|Home | Audio Magazine | Stereo Review magazine | Good Sound | Troubleshooting|
No matter what one may feel about quadraphonic sound, there is no denying that in the development of four channel disc technology, a number of the "spin-offs" of this research have resulted in significant improvements in record playback and in the records as well.
I hope my friends in the SQ and QS camps will forgive me, if I point out that most of these "spin-offs" were derived from the technology involved with CD-4 quadraphonic sound. The complexities of this system with its 30 kHz carrier, and the necessity for playback response to 45 kHz, spawned a great deal of research from phono cartridge manufacturers. In a remarkably short time, these manufacturers produced the ultra wide range cartridges with the special Shibata stylus, or some variant of it, needed to track such high frequencies. Currently, even though quadraphonic sound has a very low profile indeed, new generations of stereo cartridges are benefitting from the research done on the CD-4 styli, and are utilizing the various stylus configurations with wide bearing areas as compared to the normal stereo styli. As most audiophiles are aware, in order to record the 30 kHz carrier, CD-4 discs were cut at half the normal 33-1/3 rpm speed. Presently, very few CD-4 records are being cut, but this very specialized half speed disc cutting technique is now being used for stereo disc cutting. It is probably safe to say that this half speed technique is the premier method of cutting stereo discs today, and the results are absolutely superb. For a sample, listen to the Stanley Clarke recording School Days (Nemperor Records NE439).
Victor Company of Japan set up the JVC Cutting Center in Hollywood during the heyday of the CD-4 disc, as an aid to companies wanting to produce records in this format. The Center was equipped with the latest Neumann lathe and Neumann SX 74 cutterhead, along with the specialized electronics" and ancillary equipment for half-speed (16-2/3 rpm) cutting. Nowadays, quality conscious producers are using the-JVC Cutting Center to have their stereo tape masters processed into lacquers cut with the half speed technique.
I have visited the Center a number of times, and this past spring I was there again. On this visit I was aware of the increased activity in half speed stereo disc cutting, and I had the opportunity to delve into some details of this technique. Presiding over the cutting lab operations is Mr. Tom Nishida, with Jim Kawada (who was my gracious host in Japan a few years ago) as CD-4 specialist, and the redoubtable Stan Ricker as cutting engineer. Stan is a man of many parts... one of the foremost disc cutting experts in this country, a musician and authority on brass bands, a sometimes tinkerer and restorer of classic cars, and an audio enthusiast with "purist" inclinations. I had asked Stan to give me some straight talk on the advantages and problems of half speed stereo disc cutting, and he was kind enough to "walk" me through the process in the cutting room, from tape playback, to cutting, to reference lacquer inspection and evaluation. Most of what follows is the result of these chats with Stan and subsequent correspondence in writing and on cassette, with many of his remarks verbatim.
Half Speed Advantages
According to Stan, the half-speed technique is not exactly new. Decca of England used the technique in the early days of stereo to achieve the overall level and frequency response of their “ffrr” recordings, and they continued to use it until 1968 when the Neumann SX 68 cutterhead was introduced. In the U.S., RCA used half speed cutting as a "problem solver"; for example, coping with the high frequency energy in the sibilants of Harry Belefonte's voice. Stan lists the following reasons for using the half speed cutting technique: Most accurate high frequency phase response . . . Best transient response . . . Least possible high frequency crosstalk . . . Lowest possible distortion from the transfer system ... Permits use of CD-4 cutter stylus with one micron tip radius, two micron burnishing facets; and 35 degree back angle (instead of the usual stereo real-time 45 degree angle) .. . and this gives: Smoothest possible groove cutting and best possible electroplating ... Least noise and best high frequency response at smaller disc diameters.
The Neumann SX 74 cutterhead is of the dynamic motional feedback variety ... each channel has a drive coil and a feedback coil. Much in the same manner of similar motional feedback circuits in loudspeakers (Philips) and servo motor systems, the feedback coil senses the motion of the stylus, compares it to the input signal as to deviations in amplitude and phase, any errors "seen" by the feedback circuitry are corrected and fed to the drive coil.
While this seems fairly straightforward, it is important to realize that the amount of feedback control is not uniform at all frequencies. Feedback control at the primary resonance (about 1 kHz) is in excess of 35 dB, but there is quite a bit less at the higher frequencies, for example only 0.6 dB at 15 kHz, and 5.5 dB at 8 kHz. Now consider the high frequency boost of the RIAA recording curve, and you compound the problem. Obviously in the higher frequencies of music signals, less feed back control will be exerted.
Bringing the music signals down an octave, by using half speed cutting, and there will be more feedback control. A closely miked soprano voice may have sibilants with great high frequency energy peaks around 8 kHz. Trying to cut this in normal fashion in real-time, will probably trip all the circuit breakers in the cutting system. In real-time, the alternative is to lower the level so the sibilance can be cut ... but if this is done, the overall level will be unsatisfactory. The usual solution is to use program peak limiters in the transfer console, and high frequency or acceleration limiters in the cutter system itself. Either singly or in combination, this limiting provides less than optimum transfer. At 8 kHz in real-time cutting, there would be 5.5 dB of feedback control. By cutting at half speed, this 8 kHz sibilance would be shifted down one octave, and at 4 kHz the feedback control would be a much more useful 12 dB. Needless to say, by using the half speed technique more of the total music spectrum will be in the area where there is the most feedback control over the cutterhead stylus motion. It is this fact which accounts for the first three advantages outlined by Stan, inasmuch as high frequency response and phase accuracy are a function of feedback control.
Getting back to the RIAA recording pre-emphasis, consider that at 10 kHz the high frequency boost is 13.7 dB compared to the 1 kHz zero reference level. This begins to cause power problems, since every time you go up 3 dB the power requirement doubles. In real time, the normal cutting system starts to run out of power.
With a tape machine and lathe running at half speed, it obviously is going to take twice as long to cut a given piece of program ... 36 minutes to cut an 18 minute side. But by taking twice as long, it requires but one-quarter the power necessary in real time. In half speed cutting, stylus acceleration is but one-quarter of its real time counterpart. Thus using the 600 watt per channel Neumann SAL 74 cutter drive amplifier is equal to using a 2400 watt/channel amplifier in real time.
More than equal in fact, because the phase accuracy at half speed is better.
Stan pointed out that the use of the special CD-4 cutting styli has significant benefits in half speed stereo disc cutting. Early on in the CD-4 technique, it was found that the normal stereo cutting styli were simply too large at the cutting edge (burnishing facets) to engrave the very short wavelengths of the 30 kHz carrier which at the RIAA minimum diameter of 4 3/4 inches for a 12 inch/ 331/3 rpm records is 0.000285 inch! For this reason, a new cutting stylus was developed with a tip of radius of one micron and an average burnishing facet width of two microns. The length of the CD-4 cutting stylus is 1.17 mm, as compared to 2.1 mm for normal stereo cutting stylus. This is a reduction in length of some 44 percent, which greatly reduces cross-talk in the carrier range of CD-4 records, partly due to less mass to accelerate, and partly due to less "whipping action" at the stylus tip. As a result of this cutting stylus re-engineering, and the parallel development of the Shibata playback stylus, the high frequency limits, especially at inside diameters, are 3 dB higher, than if cut with a standard stereo cutting stylus at the same level and diameter. Stan states that with the CD-4 cutting stylus, half speed stereo cutting really shows its mettle in the production of 7 inch/45 rpm records with high density signal information.
Of course, these records are cut at 22.5 rpm.
Stan told me that due to the greater precision and accuracy in the manufacture of CD-4 cutting styi, that this in combination with the slow rate of cutting gives a much smoother groove wall, which ultimately translates into less noise in electroplating and pressing. He also pointed out that the cutter system itself is not the only part of the transfer chain to benefit from half speed operation. From the special glass ferrite playback head, through the custom-built amplifiers and equalizers, rise-time information is cut in half, and power demands at all points within the system are greatly reduced.
As you can see, Stan Ricker has put forth some very cogent reasons for cutting stereo records at half speed. On the negative side, I asked Stan about the notion that seems to be generally held that half speed cutting entails a loss in bass response. He said this was just not true, and said that if anyone knew the specifications of the Neumann SAL/SX 74 cutting system, they would dismiss this so-called loss of bass as a myth. He states that the low frequency output of the Neumann system is flat to 7 Hz, and his unit is down only 3 dB at 4.5 Hz! At half speed cutting these numbers translate to 14 Hz and 9 Hz respectively, and most certainly is more than adequate to cope with the low frequencies found in music. Stan also pointed out that the rest of the transfer system ... the tape repro head, repro amplifiers, graphic equalizers, and transfer console, taken totally as a unit, are down 3 dB at 11 Hz (due to some transformers in the chain) which becomes 22 Hz at real time playback. Again, this places no practical restrictions on bass response.
Quantity vs. Quality
I questioned Stan about the low per unit time of lacquer production, and he admitted this was a problem. The only solution he could see was the use of multiple lathes operating together.
He said it would require 3 lathes at half speed operation to equal in product output what one engineer and one real time cutting system could do in a
typical working day. Stan feels its the old story of quantity versus quality. He asserted that in today's record business it is not uncommon for four months of time and $85,000 and more to be invested in the final two track, quarter-inch stereo master, and when such a tape is handed to the cutting engineer, it deserves more than a cursory 45 minutes of his time and attention.
Despite the drawback of slow production, half speed stereo disc cutting has a growing reputation for high quality results and such record companies as A&M, Arista, RCA, Warner Bros., Epic, and others are opting for this technique. Obviously encouraged by the response to their mastering operation, the JVC Cutting Center is slated to become an executive branch office for the entire music division of Victor Company of Japan. The mastering facilities are being expanded, and will include real time cutting for those who want it. The Center will sell the special CD-4 cutting styli developed by Victor and the Adamant Kogyo Co. of Japan, to other mastering studios.
They already number among their customers such studios as A&M, ABC, Allen Zentz, and Mastering Lab (that is the Sheffield folks). The Center will also sell Japan Victor Company test records, currently being used by Stanton, Shure, ADC, Audio Technica, and other cartridge manufacturers. Plans are to offer the records to audio dealers and audiophiles as well. A new Evaluation Room is to be built at the Cutting Center for listening to reference lacquers, test pressings, etc.
The room will be set up to demonstrate the new Q-Biphonic sound. The Center will be able to accept orders for pressing in Japan. Of great interest to audiophiles is the fact the Center will import Victor Company records (you'll probably be able to buy the album Jaws again) as well as records from RVC (RCA Japan), Nippon Phonogram, and others. Lastly, the Center is to undertake original music recording here and will produce records for release in the U.S., Japan, and possibly world wide. The Center will continue to sell to audiophiles on a direct basis their super CD-4 demodulator, the CD-4/50. I have been using one of these units, and with its phase locked loop circuitry and noise gates, CD-4 recordings are as quiet as normal stereo productions. Separation is superb and distortion is very low. Stability is very good and the CD-4/50 is able to clean up the sound of quite a few CD-4 "problem" records to a remarkable degree. Price of the CD-4/50 demodulator is $350.00.
(Source: Audio magazine, Nov. 1977; Bert Whyte)
= = = =
Prev. | Next