Come summertime, and your peripatetic reporter usually visits some audio manufacturers in some salubrious foreign clime. This year it was off to the Revox/Studer plants in Switzerland and Germany (I'll report on this visit in due course) and to Decca Records in London.
As I'm sure you are aware, in the United States Decca records are sold under the " London" label. Last February, London Records invited members of the audio press corps to a New York hotel suite to audition what they had billed as their new line of "super high quality Dolby B pre-recorded cassettes." They had titillated us in their invitation by stating that there would be an A/B comparison test between 15 ips copymasters and the cassettes, and that we could expect to find them "indistinguishable." Well!! Presiding at the demonstration was Gerd Nathan, the genial physicist in charge of the cassette project, and none other than Arthur Haddy, Decca's legendary chief engineer, the man responsible for ffrr recordings. The room was set up with professional monitor loudspeakers, and if I remember correctly, the 15 ips tape deck was a Philips machine. The cassette units were British, made by the Neal Company, and they use the Wollensak transport with their own electronics, a la Advent. Various amplifying equipment, an oscilloscope, some other instrumentation, and an A/B switch completed the set up.
Pop and classical cassettes were demonstrated and compared with 15 ips copymasters of the same program material. Levels were precisely matched, and we were permitted to do our own A/B switching. Sure enough, these London/Decca prerecorded cassettes really were something special, because it was indeed difficult to distinguish between them and the 15 ips tape. Believe me-there were plenty of wrong guesses as to which was which! We were assured that the cassettes were standard productions, and not something specially made for this demonstration.
A word about the sound quality is in order. The cassettes unquestionably were wide in frequency response and dynamic range. What impressed me the most were the good motion, relative freedom from dropouts, and most especially the absence of modulation noise. For the first time, it was consistently possible to hear exposed flutes, oboes and clarinets, and sustained piano chords without the blurring and the "burbling" unsteady quaver which modulation noise inflicts on these instruments. In my opinion, it has been modulation noise above all other technical shortcomings of the cassette that has kept them out of contention as a high fidelity medium.
I should note that while the good points of the cassette were audibly obvious, I wasn't too happy with the reproduction in the demo room. This was mainly because of the loudspeakers. Typical of the co-called "professional monitor" breed, they rolled off rapidly below 60 Hz, had a peaked mid-range, exaggerated top end, and an overall harshness that was not only unpleasant but highly inaccurate.
Room acoustics didn't help either, contributing a boomy resonance around 80-100 Hz. Program material ranged from pop items like Mantovani and Ronnie Aldrich and quasi rock stuff, to classical music performed by the Chicago, London and Los Angeles symphony orchestras, which we know so well from the excellent London/Decca Records.
I asked Gerd Nathan what sort of duplication process Decca used to achieve such high quality results, envisioning all sorts of exotic guidance systems, with motion servos to prevent tape skew, saturation threshold detectors, etc. To my surprise, he told me that their system was fairly straightforward, with some significant modifications to standard Gauss duplication equipment, but that extremely rigid quality control was the main factor in their operation. I guess I exhibited some skepticism, and remarked that some time I would like to see the Decca cassette duplication facility in England. Some months passed, and I received the first samples of the new London/Decca cassettes. When I played these cassettes over my "state-of-the-art" system, I was truly astonished with the high quality of the sound. After years of conditioning to all of the sonic ailments that afflict the sound of cassettes, to hear a nice clean sound, with wide frequency response and dynamic range, and nary a trace of modulation noise was a most pleasant surprise. I expressed my high opinion of these cassettes to the people here at London Records in New York, and suggested that if they heard them over my sound system, they would really be impressed with the quality of their product. Some weeks later, I had as my guests Mr. H. Toller-Bond, President of London Records, his marketing manager Mr. Bernie Fass, and old friend Lee Kuby of Harman International, and we spent a most pleasant afternoon playing a wide variety of the new London cassettes. Everyone was duly impressed, and at the conclusion of our "musicale," Mr. Toller Bond very kindly suggested that I might like to visit the Decca cassette plant in England. The wheels were set in motion, arrangements made, and thus somewhat circuitously I have explained how I wound up in London on the 16th of July.
