by James V. White
The problem of vertical tracking error in phono pickups was brought to light seventeen years ago by Erik R. Madsen of Bang & Olufsen. A year later, more detailed evidence was published by the late Benjamin B. Bauer of CBS Technology Center, by J. D. Halter and J. G. Woodward, then of RCA Victor's Princeton Laboratories, and by Duane H. Cooper of the University of Illinois. Last year measurements by our author, James V. White, and Art Gust at CBS Technology Center, and parallel measurements by Mitchell Cotter at Verion Audio, Inc., confirmed that the problem is with us still.
THE ODDS ARE better than 100 to 1 that, astonishing though this may seem, your phonograph's sound suffers unnecessarily from as much as 5% distortion due to vertical tracking angle error. Recent data developed at CBS Technology Center show that VTA modulates the frequency of tones recorded in the groove and that 5% peak "flutter" is typical of the results with modern pickups-including models that enjoy formidable reputations.
Moreover, listening tests verify that the effects are clearly audible in music reproduction.
VTA has a simple geometrical meaning that is illustrated in Fig. 1. It should be obvious that the cartridge--and the pivot point within it--must be mounted above the record during playback so that it will not drag on the surface. The pivoted stylus shank therefore must vibrate at an angle to the vertical--the vertical tracking angle of the pickup when reproducing "vertical" modulation of the record groove. To prevent distortion in playback, the original master disc must be cut with a stylus that pivots (as shown in Fig. 2) so as to produce a vertical cutting angle that matches the tracking angle of the pickup. Any difference between the two is called VTA error.
Fig. 1: The pickup's stylus tip, in tracing vertical modulation, does
not move straight up and down, but at an angle tangent to a circle
about the effective (dynamic) pivot of the stylus assembly. The angle
between this tangent and the true vertical is called the vertical tracking
angle, or VTA.
Ideally the error should not exist, but with today's commercial products it is typically a whopping 10 degrees. Many readers will be aware that the vertical angle was standardized at 15 degrees by the RIAA in the U.S. back in the Sixties and that a German (DIN) standard, by contrast, specifies a range of angles with 20 as a central target value.
Ben Bauer, for many years the man in charge of CBS's Stamford laboratory where our recent investigation of VTA took place, was the first to conduct a comprehensive study of vertical angles, and his work in the Sixties contributed significantly to the development of cutters having angles that could be matched with pickups of practical design. A primary frustration in establishing a standard was the wide divergence among different brands of cutters and pickups at the time. In addition, the use of different measurement techniques on them were usually found to give somewhat different values for the vertical angle. This raised the question of which measurements were accurate.
Amid this turmoil, it was Bill Bachman of CBS Records, according to Bauer, who made a moderate suggestion to the RIAA standards committee: Why not pick 15 degrees? This figure had the merit of being nearly midway between the extreme values of vertical angle, ranging from less than 0 degrees to more than 40, that were attributed to the various cutters and pickups then being used. And 15 degrees was not a difficult value to achieve in cartridge design--or so it was believed at that time.
So why does the VTA problem still exist? In my view, the present state of affairs is a natural outgrowth of one simple fact: The test records that have been available for measuring VTA all give different results. Put yourself in the position of a busy pickup designer or equipment reviewer, and imagine what you would do if two authoritative measurements of VTA for a pickup you were testing differed by 10 degrees. Add to this the existence of two different standards in international trade, and I think you can see how the subject of VTA would gradually crawl under the rug, where it has been lurking for several years.
During this time, of course, significant improvements in pickups have reduced tracking forces and stylus impedances and have improved stylus geometry for reduced tracing error and record wear. But these very improvements have made the frequency-modulation distortion caused by VTA errors one of the most obvious flaws remaining.
How Big Is the Error?
The most accurate method for measuring VTA is the one that ignores all other characteristics of the pickup (and test equipment) and responds in a predictable manner only to the tracking angle. Unfortunately such a method may, technically, be prohibitively difficult to implement.
Based on the simple geometrical meaning of VTA, you might think that a direct optical measurement would he the best technique; you could observe stylus deflection with a microscope while slowly pushing against the stylus with a micrometer probe. Such techniques have been tried in several laboratories, including those at RCA and CBS. Careful observation reveals, however, that the stylus is dynamically excited by sliding in a real record groove and only statically deflected in the optical test, making its results of questionable value for assessing behavior during music reproduction. Optical measurements also are very time-consuming. The best methods rely on the use of special test records and associated playback equipment, which permit those features of waveform alteration that are caused by VTA errors to be distinguished and quantified, despite the simultaneous occurrence of other waveform changes caused by tracing error (due to the stylus geometry), groove deformation (due to vinyl or lacquer compliance), and test-record imperfections.
