Sound; certainly it sounds like their work. Side one of my copy has some surface noise, but side two is almost perfect. On the strength of this record, I've ordered Corboz's recording of the B Minor Mass, MHS 1708-10. Tchaikowsky's Serenade for Strings in C Major, Op. 48, and Dvorak's Serenade for Strings in E Major, Op. 22: Enrique Asensio conducting the English Chamber Orchestra on an Ensayo recordingMHS 1623. Another fine performance with very good recorded sound. (Ensayo, you may recall, produced the excellent AR demonstration record.) Surfaces on my copy are almost perfect. The only flaw I've noticed is that the strings sound a little hot at the top. Mike Riggs (Massachusetts) New Insights. There is a new source of wide dynamic range, smooth sounding, realistic discs for those who have tired of the program material on some of the super-demo records most of us own. Insight Records (7726 Morgan Avenue South, Minneapolis, Minnesota 55423) has promised three or four releases a year of classical and jazz material selling for $10 to $15. Their first effort is entitled "Fidelity First: An Unrehearsed Experiment." It is light, big band jazz recorded by a twelve-piece band. The record not only has impact and wide dynamic range, but is very smooth and lifelike. It has none of the harshness, funky super-ambience, or overly forward closeup sound found on some of the other available "super records." Doug Erickson, who seems to be the man behind Insight, says that during recording and mastering, every piece of equipment and every circuit that could be eliminated was done away with. There is no multi-miking, limiting, compression, or equalization. This leads to the interesting point that the records don't sound particularly good on ordinary phonographs, but are superb on wide-range systems. [Compression, limiting, and equalization can make poor phono's sound unrealistically good, but the resulting discs sound unnatural on good systems. Ed.] I could see no sign of limiting using an oscilloscope, even on two cuts which include solo voice. I measured peaks about 14 dB above standard 0 VU, but this gives an inaccurate reflection of the disc's dynamic impact, since it is cut several dB below normal industry levels. At the same time, there was little particularly bothersome surface noise either, though one side was less quiet than the other. There are flaws, but compared to the technical achievement, and to the realistic, livesounding playingthe band was jammingthey are minor. The record is called ". An Unrehearsed Experiment" because many aspects of its production were not refined before takes; the players weren't paid to rehearse before recording (that's not necessarily bad if you like life-like playing), mike placement and balance are better on some cuts than on others, and some takes are cut off without natural decay. This last comes as even more of a surprise since one is so easily caught up in the swing and naturalness of the music and sound. This record is meant to serve as an introduction to the sort of quality Insight is capable of providing and is selling for $6.50, about half the intended eventual price. It also is meant to measure the market, and as part of this, Insight is looking for feedback from listeners as to their opinion of the technical approach used and also as to the sort of music wanted. As for me, I'm looking forward to their future releases. Ira Leonard (Massachusetts)

More on Phono Cable Capacitance
Chapter XI in the BAS's continuing search for accurate data on the capacitance of your turntable's phono cables: Massachusetts member Andrew Kobziar sends word on Sony units. This is believed to apply to all Sony turntables; says Sony: "Each cable lead is approximately 210 picofarads up to the cartridge. There is little variation between models." Thank you, AK; those of you with similar information of known accuracy, please forward same to Box Seven.

Cartridge Capsules

Herewith, notes on Peter Pritchard's best. Sonus Blue Label I would like to alert you, if you don't know about it already, to a new contender for the stateof-the-art cartridge award, the Sonus Blue Label. This is Peter Pritchard's new baby, and one of the first products of his new company (Sonic Research Inc., 27 Sugar Hollow Road, Danbury, Conn. 06810). [If you write, mention the BAS. Ed.] My previous standard was an ADC-XLM Mk. II in a paddle-damped SME 3009 with nondetachable shell. The Sonus blows away the XLM. The improvements noted were: 1) more focus, especially on the human voice, than the already very good XLM, 2) more detail than I have ever heard from any cartridge, 3) deeper and tighter bass than the XLM, 4) better tracing and tracking than the XLM, and 5) very good overall balance. The detail and focus are beautiful. Strings have that resinous quality that I have never before been able to elicit from my system. The reproduction of the human voice gives the uncanny impression that the soloist is right there in front of you. The cartridge also gives a good feeling of depthit does a good job of placing some instruments near you and some farther back. Transients are great without giving the feeling that the music is all transientsan effect I have sometimes noted with the Shure V-15 Type III, The Sonus Blue Label now is my reference cartridge. The system used in this evaluation consists of an SME 3009 non-detachable, Sony TTS-2251, Dynaco PAT-5, Phase Linear 400, and Dahlquist DQ-10's. The Sonus was compared with the Shure V-15 Type III, ADC XLM Mk. II (two), Ortofon M15E Super, Grace F8C, and Grado F1+. Jeff Stake (Illinois) [ Editor's Note : The Blue Label is priced at $115 list, although it is occasionally found discounted to $80 to $85. Go shopping. The Blue Label is a Shibata stylus unit; elliptical (Red Label) and conical (Green Label) versions are available at $95 and $80 list, respectively. The Sonus Blue Label also is getting raves from the Atlanta hi-fi club, Audio Forum. To wit (culled from their publication Mixdown): "The Sonus cartridges, particularly the Blue Label, continue to represent the finest available. A second-generation (of these cartridges) now is available with an improved and strengthened cantilever; thus it appears that the production problems of the initial run have been solved. (Early Sonus tended to collapse in all but the lightest arms, and appear to work best only in damped unipivot designs. Ed.) Some success has been had using the Sonus in properly damped SME 3009 arms with lightened head shells, but the SME does not allow the Sonus to function properly." The note goes on to suggest use with, first, the Vestigal tonearm, and beyond this, the Keith Monks, Formula 4, and Decca. If you would like to subscribe to Mixdown , send $12 for a year's worth to Damon Hill, 3261 Circle Oaks Drive, NW, Atlanta, Georgia 30339. Jim Brinton] ADC-XLM Mk. II When stylus replacement time for my XLM last came around, I went out and instead bought an XLM Mk. II (serial number 243281). It turns out to be a genuine improvement, with better definition, resolution, and separation and without the slightly withdrawn quality of the old XLM. Best of all, it is substantially less compliant than its predecessor (less compliant even than the V-15 Type Ill according to High Fidelity's test report). I'd been getting very good results

using the older XLM in a Keith Monks arm. It tracked most warps without visible or audible trouble. Still, if I peeled away the grill cloth from one of my speakers, I could see some woofer bounce. The Mk. II has ended that. It does require greater stylus pressure than the original but seems to track just as well. If you like the XLM, I think you'll like the Mk. II better. Mike Riggs (Massachusetts) [ Editor's Note : This new XLM may be the current price-performance champ despite its list price of $100. It is currently being discounted to $38 at K&L, $35 at Dixie Hi-Fi, and even less at other mail-order emporia. Its combination of high output and its inductance makes it a very quiet cartridge audibly more so than the Shure V-15 Type III, for example. Jim Brinton]

X-Rays and Your Color TV

When Ron Dunlap, a physician and co-founder of Dunlap-Clarke Electronics, was a guest on "Shop Talk," the hi-fi talk show on WBUR, Boston, he suggested that it was a good idea to have your color TV checked for excessive x-radiation. I arranged for such a check with the Massachusetts Department of Public Health (727-6214 in the Boston area), and three months later an inspector came to my home and performed the 15-minute test. The testing device is called a Storm's Meter, which contains about twelve individual Geiger-Mueller tubes and costs $1200. My TV was found safe, with radiation below that allowed by the federal standard. The inspector stated that if any TV is ever found to be emitting excessive x-rays, the manufacturer is obliged to repair or replace it. The inspector also stated that only a trained technician should service color TV's to ensure that radiation be kept acceptably low. He suggested that the primary cause of excessive emission is improper adjustment of operating voltages to levels higher than those recommended by the manufacturer, and a re-test after each service call should thus be requested. Large-screen color sets (greater than 14 inches) are more likely to be radiation offenders than black and white or small-screen color TV's. (Sets manufactured after January 15, 1970 will carry a tag certifying that they are safe according to federal standards.) The testing service is free for the asking, and they will also check microwave ovens for dangerous radiation. It is reassuring to know that these devices are operating within safe limits. Alvin Foster

Noted at the AES Show

Master tape drives: Twenty-four tracks getting increasingly popular. Perhaps the most interesting drive was on the Stephens machine. (John Stephens is a California designer who started out modifying 3M drives with his own electronics. He now builds everything. His present machine has no capstan or pinch roller; tape motion and tension are wholly controlled by the feed and takeup reel motors. An idler in the tape path provides speed sensing. A couple of time delay units were present: the Delta-T from Logicon and the Time Line from Pandora (really, Pandora). The Time Line's price started at $2500 per channel. 3M introduced Scotch 250 mastering tapes to give a 4 dB signal-to-noise ratio improvement over 206 series tapes. Ampex claimed similar improvements with their 406 series. Pioneer had their top loudspeaker system with carbon fiber woofers and high polymer film tweeter and ultra-tweeter. White Instruments had a real-time spectrum analyzer for $3200.

The J's remain fairly inefficient. I don't think they will do well in very large rooms, no matter how much power they are fed, as very high levels cannot be obtained without the sound tending to "block up" and become confused. Up to and including the 95-dB point, however, they are capable of reproducing some of the most transparent, natural sound I have yet heard. Their frequency response is subjectively very flat over the entire audible range. Bass is clean, defined, and goes all the way down; highs are very smooth and unrestricted, and midrange definition is excellent. Imaging is one of the J's strongest points. One drawback is that precise imaging is available in only a small area of possible listening positions. Reproduction of the acoustic space of the recording, size of various instruments, height and size of voices, and front to back as well as lateral and vertical focus are amazingly accurate. This imaging and the lifelike quality (euphonic coloration?) of the midrange occasionally combine to produce a shockingly real sound i mage. This is particularly true on (what else?) some of the Fulton discs. The J is a most satisfying product which I will continue to evaluate and enjoy for many years to come. Desmond Fretz (Maryland) [Editor's Note : In Boston, there are few J Modulars, seemingly, but BAS member Gary Rancourt, who knows both Fretz and Coggins, has a pair of FMI 80's. These were auditioned by Mike Riggs, and A-B compared with (large) Advents. According to Riggs, the FMI 80's "do indeed have excellent definition, focus, and depth resolutionbetter than that of the Advents and remarkable for speakers in their price class. The deep bass is missing and the upper treble is a bit rough. Overall, I would characterize the sound as being less neutral than that of the Advents, perhaps a trifle boxy, but not as dry. The Advents seem to strain less at higher levels. It's a tough choice between the two, and very much a matter of taste." Jim Brinton]
Room Placement of Folded Horns
After constructing a Speakerlab SK, a speaker kit patterned after the Klipschorn, I attached two sheets of plywood to the rear of the woofer section to simulate the effect of the walls. This allowed me to move the woofer around the room while still maintaining the horn characteristics more or less intact. I first placed the woofer in the corner and took sound-level readings using 1/3rd-octave bands of pink noise at two listening positions, 24 inches and 40 inches away from the rear wall. In both listening positions, I noted resonance peaks around the 100- and 160-Hz bands (see Fig. 1). I next moved the speaker away from the corner and along the wall (see Fig. 2). Again taking sound level readings at the two listening positions, I found the 24-inch position to be still pretty bumpy, but the 40-inch position worked out reasonably well. The response still drops off pretty badly below 100 Hz, but this could be boosted with a single equalizing section. I admit this boosting would be at the expense of increased modulation distortion in the woofer, but life is a compromise. In my attempts to minimize room resonances I found it useful first to place the speaker in a convenient position and then run through the sound pressure measurements at a convenient listener position. Pick the worst resonant peaks, play a sine wave of that frequency through the system, and walk around the room listening to the tone to detect a dip in the sound. Reposition the sound level meter here and run through the sound level measurement with 1/3 octave noise again. Don't be surprised if the resulting curve is different, but not better. Or you may find a speakerlistener combination that is acoustically good, but impossible to live with. At this point you may become philosophic, take up drink, or place the speaker in a different position and try again. If you can come up with a reasonable speaker-listener position, you may be able to save money by simplifying the equalization requirement to the point that remaining problems could be solved by adjusting the bass tone controls or by building a small equalizer with two or three sections tuned to your particular resonance peaks and dips. Jim Nichol (Massachusetts)

Book Reviews

Audio Alternative: The Definitive Guide to High Fidelity , by Mark Tobak. $8.40 postpaid from Tobey Publishing, Box 428, New Canaan, Connecticut 06840. This book would be much better had its author elected to explore systematically the technical and practical problems of buying, installing, and maintaining a high-quality stereo system. If Tobak had really done his homework, he could have written a solid, genuinely valuable consumerist book about high fidelity. Instead, he's given us a large (10 - by 14 -inch) paperbound tome filled with photos, superficial information, misinformation, and folklore. In the first half, he discusses the various components of a system and reviews some individual products in each component category. The reviews are always subjective (apparently digests of reviews that have appeared in The Stereophile and in The Absolute Sound), frequently skimpy, and sometimes misinformed or out of date. We read praise of the Citation 11a's "subsonic" filter and damnation of Crown's failure to include one in their IC-150. It doesn't take much effort to look up the High Fidelity test reports that show the action of Crown's more accurately labeled low-cut filter to be virtually identical to that of H/K's "subsonic" filter. The author says nothing about the Ortofon cartridges but gives us three paragraphs on the BSR 310AXE changer. And he seems to recommend Sherwood receivers, despite their low phono overload margins, notch distortion, and poor reliability records. His general discussions leave the impression that he is less knowledgeable than the typical BAS member, who doubtless does not consider
himself qualified to write a "definitive guide to high fidelity." Tobak thinks the FCC puts a 50 Hz lower limit on FM broadcast frequency response, that a tonearm's effective mass is its most important characteristic (the lower the better) and that it has something to do with the arm's weight (true enough, but misleading if not explained), that one can safely rely on manufacturers' speaker impedance specifications, that the average audiophile can do a good job of aligning his kit-built tuner, and on and on and on. The second half of Audio Alternative talks about buying equipment, about building equipment, about installing what one has bought or built, and about getting it repaired when it fails. It's a better effort than the first half, but not much. The author provides few hard facts, little useful advice on how to cull the real gems from the rhinestones, and virtually no insight into the critical interfacing problems that are such important determinants of a system's total performance. At the end, he reviews some books and the audio press, praising the "underground" and rapping the American slicks, which is all well and good as far as it goes. Gordon Holt and Harry Pearson deserve praise for much of what they do, and Stereo Review merits censure for what it conveniently neglects. But the indepdenents have deficiencies that should be pointed out to beginners, and the commercials, however much they may limit themselves, do contain information that the consumer can use, if he's told how to find and interpret it. Again, Tobak's consumerism, wellintentioned though it be, comes up lame. There are many worse books on high fidelity, but that doesn't make this one good. It contains too many simple-minded mistakes and foggy generaliza Michael Riggs (Massachusetts) tions to be taken seriously. And on the same book : I just got the most impressive looking hi-fi book I have ever seen. called Audio Alternative. It seems not quite up to date. The technical editor is one George Ritschner, a recording engineer; technical consultant, one J. Gordon Holt. It seems to be a compilation of material from the underground pubs plus others from here and abroad. Tobak, the editor, advises one to approach his views cynically: is there any other way? It can be interesting reading, though; either basic or slightly advanced depending on where you are. I am tempted to forget the whole business, learn to play the electronic organ, and attend some live Nate Garfinkle (California) performances. Modern RecordingA Review of a New_Magazine The second issue of Modern Recording (Dec./Jan. 1976) is now on the street and some early reactions are in order. Conceived and published by one Vincent Testa, who has successfully been conducting recording seminars around the country (dba Recording. Institute of America), the new publication purports "to serve today's music/recording-conscious society" with articles of interest on the process of recording. Also promised are profiles of major engineers, producers, and artists, with concentration on actual studio techniques, mike placement, mixing, etc. In equipment reviews, they promise to keep to hardware not usually treated by hi-fi magazines and to avoid restraint if they come across a "less-than-topflight" piece of equipment. How do the first two issues shape up? First, the negatives. The equipment reviews, with Len Feldman doing the measuring and Norman Eisenberg the listening, are a cut (albeit a thin one) above what one expects from hi-fi mags generally, and of the eight test reports in the first two issues, only two of the products reviewed (the Teac A-6100 recorder and the Sound Workshop 882M mixer) can make any serious claim to being high-end items of interest to the recordist. The people stories ("A Session with Loggins and Messina") offer the neophyte an interesting look inside a real recording session, including value judgments on equipment and techniques, but offer little helpful information to the experienced recordist. And, finally, one is chagrined to note that advertiser support, although it includes some professional equipment manufacturers, also includes a number of "hi-fi" lines. One wonders what effect that will have on editorial content.

Radio-Electronics, Dec. 19'75 For those of you who want a space-shuttle control panel in your den, the Marantz 4400 is reviewed. (p. 44) Bookshelf Speakers, Part III: A bit more useful than the previous installments in this series; speaker placement and control settings are discussed. (p. 51) Power FET' s for Audio: This is the longest and most complete description of the FET amplifier system to appear to date in newsstand publications. (p. 58) Stereo Review, Dec. 1975 Tuners: Ho-hum. Music on the Air: Perhaps the rapid loss of classical music FM stations has begun to taper to a stable plateau. Includes a box about Richard L. Kaye, Executive Vice President of WCRB, Boston, describing his career and some of his pioneering broadcasting techniques. One otherthan-obvious fact is that he adds "a very small amount of reverberation" to broadcasts from Avery Fisher Hall (in New York's Lincoln Center). He admits, "I know some people will disagree with me on that. I'm trying to produce a sound the listener can relate to in terms of being in a concert hallnot what he is likely to hear on record." (p. 62)
The Stereophile, Summer 1975 Full reports on the Dyna PAT-5 and the dbx noise reduction gear; "quickies" covering the Polk Model 9 speaker, the Yamaha CT-7000, the FMI J-Modular, the EPI 100, the IAD expander, the Yamaha NS-1000 speaker, the ESS AMT-1, the Denon 103 and 103S cartridges, the Audioanalyst A-100X, the Infinity SS-1A, the Otari MX-5050 tape recorder. the Russound QT-1 patch bay, the Sony TC-161SD, the Infinity Monitor II, and the Zerostat, and a comparative report on the Dyna 400 and Ampzilla. Eight very nice pages of record reviews and two pages of mediocre (at best) rock reviews (Holt should consign Karagianis to the cutout bin). He drops hints that there will be a new magazine soon, and a record, and brings up a hitherto unsuspected compatibility problem between tape recorders.
StereOpus, Vol. 1, No. 2 Full reviews of the Audio Research SP-3a-1, Ampzilla, the Crown DC-300A, the B&O MMC 6000, the ADC-XLM Mk. II, the All-Test phono preamp, the RTR ESR-6, the Infinity Monitor II, the Rectilinear 5, and the Braun L810. Short reports on the Dyna PAT-5, the Koss Model One full-range electrostatic, and the Shure V-15 Type III. The Constructor's Corner reports on building Ampzilla and on sources of raw speakers. A little better issue this time out, though still not quite up to its brethren. A second reviewer has joined the staff.

Wireless World, Oct. 1975 In the letters section, the first item may inspire some members to contribute their ideas (e.g., as in The Idea File column) even if they are not sure they are absolutely correct. (p. 465) Television Tuner Design: This sounds like a good article, but one look at the diagrams will scare off all but the professional designer. Electronic Circuit Calculations Simplified, Part V: Discusses the rolloff frequency of RC networks and gives some simple rules for their design and for their combination to give complex response curves (e.g., in RIAA networks). Of great use to the audiophile.

