less than half that of to a motion forward or backward be able to align recording heads and an equivalent monaural channel for of one of the loudspeakers of about styli on a lathe to .001-in. tolerance the same apparent loudness. one foot, or a corresponding ran­ for each and every master, but the dom motion of the listener in the metal parts, especially stampers used roo~ (such as turning the head) to mold pressings are basically v.... r..,.. rr'oll•• while the music is playing which 0.032-in. copper as regards mechani­ With the use of twin triodes the is easily possible, in fact probable, cal strength. In the modern fast basic amplifier design may be without getting up from the chair. cycle press and in handling they doubled up (as in push-pull) with­ A typical modified arm would become stretched, so that P lh6-in. out much additional cost. Since we display a dimension of 10lh-in. may become distorted around the certainly need no more than half the from pivot to farthest stylus, an circle. Furthermore, non-uniform output power "per ear" the power offset angle of 27.5 0 and the stand­ cooling of pressings just removed supply is the same. The twin cathode, ard spacing of Plha-in. The frac­ will produce a small eccentric screens and plate supplies may be tional spacing is chosen rather than shrinkage on one side of the record by-passed through common capaci­ decimals to facilitate the rough ad­ and not the other. All these errors tors, since a moderate amount of justment by using a standard inch are of course capable of being cumu­ crosstalk between channels is per­ rule, where the points will fall into lative to the effect of at least 0.015­ missible. The only serious added the indentations at the specified in., and therefore we must have a cost will be that of the second out­ dimension. mutual "lost motion" in the P lha-in. put transformer, and means for re­ The radial error is not much af­ spacing figure of ±.015-in. or there­ ducing even that appear to be forth­ fected by the length of the arm, and abouts. However, since the record is coming. short 12 in. turntable arms are can­ started with pickups dropped in the In the "minimum" design of Fig. 2, didates for conversion. spiralling lead-in groove, this does there are only 16 one or half-watt In order to permit equal tracking not mean that we shall have trouble resistors, 4 audio condensers and 4 pressure of each point of contact, at least one and preferably both of the Fit. 2. Minimum binaural amplifier dellgn Inyolves only 16 one· or half· cartridges must be individually _It Nllston, 4 autlle capedton, and 4 tUbel not Including powe, lupply pivoted for the vertical plane, al­ though of course, if one is tempted to put up with the inconvenience, two separate arms may be operated.

....110_ Protlaetlo. Brror. No record is produced which is actually on center. The magnitude of the combined error in good commer­ cial practice can hardly be reduced under .010-.015-in. The staggered method of interleaved grooves can not operate satisfactorily because of the centering problem alone, since there is a large angle subtended between the two points of pickup. However, the radial method is the least susceptible, in fact practically immune to centering errors such as normally encountered. But there is

TEU.nCH • Novembe" 1952 12AY 7 12AY7 12AY7 12AU7 (4) 588 1 (12AX7 ) ( 12AX7 ) ( 12AX7) 1200

INSIDE INPUT I 1

+ 350 v 3SOV .. ADJUST TO 1.2 V IS K AT 2S0 rna I.ADJU ST CAT HODE 20 + VOLTAGf TO 80 V.

Fig. 3: Comple'e scllematlc of high qualify binaural amplifying IYltem. Power lupply II separale unit

tubes, not including the power compare it with a regular amplifier of O.Ol-in. this longitudinal adjust­ supply. No more decoupling or fil­ using four output tubes as a cost ment can be supplied through use of tering is necessary than with a regu­ reference. a test record signal. The Cook Series lar amplifier. The A-B comparison For equaliziI:J.g the effective high 30 test record is intended as a tem­ between pinaural and monaural on frequency response of the channels porary standard, and in order to be such a minimum system with cor­ in the room a condellSer may be palatable to non-technical users responding small speakers is per­ introduced across one or the other of employs a slowly ticking clock as fectly astounding, probably because the 220 ohm feedback re-introduc­ the source. The clock is fed into both the cheaper monaural systems are tion points in the 3rd stage of Fig. 3. channels in parallel simultaneously, so unsatisfactory in themselves. Yet in ordt!r to provide :.1 synthetic bi­ it is hard to see where such an am­ re.t ., AJI._.e_t naural signal. When played on a plifier could cost more than a few binaural arm into a binaural repro­ dollars extra. In addition to the necessity fo r an ducing system which has been ad;;. The opposite extreme would be adjustment to permit re-setting the justed for equal gain per "ear," the something along the line of Fig. 3, Ph6-in. dimension should cartridges cartridges in the arm may be ad­ where provision for magnetic car­ be changed, there must also be a justed until the clock sound in the tridges is made, together with phase­ "fore and aft'" adjustment provided room appears to be neither to the splitting so that each output is bal­ so that there is no time delay error right nor to the left, but dead -center. anced. Again the increased cost between channels-,-so that they In order to prevent insofar as pos­ over that of a straight "push-pull" operate on the same radius. A means sible inadvertent errors, the record design is very nominal, if one is to for locating to the required accur acy is made not in the 12-in. size, but

Fig. 4: (Left) Photo of binaural tone arm playing back binaural dllc. Fig. 5: (Rlgltt) Diagram Ihowlng how binaural arm II plyoted