Advances in electronics make it possible to do routine experiments that would have been considered impractical in the Sixties, when the original VTA test records were developed. Art Gust and I, over the last two years at CBS, were able to devise two test methods that fall into this category. Both methods are based on measuring only that component of FM distortion caused by VTA error, while rejecting the simultaneous FM caused by tracing error and some other imperfections. One of the best ways to isolate this factor is to use phase-coherent FM demodulation, which can now be managed inexpensively with integrated circuits.
No new test records are needed with this approach. Another way to zero in on tracking-error FM would be to use test records with special square-wave signals, though no such records are available at present.
Fig. 2: Stylus assembly construction in cutting heads differs from
that in playback cartridges, affecting to some extent the range of
vertical cutting angles that can, in practice, be designed for. The
cutting and playback angles should agree, however.
The validity of these test methods has been verified in an extensive set of experiments with a variety of pickups. We were pleased that they produced results that agreed with those obtained with the RCA test record, which also employs an FM technique. We believe this is the first time that three different techniques using three different test records made in two different laboratories all have achieved the same VTA results to within the calibration tolerance of the experiments (estimated to be 5%). The agreement among methods was observed in experiments with ten pickups typical of today's commercial products. In addition, the FM data were in agreement with those obtained by an elaborate optical method-whose precision, even so, was relatively poor: roughly 15% for the pickups used in the test.
Fig. 3: Displacement in time of waveform elements (here with the peaks "pushed" to
the left and the valleys to the right) characterizes the distortion
due to vertical tracking angle error. The input wave shape's symmetry
is disturbed as its waveform is alternately "stretched" and "squeezed";
in other words, its frequency is modulated, introducing a form of flutter.
A practical outcome of this work is a new VTA test instrument developed at CBS Technology Center. It was conceived and nurtured by Lou Abbagnaro, designed by Gust and myself, and brought to fruition with the help of Dan Gravereaux. This instrument-together with the CBS STR-112 test record-provides an accurate and very convenient means for measuring tracking angles via coherent FM demodulation. Only one band on the record is played, and the meter immediately indicates the tracking angle of the pickup with an estimated accuracy of 5%. The meter is sensitive enough to detect light once-around variations in VTA error occasioned by record warps.
Where Do We Stand?
The test results with such equipment are startling.
With a group of ten modern pickups tracking at one gram, the smallest angle we measured was 22 degrees, the largest 33 degrees, with an average VTA of 29 degrees. In conversations with Mitch Cotter of Verion Audio, I learned that these figures are similar lo measurements he made with a digital FM method that resembles the CBS square-wave method. His work, by the way, was carried out independently but in parallel with ours.
To determine what errors are involved, we must compare these VTAs with the vertical cutting angles of commercial records. The vast majority of records are made with Westrex and Neumann cutters. When we measured five such cutters, the recording angles ranged from 15 to 22 degrees.
From this we estimate that, on the average and with a random selection of records and pickups, the most probable VTA error is about 10 degrees.
This error, again, produces peak flutter of 5% on the inner bands of commercial recordings. And, unfortunately, the effect is clearly audible in reproduced music.
HIGH FIDELITY relies, in part, on tests conducted by the CBS lab for evaluating phono pickups in its product reviews. The results of VTA measurements with the CBS meter are now a part of these tests. The new data will permit you to make a more complete appraisal of competing phono pickups based on objective measurements that are more accurate than those available in the past-including those HF has published for several years based on an established measurement technique developed in the early 1960s and used since then both at CBS and elsewhere. Recent tests show that this older technique gives lower angle values whose departure from true VTA varies from one pickup to another. As a result, any pickup measured with the VTA meter will be found to have a higher tracking angle than was measured using the older method, but the true value cannot he estimated accurately from the older data. The discrepancy between results may be due to the fact that the older method relies on the measurement of second harmonic distortion and therefore is sensitive to many pickup imperfections other than VTA error.
You must therefore expect the vertical angles shown in future cartridge reports to be larger than before--that is, until pickup designs are changed to bring the angles down to match commercial recordings more closely.
And this, it seems to me, should be the first step.
It would allow the vast store of records to be reproduced more accurately-a real gain for the music-loving consumer. Remember that you can replace your phono pickup for around $100, but you probably can't replace your record library--at least not all of it--at any price.
VTA error is one of the few causes of distortion that can be abolished very simply: by establishing a single standard vertical angle and adhering to it.
Perhaps this change to a more accurate measuring method will put a healthy pressure on both cutter and pickup manufacturers to reach agreement on an international standard and reduce the errors to the point of being negligible.
Also see: