The Record Collectors Guild
A website for the Record Collector.

How records are made

(6645 total words in this text)
(48006 Reads)  Printer-friendly page
Pressing Matters
The Production of Phonograph Records

In 1952, the distinguished American music/audio writer and critic, Edward Tatnall Canby, wrote a series of articles for The Saturday Review which was published in book form as "The Home Book of Recorded Music and Sound Reproduction", by Prentice - Hall, Inc. New York. As was Canby's usual style (and that of most New York critics of the day), it was pompous, superfluous and mostly talked down to the reader. It covered a lot of ground, and despite containing a number of errors, it still made for interesting reading.

I have adopted Canby's basic framework for this article, along with a lot of re-writing and correction of the obvious mistakes. I have also added much material from my own direct experience as a recording engineer for RCA Victor in the 1960's.

... Graham Newton


To every person who has read the "Technical Information" box on the jacket liner notes of an LP record, the names Scully, Presto, Neumann, Westrex, Grampian, HAECO, Pultec, Langevin, Fairchild, Altec, Capps, Transco, Audio-Devices, Soundcraft and others may be familiar, almost household names.

These were all manufacturers of equipment and supplies that touched the process of making phonograph records, from the stage of being converted from a "master" to a release pressing intended for home playback.


It all starts with the disc recorder or "lathe" as it is often referred to... a precision piece of equipment designed with a slowly rotating feed screw mechanism and carriage to uniformly move a cutting head across the radius of the disc. The accurately shaped cutting stylus, mounted in the head, cuts a VERY precise spiral groove across a flat lacquer coated aluminum disc spinning at an exact speed of 33-1/3, 45, or 78.26 revolutions per minute, the standard phonograph operating speeds. Scully, Presto and Neumann are probably the best known manufacturers of disc recording lathes.

There were instances of early recordings (mostly acoustical) that played at speeds as low as 60 rpm, and as high as 90 rpm, but by the time electrical recording replaced acoustical, standard speeds were adopted.


Installed on the lathe, the cutting head is simply a phonograph pickup in reverse, that is, feed audio in and get mechanical motion out. Other than its greater size, the specially shaped cutting stylus, and the feed screw mechanism which moves the head across the record to make the spiral groove, the internal workings are very similar. A great deal of literature has been written about the cutting stylus, its shape and the resulting groove cut by it, many written by engineers Frank and Isabel Capps, whose company produced many of the innovations in disc recording styli over the years, not the least of which was the Capps (ANM) Antinoise Modulation Stylus and the Cappscoop stylus which provided a further 3 dB reduction in noise for stereophonic records.


Contrary to what you might think, diamond IS NOT a good material for a cutting stylus, but IS excellent as a reproducing stylus. The recording stylus is one of, if not THE most important component of the recording process, and was probably first used by Edison in 1877.

Home cutting styli were often made of a steel alloy called Stellite because it was inexpensive to manufacture, and although cutting styli have been made of other materials, most cutting styli for professional use are made of corundum, commonly known as sapphire.

Because of its lack of grain, crystalline structure and cleavage, sapphire may be ground to very accurate dimensions and angles, while retaining a very fine cutting edge. These properties are of prime importance in manufacturing a recording stylus. Believe it or not, a sapphire recording stylus will outlast a diamond and produce superior recordings.

A small sapphire rod is ground with a flat face and mounted in an aluminum shank to make it easier to install and remove the stylus from the cutting head as the need arises. The end of the sapphire rod is ground to a point with a rounded tip, and extending upwards from the tip, along the flat edge, are tiny burnishing facets that polish the groove as it is cut, producing a quieter groove. These facets are critical since, if ground incorrectly, they will affect not only the signal to noise ratio but also the high frequency response of the recording. High quality recording demands that the stylus noise be typically in the range of 57 to 60 dB below a recorded 1kHz reference level of 7 centimeters.

For microgroove recordings (16-2/3, 33-1/3 and 45 rpm), the recommended cutting stylus tip radius is .00025 inch or less, and for coarse groove recordings (33-1/3 rpm transcriptions and 78 rpm) the recommended radius is .0015 inch. You can see why a 78 played with an LP stylus sounds so bad! For coarse groove recording, the stylus was often mounted with a slight mechanical bias toward the center of the record. This ensured the thread of cut lacquer material or "chip" would be thrown toward the center of the disc thus avoiding fouling the stylus tip.