November BAS Meeting

Business Meeting Over 70 BAS members and guests crowded into Boston University's Sherman Union for the November meeting. Due to time constraints, discussion of changes in the bylaws and quorum requirements presented in the October Speaker were postponed. An overseas record-buying service is being organized by Dr. Brian W. A. Leeming to purchase discs from a British firm on a prepaid order basis. Purchases will be made only after total orders reach $100 or more, since a 5% discount is available above this figure. Orders will be placed and records shipped by air in the interest of expediency, although this will add about $0.50 per record in shipping cost. A subscriber to the Gramophone , Dr. Lemming is willing to act as an advisor on recommended recordings. For more information, contact Dr. Leeming through Box Seven. After a year of work by BAS members on packaging and circuitry for the Thermo Electron 814 and 814C electret microphone capsules, Thermo Electron has made the decision to drop this line of business and has sold the patent rights and tooling for their complete line of mike capsules to Knowles Electronics in Chicago (their largest competitor in hearing aid mikes). As far as could be determined Knowles has no plans to produce the 814 or 814C, although a smaller capsule, used by Group 128 for their mike, will be produced by Knowles. Not much is known in BAS circles about how the smaller capsule compares acoustically with the 814's. However, the electrical connections are via solder terminals rather than pins ( a la 814). Initial attempts to solder to 814's resulted in several damaged units and does not appear to be a recommended procedure. Peter Mitchell has managed to obtain about ten 814's and a few experimental 814C's which had been modified by Thermo Electron but still meet specifications. They will be available at the December meeting on a first-come, first-served basis. A short note on the BAS oscillator for those who have built kits. The value of the 39-ohm feedback resistor may need to be changed, due to variability in the amplitude-stabilizing lamp. This may be done by substituting values or using a 500-ohm trimpot in parallel with a 47-ohm resistor. (See also "BAS Oscillator Feedback," BAS Speaker , Oct. 1975.) Some questions were raised on the validity of judging preamp quality based on Holman tests of single samples, as reported in last month's Speaker. Jim Brinton noted that the samples are statistically small but felt the rankings were indicative of a trend and could be used as a general indication of quality. As a matter of record, Jim also revealed that listening tests with the MicroAcoustics QDC-1 cartridge, reviewed a few months ago, were made with Marantz 7C and Audio Research SP-3A preamps, both good performers in the Holman tests. Thus the conclusions reached remain valid. (See "More Holmania," in this issue.) Meeting FeatureReducing Distortion in Tape Recorders David Griesinger has had a long-standing interest in acoustics, electronics, and the relationship of what can be heard to what can be measured. He heads his own recording company and has recorded for Nonesuch and other labels. Besides describing the principle and operation of a predistortion system that cuts distortion in tape recorders by a factor of 10, he talked about microphone and speaker placement to achieve more realistic musical reproduction. The section of his lecture on predistortion covered so many topics so interestingly and thoroughly that rather than risk making errors in transposition of his material to text, we are reprinting (with his permission and that of the Journal of the Audio Engineering Society ) the full text of Griesinger's JAES paper of March 1975 in which much of the same territory was covered in detail.

DISTORTION MEASUREMENTS

The exact nonlinearity of the tape was determined by measuring harmonic distortion as a function of tape flux level. These measurements were made on a studio tape recorder with variable bias current. The recorder electronics had very low harmonic distortion. The tape flux level was measured with an averaging voltme-
ter connected to the output of the playback preamplifier. All levels in this paper are thus total nonfiltered tape flux levels, relative to a fluxivity of 185 nWb/m (the "Ampex operating level"), in decibels. This fluxivity usually corresponds to a deflection of 0 dB on the volume indicator. Harmonic distortion in a tape recorder is tricky to measure. There is enough wow and noise to make a notch-type fundamental canceling distortion meter unusable below about 1 % distortion at 500 Hz. When predistortion is used, there is also a rise in the modulation noise at the frequency of the third harmonic. A wave analyzer with a filter width of 12 Hz was used in these measurements. The filter width of 12 Hz was found to be wide enough to be independent of wow at most frequencies, and yet was narrow enough to exclude the modulation noise. Measurements of harmonic distortion with third-octave filters tend to be higher. The results of third-harmonic measurements without predistortion for several different brands of tape are shown in Fig. 1. The major inaccuracy in these measurements occurs in determining the flux level of the recorded tone ( 0.3 dB). Cross-field bias was used [1]. (Cross-field bias will be explained in detail later in the paper.) Each tape was biased for minimum distortion at a flux level of 0 dB at 500 Hz. This criterion for biasing was chosen to simplify comparisons between these tapes. However, with cross-field bias the bias current which produces minimum distortion at 500 Hz severely alters the shape of the recording zone. The result is lower harmonic content when the tape is played, but very poor high-frequency response. For best operation of the predistorter the bias should be 12 dB less. Harmonic distortion depends greatly on the design of the recording head. The first measurements used a metal two-channel Ampex recording head with a 25-m gap length. The bias frequency was 200 kHz. This head produced about 0.6 % harmonic distortion at a flux level of 0 dB with Ampex 406 tape. A Philips ferrite head with a 12-m gap length produced 0.3 % distortion under the same conditions. Adding cross-field bias reduced the distortion with both heads. The data shown in Fig. 1 were taken with the Philips head using cross-field bias, and they show 0.2 % distortion at a flux level of 0 dB. When biased for best operation of the predistorter, this head produced 0.3 % third harmonic distortion. The Ampex metal head also produced 0.3 % third harmonic distortion when biased for best performance of the predistorter, and the distortion did not improve when the bias was increased. This head is currently being used in this machine. (After only a hundred hours use the ferrite head shows ablation of the ferrite material at the edges of the gap. Further, at high frequencies the metal head produces a higher level without saturation.) These measurements support other reports that ferrite recording heads produce lower distortion than metal heads, at least at low frequencies [2]. It is not obvious why this is so. The heads used in these experiments were different in permeability, gap length, gap spacer material, and high-frequency loss. The fact that both the metal head and the ferrite head produce similar distortion when used with cross-field bias might be an important clue. Further research should be done in this area, preferably by an organization with access to a great variety of heads. An improvement in the distortion properties of conventional tape machines might result. The results of the distortion measurements are consistent with the results reported by Langevin [3]. They can be summarized as follows: First, if the bias waveform is very good and the recording head is completely demagnetized, there is very little even-harmonic distortion. This is equivalent to saying that the tape transfer characteristic is completely symmetric. Second, the amplitudes of harmonics higher than the third are very small. Third, when plotted on a loglog scale such as Fig. 1, the

Fig. 1. Harmonic distortion as a function of tape flux level. No predistortion. AScotch 203 tape; SScotch 206 tape; AAmpex 406 tape; MMaxell UD 50 tape; TTDK SD 150 tape. Philips ferrite recording head, cross-field bias, 15 in/s (380 mm/s), 500 Hz. third-harmonic distortion data can be fit by a straight line, at least over the most important part of its range. This is the fact which makes a good predistorter practical.
DEVELOPMENT OF THE PREDISTORTER
The shape of the curves in Fig. 1 suggest that the tape nonlinearity may be closely approximated by a power law.' Specifically, if is the flux recorded on the tape and i is the recording current,

when i > 0

+ when i < O. The constant x relates the recording field produced by the recording current to the recorded flux at low levels. N -1 is given by the slope of the distortion data in Fig. 1, and N varies from 2.5 to 4 in the tapes tested. The constant C is best found by trial and error on a given recorder. However, it can be computed from the data in Figure 1. The value comes out to be about 0.05 for Scotch 202 if i is normalized to 1 at a flux level of 0 dB. Guided by the above measurements, several circuits were constructed using analog computer techniques to produce the exact inverse of Eq. (1). After a long and sometimes frustrating development, the type of circuit shown in Fig. 2 was chosen. The majority of the circuit is designed to compute a power of the signal at its input. This part of the circuit is then placed inside the feedback loop of the recording amplifier to generate the exact inverse of Eq. (1). The first step in finding the power is to find the absolute value of the signal using an active rectifier. The next step takes the logarithm of this absolute value. An amplifier then multiplies the logarithm by the value selected for N. The antilog is then taken, giving the absolute value of the input signal raised to
Eq. (1) is only valid for low values of i. As tape approaches saturation, the recorded flux approaches a constant value, whereas Eq. (1) predicts a continuous decrease. In the predistorter design a soft clipper, along with some feedback around the power circuit, partially correct for this error. There is not enough bandwidth in the recording process, and tape is not uniform enough from reel to reel to correct the distortion at high levels perfectly. It is better to design the predistorter to produce as few high-order harmonics as possible, and record at slightly lower level. For a very interesting discussion of the interaction between bandwidth, disc predistortion, and level, see [4].
Fig. 2. Block diagram of predistorter. the N th power. This signal is soft clipped if necessary, and the sign is reinserted with an electronic switch. Timing errors in the switch do not produce crossover distortion, since the switching takes place only when the correction signal and its first two derivatives are zero (Fig. 3). This circuit has many advantages. The constant C can be adjusted by means of a dc voltage, which means that several of these circuits may be ganged to one control. Thus a multichannel recorder may be easily adjusted for different tapes or bias levels. The exponent N is set by a single resistor, the value of which can be computed from the data in Fig. 1. The value of N need not be changed for most of the tapes in common professional use (N = 3). The circuit is inherently symmetric; both the positive and the negative parts of the signal are affected identically. The circuit is completely temperature independent, stable, and dc coupled. Parts for the circuit cost about $8.00 in small quantities. Fig. 4 shows the performance of this circuit using Maxell UD-50. At flux levels below +8 dB the distortion is almost completely removed. These measurements were made by setting the predistorter for minimum distortion at a flux level of +8 dB at 500 Hz. The distortion does not remain quite as low at other frequencies, but at low frequencies it stays below 0.1% for any flux level below +4 dB. Notice that at the higher levels the error bars on the distortion-reduced data are rather large. This unevenness in the distortion is due to the lack of uniformity in the coating of the tape. About once a revolution of the supply reel there is a quick jump in the distortion. The size of the jump shown in Fig. 4 is about average for many brands of tape, although some individual reels may be perfectly uniform. Such a jump in distortion is usually accompanied by a slight drop in the tape output, and this drop can sometimes be heard if a noise reduction system is in use. Even with these faults the improvement in harmonic distortion is dramatic and immediately noticeable by ear when pure tones are played. The decrease in intermodulation distortion when two randomly selected tones are recorded at the same time is even more dramatic.

