by Peter Sutheim
I want to begin this second column by correcting a couple of well meant misattributions at the head of the first column (issue 3/79).
I was never managing editor of Radio-TV News, and it never became Radio-Electronics. (Radio-TV News went on to become Electronics World, published by a competing firm.) I was associate editor of Radio-Electronics, a Gernsback publication. The managing editor for some 25 years was Fred Shunaman, a man to whom I owe most of what I know about lean writing and tight editing, and much of what I know about how the world works. I would not wish to seem to usurp his role. [The editor's apologies.-ED.]
After a somewhat halting start six years ago, dbx, Inc., is now energetically launching 'the noiseless disc," under the direction of Jerome Ruzicka, vice president for sales and marketing. That provocative epithet identifies a line of dbx encoded records made to be played through a proprietary decoder, the model 21, but playable equally well through the company's 122, 124, and 128. I assume that readers have at least a nodding acquaintance with the dbx noise reduction system, which appeared a little over half a dozen years ago. It is a full-band, linear decibel reciprocal compression/expansion system with rms level sensing and high frequency pre-and de-emphasis.
The result of the disc version of the "dbx II" (consumer variety) system is to approximately double (in db) the theoretical dynamic range of the disc. The most immediately observable effect is the (almost) complete absence of surface noise, a phenomenon dbx shrewdly exploits by enclosing with each disc a warning to users not to expect the usual prelude of surface noise after the introductory plop of stylus into groove.
It may be hazardous to your ears, your nerves and your speakers to keep turning up the volume in those few seconds before the music starts, in search of that reassuring blend of scratch and rumble.
During an experimental broadcast of the discs on KPFK in Los Angeles, Jerry Ruzicka himself had a moment of doubt before the music actually appeared, reaching to see if the volume was up. And I've been fooled a dozen times.
Well, it certainly works. The disappearance of that steady backdrop of scrapes and ticks is a great balm. And since the decoder goes into the chain after the preamp, any preamp noise and any hum or rumble that originates with the turn table or its wiring are also nicely wiped out. However, among the records so far released, several are made from imperfect tapes--one in 1967, without Dolby noise reduction.
There, the elimination of disc noise unmasks tape hiss and, in the case of the 1967 recording, three distinct layers of print through at the end of one side.
The newer recordings are substantially better, and some of them are excellent. A recording of Scriabin works (Ruth Laredo, piano) is one of the quietest piano records this side of digital, and will probably stay quiet after more playings than a comparable direct-to-disc recording, although some slight tape hiss is evident on the dbx'd disc. The piano sound is full and well balanced, both timbrally (is that a word?) and spatially.
An especially interesting release is ''Mark Levinson Presents," a curious mixture of adulation and self-congratulation with some quite good music on it. It's a sampler com prising some jazz (with Levin son participating), some Bach (guitar and organ), some Haydn (piano) and other things. The audio quality is good to excellent. The piece for string basses and percussion called "Jungle" is musically interesting and, in its way, as demanding on a reproduction system as are some more spectacular recordings, like M & K's "Hot Stix." The record is interesting as much for its audio politics as for its content. Mark Levinson has, in general, eschewed the use of noise reduction systems in making his master tapes, preferring to go full-bore, brute force: half inch-wide tape tracks on a Studer A80 transport and his own electronics. Approached by Ruzicka, in one of those moves that seem motivated half by promotional and half by artistic impulses, to make available some of his tapes for release on a dbx ''tribute," he reportedly agreed conditionally.
He wanted the right to cancel the arrangement if the test pressings did not satisfy him.
Evidently they did. The Levin son release is numbered RTS-1 in the ''dbx Recording Technology Showcase Series." What (or whose) will follow, I wonder? A pleasant recording of guitar works (Galilei, Sor and others) by Laurindo Almeida offers beautiful playing, but "guitar" has a different meaning here compared to the Michael Newman release on Sheffield (direct-to-disc) or even the Eliot Fisk performance on the dbx Mark Levinson record. The latter two are plausible recordings of real guitars. (My vote would go to the Sheffield.) On the Almeida disc (issued as Sine Qua Non 2027), the attack is softened, the tone is lush and thick, the image is large and in distinct. The listener is almost inside the guitar.
The re-expansion of dynamic range works upward, so to speak, as well as downward, so that in solo piano and orchestral works the total dynamic range is very impressive. Unexpectedly loud orchestral passages arise from and fall back into an almost perfect silence-except for the tape hiss. I had an opportunity to compare the dbx and conven tional releases of one recording (Vox Turnabout 34169), Copland''s "Billy the Kid," "Rodeo," and ''Fanfare for the Common Man." (This is the 1967 recording I referred to earlier.) The subjective effect is that the dbx'd release is about 6dB louder than the un-encoded disc, but that would be difficult to verify. Still, coupled with the substantial drop in background noise, the improvement in dynamic range is very considerable, and the transparency of a good master tape (30 ips, ribbon microphones) is well served by the remastering with dbx encoding. What was always a noteworthy recording sounds pinched in comparison with its new incarnation. I could detect no deterioration attributable to the dbx process.