Before going into details of the Decca cassette operation it might be worthwhile to take a brief retrospective look at the cassette medium.
As you know, Philips introduced the compact cassette in 1965, and I don't think they envisioned it as anything more than a convenient mini-recorder...good for dictation and that sort of thing. Inevitably, of course, someone recorded music on the system, and ghastly though it may have been in quality, the sheer fascination of it encouraged experimentation. In due time improvements were made in tape oxides, cassette shell moulding, better slip liners, and, of course, in the cassette player drive mechanisms. The first pre-recorded cassettes appeared, and they were pretty horrible. More time went by, and more improvements in the system, yet pre-recorded cassettes suffered from high tape hiss, modulation noise, frequent dropouts, and the lack of headroom caused tape saturation and high distortion. A few years down the line and we got tape oxides with higher packing density.
Dolby B noise reduction arrived and alleviated the curse of tape hiss. Chromium dioxide was introduced with improved high frequency response.
Ampex began to duplicate Dolby B cassettes, a big step forward, but many problems remained.
The advent of high-output, low noise cassette tapes, plus recorders with good motion and electronics with relatively low distortion, and improved Permalloy and ferrite heads, provided the audio consumer with a recording medium that was impressively close to high fidelity quality. This led to the widespread practice of dubbing records to cassettes...and since the copies were made on a real-time basis at one-to-one, their quality was excellent. Thus, to most consumers, it was very frustrating to be able to make high quality recordings in their homes, but not be able to get similar quality in a pre-recorded cassette.
Alas, they still suffered from all the woes and technical inadequacies previously enumerated.
Advent set new standards of high quality in pre-recorded tapes with the use of Chromium dioxide tape and low ratio duping speeds. On the other hand, Ampex gave up the ghost and got out of the duplicating business.
Philips and Deutsche Grammophon finally decided to encode their cassettes with Dolby B noise reduction with generally good results.
In spite of these improvements in the quality of pre-recorded cassettes, most audiophiles felt that the lowly cassette couldn't challenge the high fidelity of phono discs or open-reel tapes. Undismayed by these attitudes, research and development work on cassette technology continued apace in the labs of many companies. Since the demise of Ampex Stereo Tapes, those record companies who wanted to have their productions available on pre-recorded cassettes would either have to sub-contract for their duplication, or do it themselves. London/Decca Records, always a quality oriented company, decided that if they chose to make their own prerecorded cassettes, an intensive investigation into all of the parameters involved in cassette duplication would have to be undertaken, with a view towards correcting the existing technical flaws in this medium, and thus enable the production of high fidelity cassettes. Quite a tall order! Well, back to London. And what a terribly unseasonable London... temperatures in the high eighties and humidity to match...and almost without exception...no air-conditioning! After we were ensconced at the Savoy, it was off to dinner with Arthur Haddy and his charming wife Lydia. Mr. Haddy is both chief engineer and Director of Decca Records and an altogether remarkable man. He was involved with the radar project that was of such great help in the Battle of Britain, and it was out of this experience that he evolved the idea of ffrr (full frequency range recording) and Decca records with this process (still 78 rpm of course) began to be marketed in 1947.
Mr. Haddy has always been venturesome in his pursuit of high quality sound. In 1957, I met him for the first time in New York, at the London Records offices where he was demonstrating a stereo record utilizing "hill and dale" recording for vertical modulation. The quality was excellent, better by far, in fact, than the Westrex "45/45" system that came along a few months later. But his system was not compatible, and so Decca quickly switched to the Westrex system. However, his stereo experiments had put him in a fortunate position as far as the Westrex discs were concerned, for he had been recording stereo on tape for some time, and had a large backlog of productions ready to transfer to stereo disc. Mr. Haddy has many other accomplishments, too numerous to mention. For his contributions to recording the Audio Engineering Society has made him a Fellow, and given him the Emile Berliner Award (now the silver medal). In his own country, he was on this years' "Honors List" of the Queen, and was awarded the Order of the British Empire (OBE). Along with Gerd Nathan, he has tackled the problems inherent in the process of cassette duplication with his customary zeal.