F ig. 3. Performance of circuit that raises input voltage to N th power ( N +3).
Fig. 5. Response of predistorter to asymmetric signal. The input consists of 1000 Hz and 500 Hz mixed equally. Notice that although the input has no dc component, the output appears to have a strong dc component. The measured dc component is + 0.5 major division. listening to the playback of the oscillator through the filter, the third harmonic can be easily nulled to less than 0.1%. Changing the control for C has no effect on the output level of the machine except at very high levels.
Fig. 6. Phase and amplitude of correction signal as a function of frequency if low distortion is to maintained at high frequencies and conventional bias is used. Scotch 202 tape, 7.5 in/s (380 mm/s), Philips ferrite recording head, conventional bias. wavelength of its third harmonic are large compared to the length of this critical region, the predistorter can be expected to work well. However, as the wavelength of the third harmonic becomes comparable in size to the critical region, the situation becomes very complicated. Keep in mind that we are trying to cancel harmonics generated in the tape by recording over them with their exact inverses generated by the predistorter. If we assume that the harmonics generated by the tape nonlinearity are distributed uniformly through the critical region (i.e., that they accompany the fundamental), the phase data of Fig. 6 imply that the harmonics. generated by the predistorter are recorded toward the trailing edge of the critical region. Thus they must be slightly delayed by a network if they are to still cancel the harmonics generated by the tape. However, Fig. 6 also shows an amplitude effect. Apparently the harmonics generated by the tape must partially cancel each other when their wavelengths become short compared with the critical region. Unless the amplitude of the harmonics created by the predistorter are reduced, more distortion will be recorded on the tape than was there originally. This is the major reason that the performance of a simple linearizer is so poor at high frequencies. Designing a network which would compensate for both the amplitude and phase shifts at the same time turned out to be very difficult. We had some success at 15 in/s (380 mm/s), through use of a delay line. However, two serious problems became apparent immediately. First, such a network is very dependent on the tape characteristics. A small change in the tape orientation or oxide thickness renders the network useless. Second, a network which successfully cancels harmonic distortion may cause an increase in intermodulation distortion when two high-frequency tones are recorded at high level. The low-frequency intermodulation products from two such tones may be increased by the action of the 2 linearizer. It is remotely possible that these problems could be reduced by very careful design of a conventional record head. However, experiments with different heads were not encouraging. Many give shifts similar to the Philips head, and the Ampex alloy head has shifts about 1.5 times worse. The best solution is to use cross-field bias.

He feels that the key to recreation of hall ambience is reproduction of its low-frequency characteristics, reverberation time, resonance modes, and the arrival time of discrete reflections. That each of these factors is important in aurally distinguishing the size of a room can be appreciated by noting that when only one characteristic, reverberation time, is made equal for a large and a small room, the two rooms still will not "sound" alike. To completely reproduce these large-hall effects in the home is difficult and in some instances impractical. However, proper microphone and speaker placement can yield a substantial improvement over what is experienced with most commercial recordings. Beginning at the recording end, the best microphone techniques are those that preserve phase information and pick up at least half of the recorded sound from the reverberant field. Phase information in the direct sound aids in localization of the instruments. In the reverberant field, after steady-state conditions have been reached, phase information is ambiguous as it has been "homogenized" through multiple reflections within the room. However, during transients, when an instrument begins to play a note, the phase relationships at the mikes change as the reverberant field builds up. This rate-of-change-of-phase information depends upon the size of the room and its reverberation time, and its preservation aids the ear in recreating the acoustic "sense" of the hall. A good microphone arrangement for sensing and recording these phase relationships is the spaced, omnidirectional pair. This arrangement can, however, lead to broad sonic images of soloists with the wide spacings needed for large ensembles. To recover localization in these instances coincidence mikes should be used for center fill and be mixed very carefully with the stereo pair to avoid washing out the phase information. The coincidence mikes also are necessary if the stereo signal is ever going to be converted to monaural, since spaced mikes often don't produce good monaural signals due to phase cancellation when the two signals are added. Recording with coincidence mikes alone, though, tends to give the feeling that the sound is coming to you in your livingroom through a large picture window beyond which is the concert hall. Speaker Placement. It is well known that at frequencies below a few hundred hertz sounds tend to be non-directional, and therefore carry little stereo information. But, phase differences between channels can produce audible effects even at very low frequencies. This was demonstrated by playing separate low-frequency tones through each speaker with each tone having a frequency difference of 1 Hz. The effect created as the two sources went in and out of phase at a 1-Hz rate was that of sound washing around the room, but with no particular point of origin evident. Speakers usually can be placed to optimize the effect of low-frequency phase differences recorded with a spaced-pair mike setup. Often, for aesthetic reasons, speakers are placed symmetrically in a room. When considered in terms of how the speakers couple to the standingwave modes of the room, however, asymmetric placement may be a better choice. In Fig. 10, two speakers are shown placed on adjacent walls such that one optimally excites the standingwave room resonance along the room's short dimension while the other most efficiently excites a resonance at the same frequency along the room's long dimension. When played simultaneously and in phase, the standing wave patterns of each speaker overlap to form the two-dimendimensional pattern in Fig. 11a. The 180 out-of-phase condition is shown in Fig. lib. Note that the shaded and open areas, representing regions of maximum and minimum sound intensity, respectively, reverse positions when the two speakers shift their relative phase by 180. A listener standing at or near the center of one of the shaded areas would experience a feeling of movement of bass energy in the room with phase differences in the signals fed to the two speakers.

Table 1. Some Popular and Otherwise Amplifiers Headroom Above One Watt, 2 dB 21.8 23.0 18.8 21.8 19.5 23.0 23.0 24.0 24.0 17.8 14.5 19.0 21.9 21.8 24.0 18.8 23.0 23.0 17.8 20.0 24.0 17.0 20.2 24.8 17.0 23.0 25.4 21.0 18.8 21.8 17.0 20.0 23.0 24.8 22.3 24.8 21.8 20.4 21.8
Amplifier Accuphase P300 Ampzilla ARC Dual 76A ARC Dual 150 BGW 250B BGW 500D BGW 750A BGW 1000 Bose 1801 Citation 12 Crown D60 Crown D150A Crown DC300A Dunlap-Clarke 500 Dunlap-Clarke 1000 Dynaco Stereo 150 Dynaco Stereo 400 Dynaco Stereo 410 Futterman H-3A Futterman H-4 Infinity 500DSP McIntosh 2305 McIntosh 2105 McIntosh 2300 Paoli 60M (Stereo Pair) Phase Linear 400 Phase Linear 700B Quatre Quintessence I Quintessence II SAE B31B SAE B4DM SAE B3CM SAE B25/2500 Sansui BA3000 Sansui BA5000 Stax DA300 Technics SE9600 Yamaha B-1
FTC Rated Power, watts (1 ch @ 8 load) 150
List Price, 1 Nearest $800 1600
$List/ RMS Watt (1 Channel) 3 5.00 4.00 16.00 13.33 5.60 4.20 4.90 5.60 3.96 5.00 9.64 6.00 5.16 5.33 4.80 4.80 3.63 3.00 5.83 5.00 7.40 10.00 6.20 4.00 12.00 2.50 2.30 2.60 11.33 8.00 5.00 6.00 5.00 4.17 5.29 4.33 24.00 7.27 10.67
$ List/dB Headroom 4 34.50 34.80 64.00 91.90 25.60 36.50 42.60 58.30 41.30 16.90 18.70 25.00 36.50 36.80 50.00 19.20 31.50 26.10 19.70 25.00 77.20 29.40 32.20 52.50 35.30 21.70 31.50 15.70 45.30 55.10 14.70 30.00 43.50 50.50 40.40 52.50 165.40 39.20 73.50
1. Not all prices are "fair traded" or otherwise fixed. Dynaco is a good example: the Stereo 400 is usually available for about $460, or $2.30/watt and $20/dB. 2. Headroom referenced to 1 RMS watt = 10 x log 10 (FTC power, RMS watts one channel), dB. 3. This is: (List Price) (FTC RMS power @ 8, one channel). 4. This is: (List Price) (Value from Note 2).
Table 2. Ranking of Amplifiers Based on Cost Factor Least to Most Expensive $List/FTC Watt (8) 1. 2. 3. 4. 5. 6. 7-8. 7-8. 9. 10. 11. 12-13. 12-13. 14. 15-19. 15-19. 15-19. 15-19. 15-19. 20. 21. 22. 23-24. 23-24. 25. 26-27. 26-27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. Phase Linear 700B Phase Linear 400 Quatre Dynaco Stereo 410 Dynaco Stereo 400 Bose 1801 Ampzilla McIntosh 2300 SAE B25/2500 BGW 500D Sansui BA5000 Dunlap-Clarke 1000 Dynaco Stereo 150 BGW 750A Accuphase P300 Citation 12 Futterman H-4 SAE B3CM SAE B31B Crown DC300A Sansui BA3000 Dunlap-Clarke 500 BGW 250B BGW 1000 Futterman H-3A Crown D150A SAE B4DM McIntosh 2105 Technics SE9600 Infinity 500DSP Quintessence II Crown D60 McIntosh 2305 Yamaha B-1 Quintessence I Paoli 60M (Stereo Pair) ARC Dual 150 ARC Dual 76A STAX DA300 $List/dB Headroom 1. 2. 3. 4. 5. 6. 7. 8-9. 8-9. 10. 11. 12. 13. 14-15. 14-15. 16. 17. 18. 19. 20-21. 20-21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31-32. 31-32. 33. 34. 35. 36. 37. 38. 39. SAE B31B Quatre Citation 12 Crown D60 Dynaco Stereo 150 Futterman H-3A(60) Phase Linear 400 Crown D150A Futterman H-4 BGW 250B Dynaco Stereo 410 McIntosh 2305 SAE B4DM Dynaco Stereo 400 Phase Linear 700B McIntosh 2105 Accuphase P300 Ampzilla Paoli 60M (Stereo Pair) BGW 500D Crown DC300A Dunlap-Clarke 500 Technics SE9600 Sansui BA3000 Bose 1801 BGW 750A SAE B3CM Quintessence I Dunlap-Clarke 1000 SAE B25/2500 McIntosh 2300 Sansui BA5000 Quintessence II BGW 1000 ARC Dual 76A Yamaha B-1 Infinity 500DSP ARC Dual 150 STAX DA300