R(COMEO ON ft _ J ,U$

nLE-nCH • Noyember 1952 Binaural Sound Such noise here, however, will take differently. (Continued from page 50) on a random left-right character In the case of standardization of and as such is distinctly objection­ recording characteristic curves a as a 10- in. disc, where the diameter able because of the directional effect. basic truth was finally recognized, of start of the test grooves corres­ Hence the recorded level must not wherein it is obviously both ineffect­ ponds closely to a zero point in the be lowered in an atte!l1pt to permit ual and impossible to standardize the radial error cycle. full "curve" equalization of bright recording curve. What we try to No better method than cut-and-try original material. standardize is the playback curve has been discovered yet for cutter When experimental recordings only. Thus in developing techniques alignment on the recording lathe, and were made using 6-in. microphone for binaural recording insofar as such alignment then has to be done spacing and earphone playback, microphone placement is concerned, in conjunction with a playback turn­ phas·ing was of course necessary. But the basic truth must always be kept table whose adjustment has been contrary to what might at fi rst be in mind-that playback (on the made on the basis of the Series 30 expected, phasing of playback speak- average) will take place using 2 record as a standard. speakers separated perhaps 12-15 ft. The difference in frequency re­ against a wall in a room of corres­ sponse between inner and outer ponding size and fairly random diameters of pressings is a subject acoustics. which is well explored and about As binaural disc as a medium de­ which much has been written. Addi­ velops, there will no doubt be a tional pre-emphasis is often and may great deal of expert activity in the here be applied at inner diameters realm of "where to put the micro­ to compensate somewhat in advance phone" and "studio acoustics," and for translation loss. Unfortunately, there will be many interim pet the­ there is an engineering proclivity ories expounded. Without a conclu­ toward making hard and fast rules Fig. 6: Oose-up of binaural recording sive amount of experience at this for "shop practice," to cover all such shows Ihe two bands of grooYes point, we can only suggest a few points as diameter equalization and directions in which not to go. For in­ reference pre-emphasis. The trouble ers is unnecessary for material re­ stance, the business of about-facing is that rules do not take into account corded indoors (rather than out­ the band on stage and playing into the varying character of program doors.) With any spacing of micro­ microphones in opposite corners of material. The method which is satis­ phones in excess of ear spacing, the stage wings is extraordinarily factory for a piano - velocity micro­ phase becomes random in view of unnatural in effect. Wall reflections phone pickup is not applicable to a the acoustics of both the original abuse the reality, and treatment, of bright orchestral picked up with a room and the playback room. How­ the walls to inhibit reflections pulls wide- range condenser. With mate­ ever, in playing back material which the teeth out of the binaural head. rial which is originally bright there was recorded in the open, with no The bright synthetic modern studio is a severe limit on the amount of confining walls, phasing may well be­ acoustic appears definitely out from effective pre-equalization. Fortu­ come desirable, especially if the the binaural standpoint. In general, nately in a binaural system the two acoustical environment of the play­ any studio or hall which has been channels are not necessarily matched back speakers is on the "dead" side. treated with the idea in mind of cre­ for frequency response, and in Fig. 3 By now, most of us are well aware ating an "even" frequency distribu­ it will be found that in the de­ of the fact that a real distinction tion of energy per square foot,-the emphasis circuit (plate) of the first exists between standard monaural "mix 'em up" philosophy,- is lowest stage of the "inside" channel a 2.7K and binaural systems. Any compari­ on the binaural scale. resistor is used to leave pre-emphasis son is unfair at the start because the And the one unhappy malpractice in effect above about 3300 cps. two media are not comparable on which has been in vogue for 20 years A general development probably any real basis. The extension of the of broadcast and recording,-that of not too far away is the "half-mil" aural medium by addition of dimen­ the small and odd-shaped, acousti­ point which can only be associated sion, direction and perspective is im­ cally odd and unnatural control room with cartridges having extremely low portant not only for vitalizing the is absolutely fatal for producing bi­ motional impedance. When such musical catalogs we are building but naural. Note is taken of various cartridges become available in man­ also for extending useful repertoire broadcast and recording company ex­ ufacture, the inside-outside range into fields such as plays and other ecutives who, for very good reasons, and distortion of LP's in general will documents, where the third dimen­ insist on listening to their records in be vastly improved, and the binaural sion communicates enough additional an audition room about the size and translation loss factor will be neg­ information to make the difference shape of an "average" living room ligible. As for present practice, it has between failure and success. with similar acoustics. Yet the rec­ been found generally acceptable to Of commercial necessity we have ords are produced and balanced in a maintain approximately a 50 I.Is (3300 had to rule out earphones as the studio control room which is about cps turnover) differential between playback mechanism, albeit their as far away as one could get from the inner and outer binaural bands in the rigorous binaural nature. We thus living room prototype, both in size, original recording. Depending upon descend into a morass of conjecture shape, proportion, acoustics, and the program material the outside relating to position of loud speakers, relative position in the room of loud­ band might be 50 or 100 I.IS pre­ acoustical reflection characteristics speaker and listener. emphasis, and the inner band 100 1.15 of side and back walls, shape of For binaural records, it would be or 100 I.Is + 50 I.IS, correspondingly. room, etc., all applied to the room better in these control rooms that Assuming that binaural record­ in which playback occurs. Here none earphones be used, for making bin­ ings are pressed in high grade but the broadest rules may be drawn, aural productions will probably be­ plastic, the most serious surface since even mass-produced living come much more of an art than the noise will be that of ticks and pops. rooms are decorated and carpeted regular monaural ever was.

nu-nCH • Noy_..... 1952