After a preliminary visit with Arthur Haddy at the Decca studios in London, Gerd Nathan drove us to the quaint little village of Bridgenorth, near the Welsh border, and some 140 miles north of London. By odd coincidence, Bridgenorth is only a few miles from Kidderminster where I was stationed during the War. In Bridge north, in what was an ancient hospital, where generations of British warriors came to heal their wounds, Decca cassettes are now manufactured. (Some of the locals aver that when there is a gibbous moon low in the sky, you can still hear the screams of the wounded.) Methinks some bloke forgot to turn off an oscillator that was set to 3 kHz! What once were long wards in the hospital are now admirably suited for the set up of the lines of master playback and the duplicating slaves.
The wards are equally advantageous in the production flow of cassette loading, labeling, packaging, etc.
Production chief Peter Banks gave us the guided tour of all the manufacturing facilities, and then along with Gerd Nathan, we sat down to discuss the details of the Decca processing. There are two Gauss Electro Physics (now Cetec) master playback machines, using one inch running masters, and equipped with the tape storage bins, and currently 20 Gauss duplicating slaves, all save one equipped with ferrite heads. More slave units are on order. On receipt of a slave unit, the heads are demounted carefully checked for physical specs, remounted on a special adjustable plate, and then undergo ultra-precise alignment of azimuth, zenith, etc. The normal tape guidance path on the slaves is changed to a proprietory configuration that avoids sharp bends and results in an exceptionally tight tape wrap on the heads. A master bias oscillator is set at 10 megaHertz and is routed to the heads of each slave.
Each slave is specially equipped with a meter and position switch to ensure that each head is receiving the proper bias. The running master operates at 240 ips, and the slaves at 60 ips, with a resultant duplicating ratio of 32 to 1.
Depending on the type of program material and its dynamic range, either BASF duplicating tape is used or a high energy Memorex oxide which affords about 8 dB more headroom. In another factory, Decca operates its own molding plant where the cassette shells are manufactured. According to Gerd Nathan, this is one of the most critical factors in a high quality cassette. He states there are over 400 separate dimensions that must precisely meet specs. Decca research has concluded that roller guides in the shell, no matter how good the pins or bearings, ultimately contribute to poor wow and flutter. In the Decca shell solid posts of slippery Delrin are used.
The Decca shells are welded, rather than screwed together, and they claim that uneven torque in the screw operation causes deformation of the shell.
The slip liner of the shell is another proprietary item and is graphite coated for lubrication. Another special Decca machine attaches the tape to the hub with tiny rods of aluminum, which are press fitted along with the tape into a slot in the hub.
Decca claims this cassette shell is so good that minimal azimuth shift occurs from tape pack displacement when the cassette is turned over to play the second program. Quite a claim, so to check it, when I came home I set up the new Tandberg TD330 cassette recorder that features adjustable azimuth on record. A built-in 10 kHz oscillator provides a signal, and by adjusting a small knob on the record head to achieve maximum reading on the right hand meter, you've set the azimuth. Now when you have done this and then simply turned the cassette over to the other channels and applied the 10 kHz signal, if the tape pack hasn't shifted, you should get a reading very close to the original adjustment. I bulk-erased a Decca cassette and ran it through this test better than 20 times...shaking the cassette, up, down and around, and, of course, the normal side turnover.
The meter reading was consistently from 1 1/2 to 2 dB lower from side to side. According to some experts I consulted at Philips, this is very good indeed. Just for the hell of it, I ran through about 8 different brand name cassettes in the same test. One was close to the Decca cassette, several more stayed within 4-5 dB, but a number of cassettes were off by as much as 10-18 dB! Back in the listening room at Bridgenorth, engineer John Baxter had a neat set-up where one could switch between the running master, the production cassette and the stereo disc of the same program. Now the running master and the stereo disc are made from the cutting master. They are second generation from the original master. The cassette is made from the running master, and it is third generation. Without Dolby, things would get to be intolerably noisy.
With Dolby, the cassette is just lightly noisier than the running master. In making the running master, the cutting master is decoded through the Dolby A units, and then encoded to Dolby B, through the Dolby 320 unit.
Next month we will report on the results of switching between these formats, which might surprise some people, and go on to some tests concerning signal-to-noise ratio, dynamic range, equalization and distortion, then on to the Decca studios for a look see.
(Source: Audio magazine; Bert Whyte)
= = = =