Good belt-drive and direct-drive turntables are more than adequate to ensure their intrinsic mechanical noise will not contribute significantly to the needle-drop noise. Both types face a common enemy: low-frequency noise imprinted in the record grooves themselves.
The speed stability of the lathe is achieved by massive moving parts which put a great demand on the bearings which support them. It has been possible to produce the main load supporting bearing in the reproduction turntable with a level of rumble performance better than that achievable in the lathe for many years. Thus we can say that the lower bound of rumble performance is set by the lathe and is imprinted forever in the medium itself.
In short, wideband, vertical-component noise is present on all records. It is due to rumble in the bearings which carry the heavy platter of the recording lathe, see panel (right).
Recalling that cutting an acetate is fundamentally an industrial machining job, we call the vertical noise component extending from very low frequencies to around 200Hz, machining noise and name the noise reduction, Machine Orthogonal Noise Elimination or Mach ONE.
Much more information is given here. The rest of this page is devoted to audio demos of the Mach ONE noise reduction process.
Here are two audio demos from discs of recordings made in from 1964, Prokofiev ballet music to Romeo and Juliet Philharmonia Orch' under Efrem Kurtz (which we have on a later LP EMI ESD7151 Matrix 2YEA.7427). First a section between two tracks. First with no processing and (after the beep) with Mach ONE processing.
Note, you are listening for rumble here, so don't listen just with your PC 'speakers: use headphones or decent monitors.
The second demo is of the track noise alone (no modulation). Once again, first with no processing and second with Mach ONE engaged. The level is greatly exaggerated to make the comparison easier to hear.
And for those of you who only have PC loudspeakers to listen, here is a two-octave pitch shift of the last demonstration. First part, unprocessed rumble: second part, processed with Mach ONE. A short tone separates the two.
Measuring the rumble performance according to DIN(A) or NAB, the unweighted rumble signal improves by 10dB to 18dB depending on the test disc employed.
The basic turntable (the headstock) of a Neumann lathe weights 30kg! With a vertical bearing force of this magnitude, the simple, single ball-bearing (typical in a turntable) is not practical and a ball-race must be employed. It is impossible to impart on a race the same degree of finish possible with a single ball-bearing.
The main bearing sleeve can be machined closely enough to constrain lateral movement of the turntable, so any unevenness in the ball-race will tend to resolve into vertical movement, as illustrated in the diagram above.
Additionally, because all lathes were conceived in the days of mono, lateral cut records, the vertical component of machining noise was long considered unimportant and the issue was never entirely addressed – even fully into the stereo era.
Pre VMS 66 (e.g. AM32 lathe)
Rumble (unweighted) Greater than 55 db down
VMS 66, VMS 70 and beyond
Rumble below 10cm/s 1000Hz: Greater than 70dB (DIN 45 539)
Mach ONE noise reduction is concerned with vertical movement of the stylus as it replays the record. By knowing how energy was limited in the vertical direction during recording, we can similarly limit the signal we extract on replay. (Digital filters allow us to do this without phase distortion.)
If the record has no significant vertical modulation, because, for example, it is monophonic, it is correct to set mono in the PHONO preferences dialogue in which case, all vertical modulation is ignored. When vertical modulation is not present Mach ONE will - in effect - be inoperative. You can leave it engaged, but it will do nothing.
By contrast, where a record has exaggerated vertical signals (for example a phase-matrix quadraphonic record), it may be better not to select Mach ONE because the process will remove some vertical modulation. The effect will be slight, but generally, we recommend Mach ONE for stereo records only.
1. REIBRAD, RIEMEN, ZENTRALMOTOR - Eine Analyse von Plattenspielerantriebskonzepten anhand der Rumpelspektren (FRICTION WHEEL, BELT, CENTRAL MOTOR - An analysis of turntable drive concepts based on rumble spectra) Klapproth, L. Paper of the 47th Convention 1974-03-26/29 Copenhagen/Denmark. This translated from the German: The measurement data shown here characterise the high-quality level of the turntables offered today. They are almost all of them much better than the available test records.
Klapproth's results for various turntable types are given below.
For all support issues, go here.
For Pspatial Audio sales, email: email@example.com
Apple Certified Developer. Stereo Lab, Aria 51, Aria 20, Head Space, Groove Sleuth, iLOOP and FRANCINSTIEN T-Sym are trademarks of Pspatial Audio. FRANCINSTIEN and Bride of FRANCINSTIEN (BoF) are trademarks of Phaedrus Audio. Macintosh and the Mac logo are trademarks of Apple Computer, Inc.