The Model 21 decoder is as close to the black-box fantasy as one could wish. A little larger than a typical brick, and much lighter and prettier, it is to be connected with ordinary RCA/phono patchcords be tween the preamp and power amp of a system, or in the tape monitor or external-processor loop of a receiver or integrated amplifier. Extra jacks on the 21 replicate the preamp jacks appropriated by the connections to the dbx unit, and there is a tape/source switch on the preamp, which can now be used to bypass the 21 or route the signal through it. Another but ton enables or disables the noise reduction (expansion) function. There are no other controls, except for a screwdriver adjustment in back that permits matching of levels with and without noise reduction. The unit has no power switch and no fuse and eats about 7 watts. The bypass mode seems to be completely passive: there is no change when the power plug is pulled.
So the hardware works well and has obviously been engineered with nontechnical consumers in mind, not gadget freaks. It has two immediate and related weak points. It introduces an incompatible element into the system (dbx records cannot be enjoyed without decoding), and the range of program materials prepared with dbx encoding is pitifully small, in part re-issues of older material which do not always make the best use of the dbx system's potential. The obvious next step is to cut dbx-II encoded tapes directly into lacquer without decoding, or digital tapes, or even to make direct-to-disc recordings with dbx-II encoding. I have heard a rumor that the latter has actual ly taken place, but as of this writing I can't confirm it. At first, dbx'd records will be available only from dbx dealers-which means audio stores, not record stores.
All of which raises some interesting ruminations about the present and the future of this system, examined in the con text of direct-to-disc and digital recording. After basking happily for several hours in music rendered from super-quiet dbx’d records, I put on a couple of non-dbx records. (I was going to say ''conventional," but they aren't, which is part of what we need to consider.) I listened to a couple of direct-to-disc records, including a Japanese-RCA 45-rpm recording of Beethoven's Appassionata, the Sheffield Michael Newman guitar release, and even some discs not in the "audiophile" realm: a couple of German and French Harmonia Mundi recordings (tape to disc) of early music, and some Swedish Proprius records. I conclude that people possibly don't know how quiet the background on a carefully made record can be without the aid of exotic signal processing. The surfaces were almost as silent as those of the dbx-encoded releases. And one of the discs had been played at least a dozen times.
Now it is true that some of those records cost more ($15) than the dbx'd ones ($8 and $12), though the Harmonia Mundi imports are about the same.
It is also true that as virgin vinyl becomes scarcer and careful workmanship and quality control become more expensive, we face a rather bleak future in disc sound at reasonable price points. One of the attractions of the dbx discs was expressed by dbx's Larry Blakely in a 1973 article in Recording Engineer/Producer. '...(L)ower quality materials may be used to obtain quiet playback from what would otherwise be excessively noisy compounds." What the dbx system does is shift the burden from the manufacturers to the consumers, in an us/them dichotomy that seems to be a rondo-like refrain these years. It tends to force record lovers who want quiet playback to purchase as specialized, one function piece of hardware for
$109 and pay borderline premium prices for a limited selection of records that can be played only through that piece of hardware, while it takes cess seems as though it ought to result in noise-free reproduction from even a sandpaper disc, it will still not eradicate some of the more egregious defects created by worn or poorly made stampers. I have encountered (not among the dbx records) once-per revolution swishes over whole sides of records that were a scant 35dB below peak record level. But, the review copies of dbx discs I received included some warps and wobbles. The $20-list Levinson record is slightly off center and slightly warped, which is reprehensible, but the real killer was the other wise superb Ruth Laredo recording of Scriabin: nothing in dbx's processing can remove a warp-wow. Very sad. In other words, the record buyer is get ting stuck and will continue to get stuck despite the earnest ef forts of companies like dbx, if the quality of record-making doesn't improve.
Companies like Sheffield and Crystal Clear tend to force the quality of records upward by setting and maintaining a standard and showing others what can be done and how to do it.
An approach like dbx's, commendable and effective though itis in itself and in the short run, tends to have the opposite effect: encouraging a slackening of quality. And therein lies my big doubt. Will widespread adoption of dbx-encoded discs result in a flood of 85dB dynamic-range recordings at moderate prices while we wait for the even greater potential benefits of digital technology? Or will it result in a perpetuation or aggravation of the current mediocrity at the same or higher prices while it gradually evolves a non-compatible duplicate catalog of releases in the dbx format? What do you think? For more information, up-to date lists of releases and record companies dbx has entered into agreements with, write dbx, Inc., 71 Chapel St., Newton, MA 02195.
--------
Of course, if the essential passivity of buying commercial '"product'" leaves you dissatisfied, you can always go out and record music to play back at home. ( I suppose that, in a sense, is also passive; it may be necessary for all of us to begin composing, performing and then recording our own music....) This is a subject for a whole feature article in its own right-something I'd like to ad dress myself to soon. But I've had good times in the past few months playing with some items related to location recording.
Two of these are micro phones, in this case from a Swedish firm, Pearl Mikrofon Laboratorium (PML), not well known in the USA, mostly, they tell me, because of deliberately restricted production and distribution which confined the microphones mainly to Western Europe.
Now PML have appointed a US distributor, Cara International, Ltd. Bill Cara graciously lent me one each of two most elegant microphones. One is the ST-8, a nearly coincident pair of dual-diaphragm con denser elements in one housing, each with its own FET preamp and a dc-to-dc converter for the polarizing supply built into the shank. The microphone is intended, of course, for single-point stereo pickup, a technique I favor, although not with the ideological zeal of some of its practitioners. Each element can ac quire any directional pattern from omni to figure-eight, with varying degrees of cardioidness in between, simply by altering the relationship of the diaphragm polarizing voltages.
This is done with two single element pots, one for each microphone unit, in a little combined power supply and control box connected to the microphone itself through a multiconductor cable.
The upper mike element is fix ed; the lower can be rotated up to a half turn with respect to the microphone body. Detent is pro vided, along with angle figures in degrees visible through a little hole. This makes it possible to set up very swiftly to do, for example, a classic Blumlein style stereo recording with two coincident figure-eight (bi-directional) microphones oriented at 90° to each other. Or, say, two cardioids at 110°. Or even, with the addition of a suitable matrixing circuit, an "M-S" recording, with a front-facing cardioid and a sideways-facing figure-8. It is even possible to use this design to good advantage in tricky monaural close miking or PA situations by combining the two outputs out of phase and nulling the patterns to some unwanted leakage or feedback-causing reflection.
Versatile dual microphones like this have existed for many years, most recording engineers are acquainted with various models from AKG and Neumann, though as far as I know, no American maker has ever offered one. What I enjoyed about this microphone is that it mysteriously combines the sweet, transparent sound of the (tube) AKG C-24 (alas, no longer made) with the compactness and convenience of the newer FET designs (like the AKG C-34 and C-422).
The second PML microphone I have been getting to know is the newer, smaller ST-82, a cardioid-only pair otherwise functionally similar to the ST-8.
above: ST-8 and ST-8/48 STEREO CONDENSER MICROPHONES
It lacks the versatility and the ac power supply of the larger microphone. So far, I have used it only for speech (radio theater and lectures), though that can be a stringent test for a microphone. It has a magnificently natural, easy, un colored sound. This micro phone needs to be powered from a little battery box or from mixer-supplied power, which makes it somewhat less convenient than the internal-battery powered electrets coming into wide use.
Now the bad news. The trouble with both of these microphones is that they are enormously expensive. The ST-8 costs $2,300 in the USA, and the ST-82 $1,800. Their competition is not significantly cheaper.
These prices put such tools almost completely outside the reach of the serious amateur, who cannot take an investment tax credit on the purchase, or work out some other scam to relieve the pain. I don't wish to deny PML or Mr. Cara a just return on their efforts or risks, and I think I understand why these microphones need to be so expensive, but I wonder why so far no manufacturer has ventured to try the market for a single-point stereo microphone of high quality costing, say, $300 US retail.
This is particularly puzzling in view of the existence of several single-point stereo mikes costing less than $100 from JVC, Radio Shack, Sony and Technics. Compared to the big guns, these are, in a sense, toys; they suffer from the usual inadequacies of cheap microphones: irregular and frequency-varying polar pattern, high hiss level from the internal preamp, and insufficiently high overload level. Yet I have used them happily for interviews, collecting sound effects, informal recording of small musical ensembles. But with single microphones, I know that when I encounter the limitations of a particular model, there is another model for twenty, fifty, a hundred dollars more that might overcome them. With single microphones there are models at every $10 increment from under $10 to nearly $1,000.
Nothing like that for single point stereo pairs.
Good project here for adventurous amateurs. How about a build-your-own double-figure eight or double-cardioid stereo microphones? Dynamic or con denser, take your choice. It would be easiest, of course, to begin with manufactured elements. Who will be the first to risk disassembling and recombining two perfectly fine microphones? :
Information on the complete line of PML microphones, including dynamics as well as condensers, is available from:
Cara International, Ltd. 4145 Via Marina, Suite 120, Marina Del Ray, CA 90291.
------------------------
Also see:
Showcase--Leach and Jung turned out under a craftsman's hand; Editorials
The Williamson 40/40, Power Amplifier--Return of an improved favorite after a decade