|
78 rpm
Nov 14, 2004 15:35:10 GMT
Post by Michael Messer on Nov 14, 2004 15:35:10 GMT
This may be of interest to some of you. It is copied from the John RT Davies website. John died in May 2004 and was unquestionably one of THE authorities on this subject. I visited John at his home in Buckinghamshire a couple of times & had a few long telephone conversations with him. The visits to his home were two of the most memorable evenings of my life. This man was an eccentric genius and wonderful company. Although he used computer technology during the last ten years or so to improve the sound of old records, he actually preferred to take the original 78 record and restore the disc itself. This self taught science, which has probably disappeared, was a revelation to me and other enthusiasts who visited his home. During the 40s & 50s John RT Davies worked for the BBC special effects department and according to what he told me, he taught the young George Martin the basics of recording, editing & mastering techniques. He was more of a Jazz fan & enthusiast than a blues fan, but his knowledge of early recording techniques & his skill with restoring 78 records is to my knowledge unsurpassed. John RT Davies Shine On, Michael. This forum will not accept posts with more than 10,000 characters, so I have broken this post into three separate posts. The following two posts are by John RT Davies and discuss the subjects of restoration, cleaning up & RPM of 78 records.
|
|
|
78 rpm
Nov 14, 2004 15:36:16 GMT
Post by Michael Messer on Nov 14, 2004 15:36:16 GMT
Some items of Equipment used in the repair of Gramophone Records – by John RT Davies.
An aluminum plate c. 13 ½" diameter, ½" thick machined flat (to tolerance of 0.001") endowed with a removable center spindle. The spindle is drilled and tapped 2BA as is the plate in two circles at 5 ¼" and 6 ¼" radius as well as two groups at 4" radius. The purpose of these tapped holes is the reception of screw held clamps, bridges etc. for the securing and manipulation of parts of records.
Sundry simple clamps fashioned from aluminum angle, three thumb-screw-adjustable horizontal cramps, also a collection of strangely shaped "outrigger" clamps for applying vertical pressure from tapped holes. A 1 ½" diameter (1/4" thick) piece which fits over the center spindle and is endowed with four tapped holes to receive screws holding aforementioned strangely shaped "outrigger" clamps.
A couple of pairs of "Spencer-Wells" forceps. One will carry a used cutting stylus for creating pilot grooves to bridge erstwhile digs and gouges or even across a newly rebuilt bite - hold the tool, move the work ; the other normally holds a soft-tone steel needle for gentle crut-removal and planishing upper reaches of newly cut grooves. In cases of loss of groove wall between two turns, such planishing can approximate the recorded information on the surviving groove wall onto the newly rebuilt wall... to some advantage.
A block jig to hold a dental burr. Such will facilitate the removal of a little material along broken edges so that pieces may be rejoined with the adhesive being contained within the cavity so created thus avoiding to a large extent the displacement which would otherwise be caused by the thickness of the adhesive.
A dissecting microscope – binocular – 20x and 40x. Greater magnification might seem desirable for microgrooves but tools for working on microgrooves are very difficult to create and even more difficult to wield.
"Chinagraph" black and white. White is easier to work with because of contrasting visibility and lower fluidity temperature. Black- more difficult to see what you’re doing – can effect a repair invisible to the naked eye. Melted in with the tip of a heated darning needle, surplus can be planed off with a slicing motion of a razor blade; the new surface can then be buffed smooth before pilot grooves are cut. Hold the tool, move the work
Edison amberol wax (more amenable than the two-minute material) will provide a fine, durable new surface for larger repairs, but (having poor adhesive property) needs to be bedded in chinagraph. The Edison wax needs to be cut lightly/repeatedly to prevent chipping; tendency to chip can be reduced by blending with black chinagraph (which seems to mix more readily/evenly than white).
Odd bits of post-1932 EMI (post-1936 English Decca) records. This material has a relatively low melting point and will actually flow without its nature being changed.
A miniature soldering iron with interchangeable bits. One bit is of the "spade" type with a groove cut along the width of the tip (this affords control and heat for insertion of pins (0.035" diameter hard brass) to stabilize cracks and to create bridges for rebuilding bites); another is a slimly tapered point useful for adding material (as in bite reconstruction) also for non-contact leveling of new (Edison) –wax surfaces – this is also useful for "welding" wax cylinders from the inside!
Razor blades. Useful for paring off surpluses, unsharpened edges will provide hard right-angled edges which perform very well in fining up of larger areas of new surfaces prior to grooving.
Bolsters. Sundry pieces of flat metal with suitable arc to support the edges of records pressed in the manner of European Pathe or turn-of-the century pressings with raised rims during pinning and other operations inhibited by non-flat edge conformations.
Jigging. The aluminum plate has, at the center of its underside a hole to receive a turntable spindle so that broken records may be assembled, adjusted and clamped .. and played in that condition; this is particularly useful in the case of glass-based discs. It is necessary, of course, to raise the pick-up arm mounting by half an inch.
Horizontal cramps. When using horizontal cramps, the insertion between cramp and record edge of a sliver of lead will spread the point of pressure and so obviate chipping, flaking and cracking.
Rebuilding a bite. Create and insert a wire bridge near rim. Fill in a base of new material with soldering iron to within about 0.015" of intended surface. Run chinagraph to all parts of the surface to provide adhesion for Edison cylinder wax to be built up to/above surface level and then trimmed and smoothed to surface level. Finally cut pilot grooves. (NB. Amberol wax works better than 2-minute.)
Record pressing materials. Certain record pressing materials – notably that of brown Perfects (14200 – 14900) resent insertion of pins, but the addition of post-1932 EMI pressing material will ameliorate.
|
|
|
78 rpm
Nov 14, 2004 15:37:28 GMT
Post by Michael Messer on Nov 14, 2004 15:37:28 GMT
2. PITCH CORRECTION
Not all "78"s are actually recorded at 78 revolutions per minute. While Edison's "diamond discs" are remarkably consistent at 80 r.p.m., other companies have been less consistent - usually by reason of unreliable power source or of ill-controlled elements such as lubrication or even weather! In early days, differences in recording speeds could be attributed to compromise between duration of performance and space available - (Pathe).
Variation in recording speeds were largely ignored in the days of "78" dubbed reissues when the pitch of the dubbing was directly related to that of the original. However, it has long been apparent that reproduction at correct pitch is essential to correct perception of the performance.
Although, in the past, musical pitches have varied (Steinway specified C517 c/s, "Old" Philharmonic pitch called for A452 and a half c/s) but, although some English Brass Bands continued to use the Old Philharmonic pitch through the first half of the last century, generally the New Philharmonic pitch at C522 to C523 or A440 has been that accepted as the norm since before the advent of sound recording and reproduction.
The most reliable clue to pitch is found in the vowel sounds of the human voice which remain a useful constant regardless of language. Even so, it is often difficult to discern exactitude within a semi-tone, and other yardsticks will be necessary to assist a closer approach.
While the classic music fraternity may feel unhappy when trying to relate the pitches employed by earlier composers to a modern equivalent, the performances during the sound recording period present less of a problem.
There is a period in "popular" music during which band leadership tended to dictate a change in keys employed. During the first quarter of the last century, leadership was dominated by the violin and consequent use of "sharp" keys whereas, with the ascendancy of leadership by brass or reed instruments after 1920, keys employed became predominantly "flat".
Comparison of apparent keys employed throughout a recording session during which recording speed seldom varied significantly will usually serve to show up unlikely keys and suggest an adjacent set. If doubt persists, this may be settled by making a judgement upon the feasibility of a given phrase being mechanically possible for the instrument concerned or, in the case of a reed instrument, the transition from one register to another.
For the most part, therefore, in recordings of musical instruments and a voice, three areas of consideration will almost certainly establish the pitch of the performance. There will, of course, be recordings involving, for instance, voice and guitar only when no reference has been made to any standard pitch as the instrument was tuned: others involving voice and piano when the pianist may or may not have the ability to accompany in unexpected keys. A solo voice recording may have no relationship to any standard pitch and may even vary during the performance in which case reference to another recording of the same voice in known pitch may be the only clue to be followed.
3. CONTENT OF SURFACE NOISE
Wouldn't it be NICE if surface noise were just surface noise - a mere interruption of listening pleasure, devoid of any informational significance? Certainly this viewpoint has been the basis for much experiment aimed at reduction of noise without consideration of the wealth of valuable peripheral information contained within it: minute detail which reveals the taking of a breath, the movement of a finger on a guitar string or the manner in which a brass player breaks from one harmonic to another or that in which a reed player uses his tongue; upper frequency reverberations which tell of environment ... the size and shape of a room... which affect the manner in which a musician plays. Even more important are the upper harmonics which identify not only instruments of differing principle but also those of similar principle (such as trumpet and cornet)... very often, such very individual information will allow separation of a "section" of trumpets from a numerically indeterminate group into a group of separately recognisable voices. Thus it is important to preserve and make use of this information - and to take a closer look at how the ear hears surface noise and what it can do with it.
Surface noise not only contains an indispensable abundance of peripheral information but can be shown to provide another extremely valuable asset in the form of an aural "anchor". A constant undercurrent of noise will readily be converted by the ear into an ignored continuity-base for the listening process. The interruption of such constant undercurrent - whether by click or silence - will immediately restore the noise element to perception. A recent American CD devoted to the Boswell Sisters offers an example in which the engineer, believing that a pause in the performance contained no information, on several occasions briefly removed ALL sound - so destroying the continuity of the otherwise constant undercurrent - and requiring aural readjustment on each occasion. Had the engineer left just half of the apparently useless noise, the result would have been the same. It was important that he maintain all the continuity merely to assure the ease of listening which he intended. Automatic Level Control devices will similarly destroy the "anchor" by "pumping".
Assuming that it has been possible to smooth the aforementioned undercurrent, more can be achieved by positioning the noise by colouration. Just as a visual artist will establish or enhance perspectives and distances by colouration (primary colours in foreground being reduced toward blue in middle distance and purple in far distance), a sound engineer can apply a very similar system so that, of choice, a curtain of in eradicable noise can be transferred from a position in front of the programme to one preferably immediately behind the principal sounds or, in some cases, even futher back where its distraction can be minimized. There will be occasions when a shift of such position my seem appropriate during a performance... but, if in doubt, leave well alone: almost any such alteration my be perceived - in which event it will be well to remember that "if you can hear what has been done, then it has been overdone".
Colouration adjustment, largely made at the top of the principal recorded range and in the area immediately above may often seem to interfere with tonal adjustment for "reality" in the same area. Experiment will show that the establishment of a character will call for establishment also of a "horizon" upon which it is possible usefully to alter the perception of a part of the tonal landscape by altering - NOT - at the obvious frequency area - but in an area representing what might be called an anti-node at the other end of the horizon balance (imagine a see-saw). Thus it could be that the demand for colouration of a segment of tonal character at, say 4000c/s may be achieved by an opposite adjustment in an area perhaps around 450c/s. A tell-tale of upper frequency loss appears in sibilant (sss) sounds which, deprived of higher frequencies, will tend to sound more nearly "ssh".
Realising that the exclusion of surface noise must bring about immense loss of peripheral information, it is necessary to seek amelioration of such noise whilst causing least damage to that information. It will be appreciated that simple or spot filtering will not answer such a requirement. At the other end of the scale, a computerised sampling of noise applied variously across several frequency bands to weed out or otherwise reduce or nullify those noises appearing in a programme - while capable of delivering dramatic noise reduction - will, if it does not actually destroy much of the peripheral information contained among the unwanted debris, cause a continuing variation among the parameters applied so that the ears are denied a stable platform on which to function.
Whether by mechanical or computer-screen methods, interferences of a benevolent nature covering no more than a couple of milliseconds will safely achieve the least destructive results in comparison with those engendered by any "blanket" system of processing. It has been shown all too often that over-driving of blanket systems can destroy peripheral information to the extent that notes played on an alto saxophone are indistinguishable from those played on a trumpet!
Apparatus which serve to curtail or replace obviously unwanted transients bring with them hazard to other transients, notably those arising from onsets and harmonic breaks peculiar to brass instruments.
Thus, any of these electronic solutions can readily destroy that which is intended to be preserved, and must be applied with extreme care on a moment-to-moment basis if a useful and seemingly non-destructive result is to be achieved. This, in turn, will demand the least processing rather than the most!
Shine On, Michael.
|
|
|
78 rpm
Nov 15, 2004 11:47:45 GMT
Post by Richard on Nov 15, 2004 11:47:45 GMT
Excellent info and long live the 78 I met John RT back in the late 60s as I recall (?) and had been introduced to him through a friend who was, and still is an avid 78 collector - he currently has about 8,000 (to my few hundred) malnly jazz 78s and really has had to have the floor strengthened to take the weight As in MMs post I have a few 78s that have been repaired by various (dubious) means. For instance, one of the common tricks to repair an edge crack was to push into the edge of the record a steel gramaphone needle with a hot soldering iron - it works! I agree with Michael that if you filter to lose the fishfrying sound, you will lose some top and really the answer is to try to find as clean a copy as possible, although sometimes a well nigh impossible task. Look at it this way, back in the days of 78s everyone just accepted the extra surface noise as the way it was.... As an example of a 78 in action so to speak, I have attached (hopefully) a 1929 version of a Lonnie Johns\Eddie Lang duet which just shows how much presence the is to a 78 - naturally, one should play this as loud a possible to get the full effect and then think what a really good copy with the pops and bangs would sound like. www.weekendblues.org/Music/Lonnie%20Johnson-Eddie%20Lang%20-%20Two%20tone%20stomp%20-%201929.mp3
|
|
|
78 rpm
Nov 15, 2004 16:53:14 GMT
Post by martin w on Nov 15, 2004 16:53:14 GMT
Great thread. Interested by Michael's comment about prefering the originals complete with snap, crackle and pop. Doing my "blues evangelist" bit a while back I converted a work colleague who much prefers the uncleaned-up sound because, "it's like looking through a window onto the past." I think he's got a point: if you clean it up too much, you do lose that sense of history which is part of the magic
Cheers,
Martin PS Michael - saw Martin Simpson last night (with Eric Bibb) - sends his regards.
|
|
|
78 rpm
Nov 21, 2004 23:50:14 GMT
Post by Alan on Nov 21, 2004 23:50:14 GMT
|
|