May 1975, No. 8

June 1975, No. 9

July 1975, No. 10

N: Phono Cartridge Noise: A New LowADC cartridges (Zwicker) P: Comments on Records, Cleaning, and the Discwasher (Maier, Zwicker, Borden, Brinton) (see also 9/75, p. 13) P: The Allison:One Speakera BAS test report (Brinton, et al.) Aug. 1975, No. 11 N: Options with the 814 (Mitchell) R: British Record Recommendations (Mitchell) N: A New MagazineSound Advice (Riggs, Mitchell) N: Measuring Speaker Impedance (Davis, comment Cohen) N: Adjusting Volume Levels (Leventhal, comment Mitchell) N: Phono Load Capacitance (Riggs, comment Mitchell) N: Mobile FM (Horrall, comment Mitchell) N: Are Some Red Apples Really Green ?comments on 6/75 headphone amp article (North, comment Mitchell) N: Addendum to Allison:One Review (Brinton, Mitchell) N: Some Interesting Test Data on Recording Tape (McKenzie, Mitchell) S: Sam Walinsky and Richard DeFritas (Hybrid Systems)time delay reverb device S: Joel Cohentime delay device S: David Ranada and David Satzmusic, performance, and recordings P: A Transformerless Balanced Line Preamp for the Phantom 814 Microphone, Peter W. Mitchell (see also 9/75, pp. 2, 4-7) P: Using the BAS Oscillator, Peter W. Mitchell (see also 9/75, p. 3) N: Errata on A Transformerless Balanced-Line Preamp for the Phantom 814 Microphone (8/75) N: Errata on Using the BAS Oscillator (8/75) N: Allison:Oneletter from Roy Allison (comment Brinton) N: A Transformerless Balanced-Line Preamp for the Phantom 814 Microphone Revisited (Davis, comment Brinton) N: Checking for Sloppy Loudspeaker Quality Control (Nichol) N: A Nichol Sound Level Meter (Nichol) N: A Cheap Earphone (Nichol) N: Everything's Up-to-Date in BuffaloBuffalo audio scene (Temple) N: Some Sound Advice for Sound Advice comment on A-B test techniques (Shanefield) N: More on Sound Advice (Mashey) N: Record Catalog AvailableCanadian Broadcasting System record Catalog (Gupta) N: Serendipity with a Technics Turntable (Leonard) N: Attention Rabco SL-8E OwnersRabco modifications (Slindee) N: Discwasher Fluid (Mashey) N: Speaker Magnetism Strikes Againspeaker magnets affect color TV (Mitchell) S: Dennis Colin, Mark Davis, Rene Jaegertheory and demonstration of phase shift N: More on Phase "Error" Cancellation in Tape Recorders (Davis) P: BAS Test Reports: (1) A Preliminary Look at the Koss Electrostatic Speakers, Alvin Foster (2) Tuner ComparisonsSequerra One, Kenwood 700T, Kenwood KT-8007, Citation XV, McIntosh MR67, Larry Hardin (3) Comparison Test: Tandberg 9100X and Sony 377, Mark Davis. P: Membership Preference Questionnaire

Tandberg of America, Inc. ". We will do any modification possible to eliminate the RFI." Note that in the full list provided with the RFI packet, many responses are not nearly so positive as the above, and that many audio manufacturers are not listed; many of the manufacturers that are listed are TV, organ, and mass-market component manufacturers. Note also that, with the bill hanging over their heads, many manufacturers will be bending over backwards to be helpful. They would rather be able to show Congress that they are solving all of the complaints for free, and that no law is required, than fall into another FTC-type ruling. If you are really interested in solving your RFI problem, write to: The Honorable Torbert H. Macdonald, Chairman, Subcommittee on Communications, Room B331, Rayburn House Office Building, U. S. House of Representatives, Washington, D.C. 20515. If not quite so interested, drop the SASE to the ARRL for their information packet. Otherwise, if you do have an RFI problem, write to the manufacturer. Chances are steadily improving Harry Zwicker (Massachusetts) that he will be helpful.
Feedback on Service: Satisfaction with EV
Words of praise for the service people at Electro-Voice and to Harry Arnold and assistant "Andrea" in particular. Recently the voice coil of a 20-year-old T-25A midrange driver opened. For $21.50, EV rebuilt and returned it via UPS. Alas, it was lost in transit. EV scoured its shelves and promptly sent me another of this long deleted speaker. A 21-year-old EV crossover network has also failed. The company has promised to repair it for no more than a nominal fee, if not for free. This kind of service on units so long out of warranty is, I think, remarkable. Even more so is the first letter I received regarding the T-25A: EV regretted having to charge anything for the rebuilding job, but they stated that the recent surge of superpowered amplifiers has forced a change of repair policy. Richard A. Wolf (Texas)

full-range electrostatic units, but so far they still beam. Even the new Koss model I has more high-frequency directionality than I would have expected. The Janszen 130 provides a good R. Graham (Massachusetts) compromise by angling the elements outwards. The Rectilinear III: Three Versions of One In response to the comments of Carlos E. Bauza ( BAS Speaker , January 1976) on the Rectilinear III, I thought readers might be interested in my experiences with this speaker. After auditioning a Rectilinear III in a showroom in 1974, I became curious about the unfamiliar sound coming from a speaker with which I am intimately familiar. I called Rectilinear. A representative stated that the midrange driver had been changed because of its low power-handling capacity (15 watts) and that the earlier midrange units were far superior. The super-amp boom in conjunction with a corresponding misuse of super amps had led to a high midrange failure rate; hence the change. It has also been suggested (by a party not connected with Rectilinear) that Philips is no longer the supplier of Rectilinear ' s drivers. I am familiar with four Rectilinear III systems and have heard several of the new models (post-1973). There is no similarity between the new and the old. The latest models have a completely anomalous character that in one case sounds more like a jukebox than a Rectilinear III. I have since been told that the Rectilinear III-a has a new bass driver; perhaps this was the jukebox. I suspect that it is these later models that have led to the verbal abuse that the Rectilinear III sometimes receives. And I suggest, if Mr. Bauza has a model manufactured between 1967 and 1971 (I don't know when the midrange was changed), that he relax and enjoy them. The Rectilinear III does in fact have the somewhat "remote" sounding character described in High Fidelity magazine, but when properly driven with a 400-watt amplifier and fused at 1 1/ 2 amps (as recommended by Rectilinear), they produce unclipped peak levels of 105 to 107 dBc in my somewhat inefficient room. Not earth shattering, but loud. Some listening fatigue may set in at these levels after extensive periods, probably due to increased driver distortion. Jon B. Elwell (Rhode Island) $/dB and Product Pricing Formulas: Comment From a High-Line Manufacturer The Boston Audio Society never ceases to amaze me. It is a unique entity that provides valuable information to its members. In the December 1975 issue I read the method of computing amplifier power [sic] on the basis of dollars per watt. While I have no quarrel with your analysis, I do find that certain criteria of costs have not been evaluated. These are the quality of components and of manufacture, and the mean time between failure (MTBF). The Barrett Group Corporation has over eight years of military experience. As a result, our disciplines are "militarized," with our equipment designed for much longer and more rigorous use than would be found in a consumer product. Certainly it costs a bit more, but such equipment is designed not to be obsoleted by next year's model and to perform predictably under adverse conditions. This brings about a new type of design philosophy when a company combines aerospace and audio technology. If all our predictions are accurate, the industry should see a pattern of similar new equipment designed for longevity of use, high reliability, non-obsolescence of design, and most of all, the value for valuea bond of mutual understanding between the customer and the manufacturer. [The cost of such qualities will lower a product's ranking on a strict "$/dB" or "$/Distortion" rating.--Ed.] Here at Sequerra we do not intend to compete against any other component product. Each audio manufacturer, for economic reasons, provides a product to the marketplace that has the potential for the highest return on his investment. This means that the product was meant to be sold in volume at a prescribed formula that insures his ability to make money. 7

On page 13 of the December issue, under "The Idea File," the BAS Speaker has again brought out what I consider the most important analysis of audio industry practices: What goes into the pricing of a piece of audio gear? May I suggest that someone from your Society visit our facility and go through our costing on the Sequerra Model I? I am sure you will be surprised to see that the Sequerra Company does not follow the established practice of other manufacturers in the industry,. [namely] pricing the end product with a formula of five times the bill of materials cost or four times the manufacturing cost, whichever is greater. This means that in a $200 retail priced item, the actual material cost would be approximately $40 and the total manufacturing price would be $50. We at the Sequerra Company do not subscribe to that formula. Let us therefore come up with a scale of audio consumer products value for value. Frederick E. Barrett [Sequerra is now located at a new address: 143-11 Archer Ave., Jamaica, N.Y. 11435.Ed.] Comment: Included in the $/dB publication was, of course, the disclaimer that the "rankings" in terms of $/dB were not at all the complete basis for selection of a power amplifier. The intent of the article was to re-examine the possible insanity of super-powered amplifiers working into super-inefficient loudspeakers. Mr. Barrett and his audio products are presently at the extreme right-wing of hi-fi, in company with perhaps Levinson, Lux, SAE, Audio Research, and commercial dbx and Burwen, and as such they merit a high price for the same initial performance specifications as the average brands. Where a product is to be in daily use, all day every day, and where a failure in a rugged remote recording location would spell catastrophe, one is forced to pay for an MTBF of several years. In the home, the audiophile must determine if the same cost is worth it. Should the audiophile's dbx 117 be as ruggedly made as its professional counterparts, or is this a waste of money better spent on sound quality somewhere else in the system? Is the subjective effect of Levinson's superb control action worth the price? These factors can never be ranked with a $/dB figure of merit. Caution is in order, however, in judging the quality of a product strictly by the name on the label. Almost all high-line manufacturers could or do offer less meticuously made wine in the same labeled bottles. With any new such product it should be proven, not assumed, that it is a Harry Zwicker member of the "first-line" family.
Three Moving Coil Cartridges
Ron Dunlap of Dunlap-Clarke, Ira Leonard, and this author have briefly compared three moving-coil cartridges: the Fidelity Research 1 Mk, II, the Supex Super SD-900E, and the Denon DL103S. Our only conclusion thus far is that except for the Denon, all exhibit a rising high-frequency response starting about 6 kHz and peaking about 20 kHz; only the Denon is reasonably flat throughout this region. All three cartridges are similar (flat) from 1,000 Hz down. As expected, the Supex and the Fidelity Research sound brighter than the Denon; however, neither the listening session nor the testing apparatus was sufficiently extensive to provide more detailed conclusions. Our measured frequency response correlates to some degree with the published specifications in the new edition of Sound Advice magazine. Interestingly, Sound Advice preferred in general the cartridges that exhibited, by their tests, about a +5-dB peak around 20 kHz. Equipment used for the tests included the CBS STR-100 test record, a Shure 3009/S2 tonearm, Dayton-Wright speakers, a Mark Levinson pre-preamp, and a Dunlap-Clarke amplifier. Alvin Foster (Massachusetts)

Onkyo 4055 Tuner

The Onkyo 4055 must be considered a top contender in the $200 tuner class. (Locally, K&L Sound, Watertown, Mass., sells it for $175.) It is extremely sensitive and very quiet, with excellent stereo separation. It seems unusually effective in suppressing the WCRB SCA "birdie." I can receive nine listenable stations below WBUR (90.9) on the dial, plus WMEA from Portland, Maine (90.1), in stereo, if conditions are right. My location is 20 miles southwest of Boston in a good reception area, but my antenna is a 1951 TV design used without a rotor. The Onkyo has a useful accessory on the back: in addition to scope outputs for multipath, a switch can be used to yield an audible signal for multipath detection. This has been found useful for orienting a friend's rotatable antenna, and I don't see why all manufacturers can't provide this feature. [We wonder if this output signal is also of use in fine-tuning a station? Also, see a related note in the January 1976 issue of the BAS Speaker. Ed.] On the negative side, the muting threshold is set too high, and pops are heard when passing unwanted stations. The appearance also strikes me as kitchy, and the tuning mechanism lacks the smooth action of some competitors, for example, the Pioneers. AM reception is rudimentary. I have previously used a Dynaco FM-3 and FM-5, and I find that the superiority of the Onkyo 4055 over the FM-5 is more significant than that of the FM-5 over the FM-3. [No Onkyo 4055 tuners were tested at the BAS tuner clinic, so comparison with, for example, the Pioneer 7500 is not immediately possible.Ed.] At about $175, check out the 4055 before buying a more expensive tuner. David F. Temple (Massachusetts)
Book Review: Hi-Fi in the Home
Hi-Fi in the Home , John Crabb, fourth edition, 1974, published by Blandford Press, London, distributed in North America by Transatlantic Arts, Inc., North Village Green, Levittown, N.Y. 11756 at $8.75. This 330-page, hard covered introduction to high fidelity is written by the editor of HiFi News and Record Review. This is an excellent volume which should not be confused with the average book of this genre. It is not aimed at casual readers with limited endurance or intelligence, as so many such books are, but rather at individuals genuinely interested in learning something nontrivial about hi-fi and music reproduction in the home. Conversely, Crabb does not assume technical sophistication on the part of the reader, but only a willingness to follow a few logical arguments through to their conclusions. It is certainly the best introduction to the subject of which I am aware, and even advanced audiophiles will find much of interest in its ten chapters. Beginning with a discussion of the nature of musical sounds, the author proceeds to deduce and define those technical specifications that are relevant to music reproduction equipment. This leads to a detailed discussion of the items that make up a hi-fi system. Following this are chapters on selecting equipment, apportioning costs among items, and installing them in the home. A short list of recommended (British) recordings is included, with selection based on outstandingly natural sound and/or superb performances. The book concludes with thoughts on future directions for sound reproduction, an extensive glossary of technical terms, and a useful bibliography of further readings on the subject. For someone just starting in high fidelity, or for anyone whose motivation is simply to improve his understanding of sound reproduction in the home, this book is highly recommended. And when a beginner friend buys this book, borrow it from him for yourself; it provides interesting and even stimulating reading. Stanley P. Lipshitz (Ontario)

Modern Recording Techniques at Discount
BAS member Robert Runstein's Modern Recording Techniques is widely held to be the best book on its topic. Indeed, it may be the only thorough treatment of the subject for those more interested in musical recording than in sound reinforcement, radio, TV, or film sound techniques. Now Runstein is offering the book to BAS members at a modest discount from its $9.95 list price. Runstein will mail copies to members postpaid in exchange for a $9.25 check. Make out checks to Robert Runstein and mail to 44 Dunsmore Avenue, No. 610, Framingham, Mass. 01701. If this seems too modest a discount, consider that most authors profit little from their books and that by purchasing through Runstein, we may keep him eating long enough to write further on the subject. The BAS plans to write further on Runstein's booka review should appear within the next two months.

In the Literature

[ Major contributions this month come from Dan Shanefield and Dana Craig.] Audio, March 1976 Behind the Scenes: In one of the bravest and most damning editorials in memory, Bert Whyte attacks the acoustics of the well-regarded Minnesota Orchestra Hall. Response in later issues should show the other side of this issue, if any. (p. 14) Short articles include one man's view that enjoyable rock-level music may not be damaging to one's hearing after all (p. 32); a new stylus-suspension mechanism in the offing from AKG (p. 24); Heyser tells about loudspeaker IM distortion tests (p. 38); and reviews of the $1100 Tandberg receiver, a precision cassette gauge for checking mechanical tape path alignment, plus a slightly ambiguous review of the EV Interface A loudspeaker system. Audio Amateur, 3/75 Highlight of this issue is a visit to Audio Research, and a discussion with William Z. Johnson, president. (p. 5) A Bilateral Clipping Indicator: A very simple amplifier clipping indicator, with LED's to indicate positive or negative music peaks; can be built by anyone. It will soon be available as an Old Colony "kit." (p. 3) What is PCM? Discusses this still futuristic signal-processing system which has application to very expensive but "noiseless" tape machines. [Odyssey Y33200 is a PCM recording; listen to its unusual but quite audible background noise before you hope for too much from this technique.] In Defense of the Ear: Defends the proposition that we should, if trained as audiophiles, be able to spot good and bad sound strictly by earwith minimal recourse to test instruments or auditing hours of test tones. Audio Engineering Society, Journal of the, Nov. 1975 A Wide-Dynamic-Range Program Equalizer: Richard Burwen describes his new equalizer, which has four shelf response controls, two peaking controls, and 120 dB dynamic range using discrete and IC op amps (Harris HA 2-911). Circuit diagram is included. (p. 722) Audio Engineering Society, Journal of the, Dec. 1975 Horn Employing a Piezoelectric Driver: Describes new family of Motorola devices. (p. 796) AM Stereophonic BroadcastingAn Historical Perspective: AM is currently looking toward stereo "and the possibility of an expanded AM quadraphonic broadcast service is suggested." SQ is favoredno surprise since the author is at CBS Technology Center. (p. 802)

really not a product of tube-designed amplifiers and that Class A amplification really does make a difference. " Note that most tube-sounding amplifiers are class AB push-pull, in which the crossover distortion is anything but absent unless the notch distortion is carefully designed out, i.e., tube sound is not the result of class A operation. High Fidelity, March 1976 The Many Paths to Noise Reduction: Opinionated but not definitive; noise reduction tradeoffs must be a matter of personal taste. No unit can be absolutely inaudible in operation; some music must be lost along with the noise. (p. 48) Rx for RF Interference: Timely, but if you have a problem don't look here for a quick-fix. (p. 56) Rossini's Barber: The TV version from PBS in January receives a rather unfavorable review in the Musical America section. (PMA-22, in many libraries) New York Times Magazine, Feb. 22, 1976 A well written "human interest" article about master flutist Jean-Pierre Rampal (the author is a flutist himself), and Monsieur Rampal is a sufficiently interesting character to make for good reading. Popular Electronics, March 1976 An LED-Readout Audio Power Meter: A build-it article for a two-in/one-out LED voltmeter which, if one assumes constant impedance (as do almost all VU meters on power amplifiers), gives "power output" readings. Until current-times-voltage multiplying units become available on the market (soon), this approximate approach will have to do. Although this unit can indicate rapidly changing peaks, there is no "hold" feature, so your eye will have to be quick to use this box. It is attractive, however, because of its simple (if brute force) approach and low cost: $38 in kit form, without box. A MITS product. (p. 35) The Care & Feeding of NiCd Batteries: Of general interest. (p. 39) Radio Electronics, March 1976 [This issue is worth a visit to the library.] Tests of the Heath Modulus System plus matching power amplifier. Although receivers are anything but the vogue within the BAS, this Heath tuner/preamplifier is worth knowing about. As usual, little can really be learned about its sound from RE's review. As for the power amplifier, it is inexpensive and low power, and with -60 dB distortion products, would not seem state of the art. (p. 33) Review of Sound Guard, the new record lubricating system from Ball Brother's Research Co. of Colorado. Impressively done review, complete with data on record wear, noise, harmonic distortion (with spectrum analyzer), and CD-4 carrier loss versus number of record plays, for lubricated versus non-lubricated discs. Read the advertisements for this system before reading the article. An expensive product, but perhaps discs could come direct from the factory pretreated to save trouble and costif the system really is as useful as this article would have us believe. (p. 41) Understanding Tape Specs and Turntables for Today's HiFi Systemsbasics. (p. 48 and p. 38) Stereo Review, March 1976 For those of you who read between the lines to find unfavorable reviews, read (after the review of the other-league Lux L-100 integrated amplifier) the Miracord 825 review: ". rumble was about -33 dB. With ARLL weighting it improved to -50 dB," followed by the Frazier Concerto review: "slightly constricted" with a "lack of openness," and finally in the Realistic STA-90 review, the statement that it is good for urban areas "but perhaps not nearly so ideal for fringe areas or for rural listeners." If not downright unfavorable, these are at least nonetoo-subtly unfavorable comments. 19

Chromium Dioxide Pro and Con: Debate between Andy Petite (pro) of Advent and Tor Sivertsen (con) of Tandberg. (p. 65) Optimizing Cassette PerformanceThe Problem of Azimuth: An interesting article that helps explain some of the vagaries of cassettes. Must reading for anyone who wants to test cassettes. (p. 68) Wireless World, Dec. 1975 Current Dumping Audio Amplifier: The new Quad 405, made in England, uses feed-forward rather than feedback to correct for non-linearities in the transistors. Both of these distortionreduction methods were invented by Harold Black of Bell Labs, but the former has previously been used mainly in automation systems. It is claimed that biasing and other adjustments are less critical here than in feedback-type circuits, and that the resulting class A amplifier has practically zero crossover distortion and improved reproducibility (i.e., all units live up to specs, not just those that have been carefully tweaked). However, two caveats as culled from the manufacturer's literature: the raw specs are not outstanding. Power is 100 watts/channel at less than 0.01% THD, but the use of limiting is suggested with Quad's own electrostatic loudspeakers, which reduces output to one-half power. Also, less than 20 watts are delivered to the loudspeaker if its impedance drops to 2 ohms, as can happen with the AR LST (among others). [ More on the Quad in a month or two.Ed.]

February BAS Meeting

Business Meeting Well over 100 members attended the February meeting, held at the now standard location ( GTE Laboratories in Waltham). Ira Leonard had copies of the latest Sheffield disc and "Fidelity First An Unrehearsed Experiment" from Insight Records for sale, and Scott Kent was selling copies of the record "Angle on Harpsichord" at $5 each. The latter is technically an excellent demonstration-quality disc made on a highly modified Revox with absolutely no noise-reduction equipment. Bob Borden expressed his satisfaction with a PLL modification that Scott Kent performed on his Kenwood KT 7000 tuner. Jim Brinton announced that the Mark Davis phono preamp has undergone some design changes that have even further reduced the noise and distortion. A final specification sheet is not yet available, but the price will be "in the $150 ballpark." Al Southwick read a scathing attack on RCA disc quality from the March 1976 issue of Consumer Reports ; perhaps this was in atonement for last month's speaker ratings. Dr. Brian Leeming announced that more orders are needed to keep the BAS overseas record buying service active; he had received only four orders. Records ordered through Dr. Leeming will cost between $4 and $6, depending on the list price of the record. Members who wish to order may do so at the next meeting. Or, if you have a Gramophone catalog, send your request for any listed items (specifying label, serial number, and price) to P.O. Box Seven. In addition to his usual recommendations from the Gramophone catalog, Dr. Leeming offered to make available to BAS members a list of 100 or so records (not all recent) that were rated highest in the Penguin Stereo Record Buying Guide. This Guide , published in England, is a compilation by three reviewers. The question of changes to the BAS bylaws on quorum requirements was raised. It was moved and voted that the bylaws remain unchanged until the next business meeting in September.

AUD: What are the hardest to fix lines of equipment? GL: Marantz is poorly designed, and requires complete chassis removal just to change a pilot lamp. CVA: You get what you pay for. The cheaper the unit, the harder it is to service, the higher the labor charge, and the higher the parts cost. Higher priced equipment is less expensive on all three counts, and the documentation is far superior, again lowering the bill and improving the final results. GL: Often the Japanese do a much better job with their repair manuals than the domestic manufacturers. Sony, Marantz, Sherwood, and Pioneer put out beautiful documentation. AUD: What companies are seen most often in the shop? GL: This can be misleading because of the large number of units actually in the field from Sherwood, Kenwood, Marantz, etc. Often it simply depends on what is on sale locally. Named for infrequent servicing were Phase Linear and SAE, while a genuinely poor line is the Sansui 5000/2000/3000 receiver line. These had to be modified by replacing the driver boards. HN: Dyna 120. He does not accept these because his fire insurance is insufficient. If you want to have an intermittent, buy a Pioneer, especially the 828 and the 747, where the $40 (cost price) tone control board must be replaced. The 424 had some problems in the beginning. Yamaha had some problems initially because of faulty transistors. SAE sometimes requires a new bank of transistors for $12 to $20. Nothing is really that good. CVA: You get what you pay for. Sees many KLH and Kenwood because they are pushed locally, but Kenwood is easy to fix. Sees very few Sherwoods (nods of agreement from the others on the panel). Likes Pioneer, in difference with HN; they are hot in the front end with fine sensitivity. Dynaco 120, 80, and SCA35 are nightmares. AR receivers and amplifiers provide lots of headaches. AUD: What are the reasons for not doing warranty work? Is it true that you don't make money on this work? (Comment from Ron Dunlap: If a company does set up a good service policy with decent repair rebates, this adds to the cost of the unit to the customer. Shouldn't those shops that refuse warranty work urge the customers to write to those manufacturers with flimsy warranty policies?) CVA: They might pay me $6 per unit; I don't do warranty work. GL: Differs. They give me free advertising and 28 filing cabinets full of documentation. Some (Panasonic TV) do pay well, but most do not. But even with the low rates, the high volume is useful for follow-on repair work after the warranty expires. Lots of customers know about him through warranty contracts. I do make money, but someone without the high volume would not. CVA: It isn't fair for me to make up for the low warranty rates by charging extra for out-ofwarranty work. I try to charge the lowest rates I possibly can. I do warranty work for four companiesthree rock amplifier companies and Akai. Akai refuses to sell me parts or service information unless I enter into a warranty agreement (a "gotcha" business). AUD: Why is QC so poor on audio products and apparently not so with, say, color TV's? GL: This is a problem with all portions of American and Japanese industryTV, cars, and audio. RCA had admitted that 20 to 25% of their equipment is not quite right out of the box, but the assumption is that the customer will not know the difference. With TV, Leger stops by each new customer's home to check out the new unit. Inspection is an overhead burden to industry and is shortchanged by management.

SK: Tuners. But measuring sensitivity is difficult without an RFI-screened room, which few of us have (H. H. Scott Co. and Leger do). Tape recorders very rarely meet spec, mainly because they are set up with a particular type of tape and, for playback EQ, with a test tape that has been used so many times that it is no longer reliable. At 15 kHz, oxide loss can reduce output by 6 to 10 dB at 7 ips for 1/4 track. Buy your tape in bulk and have the machine set up for it if possible. Deviation can be 2 dB within a single batch, even in studio tape. AUD: Can units be brought up to spec? Or are the published specs simply wrong? CVA: In general, only a touch up will bring them into spec. It is only the QC that is poor, not the design. [Or hyped specifications in the sales literature.Ed.] PM: Do any tape machines meet spec out of the box? SK: Revox. HN: Not Tandberg, until it is adjusted. AUD: How are AR products? SK: Tuners were not good. PM: Differs. All tuners tested in the BAS clinic were fine, while none of the Pioneer tuners met their sensitivity spec. Separation of the AR wasn't too good. SK: But we only see equipment that is defective. AUD: Is it true that some equipment manufacturers use two different sets of specs, one for sales literature and one for the service shop? GL: We repair to better specs than published. PM: But there are cases where the service manual, for example, for a Sony tape machine, lists twice as poor a frequency accuracy without Dolby as the sales literature states for Dolby on! AUD: And Teac has cases where the S/N ratio listed in the service manual is 10 dB poorer than that in the specifications. AUD: How should we go about selecting a good service shop? GL: You can try to look at the test equipment, but check also to see if it is actually in daily use. You will have a hard time knowing if the equipment is good, much less if it is in calibration. The Sony/Superscope (especially Superscope) in Woburn is terrible; they had a wow-and-flutter meter that couldn't possibly be used with a good deck. Even Sony test tapes are terrible. Check for shops with an IM (not THD) distortion meter. [After a lengthy discussion, and in spite of the previous comments that much equipment out of the box is not up to spec, it was not recommended that new equipment be returned immediately for checkout, except for tape devices. Usually the out of spec items will not be that far off unless a problem is audible.] SK: The surge of discounters has added to the out-of-spec problem. In the past, the dealer checked out and repaired all of the units that were sold, plus offering service for a year. PM: To what extent are manufacturers trying to make the servicing job easier? GL: They are trying more and more to improve serviceabilitywith manuals, seminars, schools, and logical troubleshooting aids. CVA: The more you pay for a product, the easier it is to service. AUD: Can you recommend some good test tapes? GL: Magnetic Reference Laboratory, advertised in the professional magazines.

SK: I disagree. Studio Sound published a review of test tapes: the only good ones they found were AGFA and BASF, available only in Germany. All others (Ampex, MRL, etc.) were found to be anomalous, especially MRL for phase alignment. The only good domestic ones I know of are 15-year-old Ampex tapes with McKnight's voice on them. I make my own on a Revox that can outperform any test tape in phase alignment. But it takes two days to make each one. PM: Don't trust any Dolby level calibration tape; their levels are 2 dB off, even those straight from Dolby Labs. GL: Yes, Advent has had a very rough time obtaining test tapes. SK: Ampex's "187-nanoweber/meter 0 VU" tapes are good. Probably the same is true of those from Standard Tape Labs. Phase accuracy may be poor, but this is not critical for most decks because tape skew is worse than misalignment unless a good transport with a back-coated tape is used. AUD: Is there any relation between the good technicans and those who are also avid audiophiles? All: No. Anyone technically skilled can do a good job at the test bench. AUD: Do you handle used equipment sales (customer items left in the store)? All: No. PM: Who are the best technicians? Trained engineers? HN: Anyone with ten years of experience is better than, say, a fresh engineer. CVA: My two best technicians were trained in the service and had a lot of experience. Trade schools and manufacturers provide the best training. GL: Engineers are the last people I would hire. And the problem of finding good help is really serious. It adds to the delay in obtaining service. Lack of logical troubleshooting instincts is the single biggest problem. AUD: Don't women make the best technicians? CVA: I have had two. One was the best and one was the worst I've had. SK: I have no employees in my shop, but I've seen many women in industry who would have been good. GL: We would hire them, but they don't apply. AUD: How about amateur radio operators? Aren't they good servicemen? GL: This used to be true. CVA: They are usually heavy on tube-type stuff, but not quite so good with solid-state. AUD: Is the "low status" of the repair technician part of the problem? GL: The attitude is the the real problem. People simply do not understand servicing. Not just fixing , but double-checking all possible sources of problems. Returning a really working unit to the customer. PM: What proportions of service problems are due to 1) wear and tear, 2) parts failure, or 3) incorrect use by the owner? GL: 98 are parts failures. CVA: In receivers and amplifiers, I usually find blown channels. I insist the customer bring in all their cables and speakers. Often these are the cause of the problem (i.e., incorrect use by the owner).

To get a more graphical idea of what is going on, I added a few other interesting amplifiers and some kits to Mashey's original list and computed the corresponding data. Many high-power amplifiers are available in metered and unmetered versions, e.g., SAE Mark IIIc/Mark III CM, Marantz 250/240, etc. Rather than rank them all, I graphed them on semi-log paper. ( The reference data are given in Table 3.) In this graph (Fig. 1), the vertical distance between amplifiers is the dB difference between them, not the power (watts) difference. Clearly the "best" amplifiers in terms of our revised figure of merit are those at the upper left corner, and the "worst" are at the lower right. The problem (and I leave it to you to solve) is to compare amplifiers lying on the diagonal from the lower left to upper rightmore power but also more money. How might we use all this data when considering a purchase? Suppose you own a Dyna Stereo 120 that is worth $200 as a trade-in or if sold to a friend. You want to buy a Phase Linear 400 at $500 list. So it will cost you $300 net for an additional 5.23 dB, or $57.36/dB. Assuming the sound quality is otherwise the same (it probably isn't), this may seem like a lot, but it may be what you need. So much for power. I myself find dynamic range to be the most important specification of a hi-fi system. Residual noise is more distracting to me than frequency response aberrations or moderate distortions, but others may feel differently. Conceivably we could invent more elaborate figures of merit, such as
where, as before, the smaller the number, the better. However, I don't think it is worth the effort. By the way, what amplifier did I finally buy? I didn't. I decided that a multiway system with several smaller amplifiers and active crossovers was a better approach than an inefficient speaker system and a monster amplifier. This, of course, leaves me pretty much on my own. So I am taking some courses in circuit design, and maybe Professor Bose's course in acoustics at MIT.
Table 3 - Reference Data for Fig. 1 ( K = kit, M = meters) Watts/ Channel (81) 70 25
Manufacturer Audio Research Bozak C/M Labs Dynaco
Model Dual 929PV (no meters) CM912 ST 120 ST120 (K) ST150 ST150 (K) ST150 (M, K) ST400 (K) ST400 (M, K) ST410 (K) 1 Citation 16 AA-1640 (M, K) AA-1640 (K) AA-1505 (K) AA-1506 (K) 200 700M M6000 M4000 M2000 M1500 MC2100 MC250 510M 240 Spec Mark XXXI B Mark IV D Mark III C TA3200F TAN8250 TAN8550 Tigersaurus 250 (K) Tiger.01 (K) Universal Tiger "B" (K) 215 (K)

List Price $ 125 139

dB Ref. 1 Watt 17.0 21.8 21.8 21.8 17.8 17.8 18.8 18.8 18.8 23.0 23.0 23.0 21.0 21.8 23.0 23.0 15.4 17.8 20.0 22.3 24.8 22.6 20.4 18.8 20.2 17.0 24.1 24.1 21.0 21.0 24.0 16.5 17.0 20.0 23.0 20.0 21.8 20.0 23.0 17.8 18.5 14.0
$/Watt $11.90 5.66 4.99 6.00 4.48 3.15 4.92 3.32 4.45 2.50 2.92 2.00 5.19 5.30 2.45 2.20 4.57 3.00 3.50 4.41 9.98 8.31 9.05 9.27 5.71 8.60 3.91 3.52 4.76 2.78 3.60 6.11 6.00 5.00 4.50 4.00 8.67 10.00 1.55 2.58 1.79 5.56
$/dB $35.00 38.94 34.36 41.28 15.11 10.62 19.63 13.24 17.77 21.70 25.39 17.35 30.90 36.47 21.30 19.13 10.39 10.11 17.50 33.63 120.77 66.15 48.77 36.97 29.70 25.29 41.49 37.34 28.57 16.67 37.50 16.67 17.65 25.00 39.13 20.00 59.63 50.00 13.43 8.71 6.76 9.93

EPI Harmon/Kardon Heath

Integral Systems Kenwood Luxman

McIntosh Marantz

Pioneer Quad SAE

SWTPC*

* Kits are mono; price shown is for two.
Fig. 1. Amplifier rankings: cost versus power
A Filing System for Audio Articles John E. Gombos
"I just bought new speakers. Now, how should I position them in my room?" "Which amplifier should I buy?" When friends asked me questions like these, I was often tantalized with a vague memory of having seen a magazine article on just such a subject, but I could not remember where. I decided I had to find some way of locating these lost articles quickly. I did so eleven years ago, and now I can find any particular article in less than a minute. At that time, I began an index card file system for all my audio magazines. This system consists of a 3- x 5-inch card for each magazine article. The cards are filed by subject. When I want to find a particular article, I look up the subject in the file, and a card tells me where to find the article. Let me show you how simple this is. Here is how each card is completed. I skim through a magazine with blank cards and a pen in hand. When I find an article to be filed, I write on the card: title of the article (or subject if the title is misleading), magazine title, date of issue, and page number. I allow two lines for this and the rest of the card is used for comments. Occasionally, I have to use more than one card for an article, as when two or more subjects are covered together. An example might be: "Which are bettertapes or records?" Here I would complete two cards and file one under tapes and one under records. This is my sorting system. I use two levels of classification (i.e., division and subdivision). The divisions are: Amplifiers, Antennas, Cassettes, Cassette Players, Eight-Track Players, Graphic Equalizers, Headphones, High-FidelityGeneral, Loudspeakers, Magnetic Tape, Microphones, Mixers, Musical Instruments, Noise Reduction, Preamplifiers, Phono Cartridges, Quad Decoders, Quad Software, Records, Room Acoustics, Tape Cartridges, Tape Recorders, Television, Tonearms, Tuners, Turntables, and Video Recording. The subdivisions are: Accessories, Commercial, Construction, Design, Installation, Maintenance, Operation, Repair, Specifications, and Testing. The divisions are self-explanatory; however, the subdivisions require explanations. Each subdivision is defined this way: AccessoriesMinor devices and software used with each major category. CommercialAll product reviews and how-to-buy articles. ConstructionHow-to-build articles. DesignArticles on theory and design. InstallationHow to set up equipment and connect it to the rest of the system. MaintenanceHow to keep equipment working in optimum condition. OperationHow to use equipment to its full capabilities. RepairHow to fix malfunctioning equipment. SpecificationsArticles defining equipment specifications. TestingHow to test the equipment.

Copyright 1976 John E. Gombos
These same subdivisions are used with each division to provide two advantages. (Obviously some do not apply to all divisions.) First, this degree of classification suffices to define specific articles. Thus it solves my original problem. Second, it lets me look across divisions at one subdivision. For example, examining all repair articles lets me put together general rules for repairing any component in my system. Two minor problems developed with the solution of my original problem. One was how to handle magazine corrections appearing in later issues. I solved this by adding a third line of data to the index card. I listed the date and page number of the correction below the date and page number of the original article. This solution has proved to be an advantage in two ways. First, it alerts me to errors in the original article, and second, I use the same technique to list continued articles so I can read the whole article as one. The second minor problem that developed was how to deal with the isolated audio article in ,non-audio magazines. I wanted to save the article, but not the magazine. The obvious solution was to remove the article and discard the magazine; however, I feared I would lose the loose pages. I decided to store these loose pages in folders. One folder was used for each file division (e.g., "Amp Folder" is used for amplifier articles). On the index card I list the folder in place of the magazine title. This allows me to locate these loose articles as quickly as the other articles. Now to make the whole system clearer, let's look at my original two questions. The first asked about speaker placement. To answer this question, I look in my files under Loudspeakers Installation. In this category, there are fifteen articles on positioning speakers in rooms. By reading these, I get the expert's advice on how to position my own speakers. My second question concerned selecting an amplifier. For an answer, I look under Amplifiers Commercial. Here are two types of articles. The first discusses amplifiers in general: what types are available and how to buy one to fill your needs. I found twenty of these. The second type is product reviews of specific amplifiers. I found 150 of these. I usually read the first type to get some idea of what to look for; then I read the second type to limit my choice to models that meet my needs. To further clarify what I have said, let me illustrate the information I found for one particular amplifier, the Dynaco Stereo 400. (Note: this list is not necessarily complete.)