In both LP and coarse groove recordings, the included angle (that's the angle formed by the groove walls) is 88 degrees plus or minus 5 degrees. The top of the groove is supposed to measure not less than 4 mils across, for a recording to be played with a 2.5 mil stylus, and in the case of LP's played with a 1 mil stylus, no less than 2 mils across at the top.

A tremendous improvement in cutting "fine groove" records occurred with the introduction of the hot stylus. This is no more than a winding of 6 or 7 turns of "Nichrome" resistance wire around the shank of the stylus, near the tip. Connected to a power supply with adjustable current and metering, is used to bring the cutting stylus to a carefully controlled even temperature. The hot stylus momentarily softens the lacquer surface of the recording disc while the groove is being cut. The lacquer thus offers less resistance to the cutting process, the cut is smoother with less distortion, and the necessity for diameter equalization is virtually eliminated, being reduced to about a 2dB difference between the outside to inside of a disc.


Over the years, since the days of wax, recording blanks have been called by various names including "instantaneous discs", lacquers", "acetates", "soft-cuts" and others. If there is a "proper" name, it is probably "lacquer" because of the fact that they are lacquer coated with a compound of cellulose nitrate, and acetate had little to do with it, although it has become a common name for a lacquer coated disc, and many professionals still use the term "acetate". Habits are hard to break.

The term "waxing," still persists to this day, even though actual wax is as out of date in disc cutting as the horse and buggy is to the automobile. In the "bad old days" of recording, there was no magnetic tape and the recordings had to be cut originally on huge thick blocks of warmed, essentially beeswax. Modified versions of the old solid block of wax were in use until not long before World War II.

With the acceptance, in the early to mid thirties, of the much more satisfactory lacquer-cutting process, wax went the way of the Dodo bird. Thus, the final product before "processing" became a single lacquered disk onto which the actual recording grooves were cut.

If you are old enough to ever have made an instantaneous recording, those one-of-a-kind records that studios, department stores, vending machines and the like made for you on the spot, you have seen a lacquer disk. It is made of metal (although the very cheap ones had a paper base), usually aluminum, coated with a thin layer of black, smooth, shiny lacquer. For professional master work, the lacquer surface must be perfectly even... just like a black mirror, and as a matter of fact, you can actually see a good reflection of yourself in a lacquer!

Different grades of "blanks" were offered by the manufacturers, distinguished by the evenness of coating and thickness of the metal itself. Simply put, all the discs of a given grade started through the manufacturing process the same way, but subsequent inspection determined the quality level that the disk met before it was shipped out of the plant... a "double face master" being the highest level sold for the highest price. A disc with a blemish on one side, but otherwise master grade would be a "single face master" and sell for less. More blemishes and defects produced lower quality levels that sold for less and less down to the level of the home recording user. Needless to say, for the making of commercial pressed records or radio transcriptions there was only one standard - the best that can be had for the purpose.


A high quality professional disc recording lathe system is very impressive, to say the least... accurately controlled to produce grooves that lie precisely even, next to each other, and such that the depth of the groove being cut will remain uniform and exactly as set, throughout the recording.

The mastering engineer controls all aspects of the overhead cutting mechanism, which extends from the outside edge, across the record, to the center pin, like a bridge over water, and from which the recording head with its stylus is lowered to the record surface. The spiraling grooves are cut by moving the head very slowly, across this bridge, usually by a type of screw feed. Think about a home phonograph playing a record, where the arm follows or "tracks" the pre-set grooves in the record. A disc recorder working upon a blank disk, has no such thing to guide it's motion, and so it must be mechanically moved by a special mechanism to cut spiral grooves. Provisions are made to allow speeding up the mechanism to make the faster spirals at the beginning and end of the record, as well as disengaging the feed-screw drive to permit making a "lock" groove at the end of the record. In earlier days, or with less automated professional equipment, the mastering engineer usually did this by turning a small hand operated crank to speed up the spiraling process.

The record grooves are very small, measuring on the order of 2 mils (thousandths of an inch) for microgroove and about 5 mils for coarse groove recordings (transcriptions and 78's). So that the mastering engineer can look at and measure this tiny area, professional disc recording systems come with attached calibrated microscopes, that can be swung over the groove while the record is being cut. Through the microscope a good mastering engineer can quickly determine if the cut is smooth and of the right depth with the grooves far enough apart so as not to interfere with each other. (Sometimes the pattern of one groove, cut too close, may distort the wall of an adjacent groove, giving a strange almost ghostly "echo" that can be heard just before the first sound or after the last sound on some records.)


The continuous thread of cut material that the stylus throws off as the groove is made is called by a few names... chip, swarf, thread and probably a few others, although "chip" was probably the most common. Most home disc recording enthusiasts ran afoul of this exasperating byproduct. On occasion, it would throw itself nicely to one side and lie flat on the disk in neat rows out of harm's way, but far more often, if not attended to, it would entangle itself on each turn with the chip cut on earlier turns. Catching under the stylus, it quickly builds a rat's nest of tangled material which ruins the cut since, while recording, you cannot lift the stylus away from the record in order to clear it without interrupting the recorded sound and the groove itself. "Chip" accumulates static electricity as it is cut from the groove, and it sticks obstinately to anything and everything in sight, but mostly to the uncut lacquer surface that the cutting head is moving towards! A dry paint brush was usually used to push the chip well out of the way of the advancing cutter head.

Chip is flammable, and can, if a quantity of it is burned, produce a surprising amount of acrid white smoke. Practical jokers working for radio stations were known to take the chip pot from the disc recording room, and set it inside the door of an announce studio just as a long newscast had started, tossing a match in the pot and let it fizzle and sputter producing volumes of smoke to fill the studio to the exasperation of the poor news reader who desperately tried to keep his composure!

The solution to the chip problem for the professional was fairly easy. A small flared pipe mounted near the inside edge of the cutting head was connected to a vacuum system. This drew in the chip, swallowed it whole, and captured it into a large bottle of water for flammability protection. The chip nevertheless can still make agonizing trouble. Many a record master has been spoiled by chip at the last minute when all else was perfect.


Once started on its spiral course, a cutting stylus cannot be stopped without ruining the disk. Everything must be pre-set and double-checked beforehand, from the visual selection and flatness of the disc blank, to the condition of the sapphire cutting stylus, chip suction and all the audio connections and settings. An appropriate size recording disc blank is chosen... usually 14" diameter for a 12" final disc size. A silent groove test cut is made, outside the diameter of the finished disc. This is examined under the microscope to check for correct groove size, and sometimes played back to ensure that the noise level is appropriately low. Sometimes discs are rejected at this stage, or even a cutting stylus might be changed, before things are ready to begin the actual recording.

The vacuum pump is started, tape players are cued to the beginning of the recording, the cutting turntable set rolling, and, when all is ready, the stylus is carefully lowered to the record surface, the fast beginning lead-in spiral is cut, the tape playback is started a turn or two after the spiral has ceased, the audio feed to the disc recorder is enabled, and the record is on its way. Depth of cut is periodically checked through the microscope, and of course the volume levels must be watched even though automatic variable pitch will protect against most instances of overcutting to adjacent grooves, there are still the odd nasty surprises that get past the tape mastering engineers. The running time must be checked at the 1/4, 1/2 and 3/4 way through, to ensure that the disc reaches the desired end point necessary for standardization of the finished product.

At the end of the record, after about two blank revolutions, the final spiral lead-out groove is made, again by speeding up the lead screw of the cutting lathe, either automatically or manually, followed by a lock groove. In earlier times, an eccentric lock groove was added on a special machine designed for this purpose. The master is now done, except that it must be visually examined very closely for flaws... once made, it is NEVER played.

For moving to the processing plant, masters are bolted by their center holes into an elaborate box with separations that keep the actual surfaces from touching; lacquer is extremely delicate and easily scratched. (Masters are conveniently stamped or marked for identification in this manner with a scratching stylus, the marks appearing on every finished disk.)


There are only minor differences between records, other than the most obvious of size and speed, from 7" 45 rpm through to the 16" 33 rpm transcriptions once used by broadcasters. The greatest differences, however, are found in the groove sizes. Coarse groove 16" radio transcriptions used a slightly narrower and more closely packed groove than the average 78 "standard", but it is played at 33 rpm. The groove used for LP's and 45's is smaller still, and much more closely packed. The problems of groove accuracy with LP's are greater in all respects, than with the old larger groove, and for a while after the debut of the LP in 1948, there were dire mutterings and groanings throughout the industry that small groove cutting was all but impossible.

After unheralded agonies of experiment and slow experience, the cutting of small grooves became less terrifying, and ultimately, no more trouble in this area than the old, and with minor adjustments to stylus size and depth of cut, any sort of record may be cut on the same equipment.

The record mastering engineers are usually found in the studio environment but the plating, pressing and manufacturing of the actual record is usually done at another location. Once the finished lacquer is in its shipping box we enter a new world, a world of more or less mass reproduction, where the object is the making of many from one.



Making a metal stamper that will actually mold records from hot shellac or plastic is essentially a multiple plating process - depositing metal upon metal in the traditional manner, by means of an electric current that transfers metal through a plating solution directly to the surface being plated. At this point record making is surrounded by tanks. Rows and rows of containers filled with poisonous-looking green and yellow and orange liquids, some steaming, some sloshing about, as objects are swished through the depths - this is the necessary scene, and by its very nature it is not likely to be housed in the company's plush front offices! Plating plants are off in far city corners, or housed in strictly industrial park areas.


If we are to make an actual metal impression of the lacquer record; we must either flow on a metal and let it harden, or plate it on cold. The first process obviously being impossible (lacquer is a soft, highly inflammable plastic!), some form of plating is the only answer. But how to plate metal onto a non-metal?

Strangely, it's easy, and a number of methods have been used. The oldest, when master records were cut into wax blocks, was to apply an extremely thin coating of graphite, a form of carbon that we know as pencil lead, which conducts electricity. To this coating, thin enough not to disturb the record groove shape unduly, one could actually plate metal, which would take on the shape of the record grooves. When enough metal had been backed up on the plating, it was an easy matter to strip it from the wax and the graphite, leaving a mold of metal.

Three ways are used to get the first microscopic layer of metal onto the lacquer. Silver spray is the newest and trickiest. Still used, also, is the silver pan bath, similar in its chemistry; it takes longer, and is not as accurate or trustworthy. The gold-sputtering process is more complex, but not necessarily better than the silver-spray, and possibly less uniform. The silver methods are more widely used. In the intriguing gold process, the master is mounted in a vacuum chamber between a gold cathode and an anode. A 3000-volt direct current creates a glow discharge, as in a vacuum tube; molecules of gold are deposited on the lacquer by secondary emission, leaving an extremely thin layer similar to the silver layer in the other process. Upon this molecule-thin layer of gold, the usual plating is done.

Lacquers are treated in more ingenious ways, to give metal coatings at first only a molecule or so thick. The most dazzling to watch involves silver nitrate and a simple spray gun. The lacquer surface is "sensitized" by being dipped into a solution of stannous chloride which is washed off in a water spray, leaving a very minute coating. Silver nitrate solution is sprayed at the disk-and, lo, the dead black grows into a perfect mirror of silver in a few seconds. Silver has been deposited in an extremely thin layer by chemical (replacement) action, the stannous chloride acting as a catalyst to promote the process. The newly silvered disk is washed (washing is almost a fetish in plating plants) and moves on to its next treatment.

From this point on, the objective is to build up a solid metal backing on the thin silver coat. This may be confusing, since this "backing" is actually being deposited on the front of the original disk. Remember, that we are making a negative mold, its surface now of silver, in direct contact with the lacquer surface and facing away from us as we look at the record. We are really looking at the silver from the rear, and we are about to add more metal to that rear in order to stiffen it. The original record will eventually be stripped away, leaving the other side of the silver, the facing-down side, as our mold.

The beginning of the build-up of this metal backing support, a layer of very fine-grained and delicate copper or sometimes nickel, is laid down slowly on the silver surface we have just made. It is fine-grained in order not to disturb the tiny groove patterns and to hold them accurately in place. After this, a more coarse grained metal can be piled on, although in some plants, another layer of fine grained "pre-plating" was added first, to be doubly sure. The coarse coating is desirably much faster, since at the slow fine-grain pre plating speed, it would take perhaps weeks to build up a strong enough layer of metal. One company used special rotating disk anodes that swish around close to the surface of the metal record and do the plating job to required thickness in a few hours. Other much slower systems use the usual immersion tanks with moving arms to swish the contents about.

Enough metal is thus put on the back of the silver surface to support it rigidly. Whereupon, with a quick blow of a special hammer and perhaps a wiggle or two of an inserting tool, the entire silver-copper mold breaks free from the lacquer, and we have a negative in metal, the back side or down-surface of the silver a mirror-image the original grooves. The lacquer is usually damaged in this separating process and cannot be used again; so the new metal negative, or "matrix", is now the only form in which the grooves exist. Being a negative, the grooves become sharp ridges with flat valleys between.


This metal negative or "matrix" can be used to stamp out actual records, and occasionally this is done for small run manufacturing requirements. Since it is now the only existing copy of the original lacquer disc, in most cases it is used to produce a "mother", in reality, a positive metal record that can actually be played for testing purposes. Playing is not this mother's destined role and instead, the mother is submitted to another plating operation much like the first, ending in another metal negative which is the desired "stamper". Since the mother, unlike the fragile lacquer original, is made of metal, it can be re-plated many times over, producing negative after negative in metal, all identical with the first. We now have a source of negative offspring stampers, each of which can press out as many as a thousand or so actual finished records. Multiple mothers can be made from the original metal matrix so multiple identical mothers can be sent to other pressing plants or countries for their production requirements, and thus huge quantities of records can be produced in a very short time if need be, all with exactly the same quality level of the release pressings in the originating country or plant.

But back, briefly, to the plating room and our original metal matrix one- of-a-kind negative, its silver surface just neatly stripped from the lacquer. The first step in preparing for the creation of the mother is to remove the silver that now contains the direct groove Impression. Remember again, the silver was next to the lacquer... it went on first and is on the bottom of the plating that was subsequently heaped on top of it! Strip the lacquer away, turn the entire plating over, and you are looking at the other side of the silver, the down-side, that was next to the original lacquer grooves. Removing the silver coating is necessary because it would quickly oxidize and corrode in the air. Fortunately it is a molecule-thin coating, and the underlying harder metal has virtually the same sharp image as the original. A swish or two of chromic acid takes the silver away faster than it was deposited by spray in the first place.

With the silver off but the groove image still nicely metalized in this negative matrix, the whole plating process is repeated to "grow" a new image, the mother. But first, before plating, a separating solution is applied to the surface, so that, though metal will plate on metal, it can be stripped off later on, otherwise the entire thing would become a useless solid metal block.


There's a lot to be done to this fourth-generation stamper (positive original; negative metal matrix; positive mother; negative stamper) before the fifth and final generation, the release pressing that you will buy, can be produced. A hard but ultra-thin chrome surface must be plated on, for wear. Then there's the little matter of that hole in the middle, a very crucial business, not to mention the stamper's rear side, which must be shaved to the right thickness to fit the presses, and the outside edge which must be trimmed to size. The rear surface is machined away in a lathe operation where a precision gouge neatly scrapes a spiral track from the outside of the back, right to the inside, shaving off all the irregularities, leaving a mirror-bright flat backside. The edges trimmed, the stamper then goes to the centering machine.

If you are mathematically minded, you may appreciate the problem involved in finding the exact center of a spiral, especially when that spiral is made of wavering, wobbly Lines. Impossible! (Don't be confused by the disk itself-the center of the disk is not what we want.) Of course, the original lacquer had its hole in the exact center. The hole was what determined the original spirals. But unfortunately that hole, spread and stretched in the plating, cannot be relied upon. One must start anew and make a new hole.

How? Simply by trial and error. As we all know, to our pain and distress, the slightest deviation from a perfect center hole is fatal to recorded music of any sort. Yet the man who makes the hole in the record does it just the way you and I might... by playing the grooves with a sort of needle and watching the way the arm wobbles back and forth. He is helped, though, by a magnifying device that reads the arm's motion on a big dial; when the dial's wobbles become reasonably uniform, he brings down a punch that cuts the hole. And frequently, with the record having slipped, it is in the wrong place! The first man I saw at this job tried one master five times, and gave up; it was off-center every time in a different direction. Every single record ever made goes through this trial-and-error search for the perfect hole, no better way having been found to do it.

As soon as the master has its hole, it loses it. Another punch knocks out a big disk a couple of inches across from the middle to fit the center of the press. But the disk's position is exactly determined by the small hole, and so the essential information is preserved, the final hole in the record to be made in the actual pressing according to it.



Wax and shellac are terms we still associate in popular language with phonograph records. But not only has wax vanished as a medium for the actual cutting of records; shellac, too, is gone in favor of the synthetic plastic products.

For that matter, no record was ever made of pure shellac. The old-style 78 rpm "shellac" record was made of a molding material that we now call a thermoplastic ("melts with heat"), in which shellac was greatly extended by assorted neutral filler materials, among them the carbon black which gives the black look to most records. Formulas for record material were and still are highly controversial and highly variable-quiet surface, hardness, resistance to breakage, and other features being more or less at odds with each other. Whatever a record material is compounded of, it must be both hard and smooth in the cold state, and capable of melting or softening to the proper degree with heat.


Pure shellac would never do. It would be as brittle as thin glass. Yes, record shellac is, indeed, the same material as the shellac we find in the familiar varnish. Shellac or "lac," its more correct name, is a natural resin, not unlike the gummy substance found on pine trees. Its source, however, is the Far East, where the female of a bug that infests certain trees of sorts unknown in the West, coats itself for protection with the gummy stuff. Ground-up insects constitute a good part of the first shellac product, but various stages of purification are represented by our orange and white shellacs, which are no more than simple alcohol solutions of lac. This same substance serves as a binder for the assorted ground-up an powdered fillers that have made up various phonograph record materials these many years, the percentage varying according to the formula.

An interesting economic situation arose due to the geographical location of the lac-producing bug and its special tree, largely within what was the British Empire before Indian independence. Britain controlled the world supply of lac, and all phonograph records made outside the Empire depended upon British exports. It is no coincidence that English shellac records have long been known as the finest made. Whether more shellac was used than in other formulae is a question. Better grades of lac (which, being a natural product, cannot be absolutely standardized to one grade) are more likely the answer, since the optimum percentage is not very high. Much of the quality of a shellac record was also due to the degree of uniformity and fineness of grain of the neutral filler or extender that was bound together by the lac.


Though millions of so-called "shellac" records may be found in the USA and Canada, a large number of them contain no natural shellac at all! Instead, the binding material is vinylite, a synthetic resin plastic made in the USA by the company that developed the first important thermosetting plastic, bakelite, in the early 1900's. Now the British must buy vinylite from the United States. The later-day American "shellac" record was a part-vinylite disk, the extender, including the traditional ultra-fine powder, carbon black, being much as it always was, the binding material vinylite instead of lac.

The plastic "unbreakable" records were also vinylite-made. Unlike natural lac, this new plastic is flexible, yet strong and hard enough to be used in its pure state as a material for record making. Pure vinylite is not cheap, and, strangely enough, a certain amount of cheap extender or fill mixed with the pure product produces a record claimed to be even better than one made with vinylite alone. Just how far the addition of other material should be carried for best results, and what materials should be used (discarded records ground into powder, for instance), are matters for complex economic calculation. The majority of top-quality LP records, shiny black in color, long-wearing, flexible, are made with a large percentage of vinylite, but significantly less than one hundred per cent. Many fine records were made of the pure product, dyed red for looks, as were many radio transcriptions - not a little of the appeal being in its attractive appearance, its window-clear transparency.

But the simple fact is that we bought records that varied from the pure vinylite type, always transparent, often colored, through dozens of formulae containing less and less plastic, more and more of the various extenders, until we reached the shellac-less "shellac" record of the popular 78 rpm standard-speed hits and a few remaining classical albums. Since that record was easily comparable to the best of the earlier true shellac disks, it is clear that in the synthetic plastic we had a material of enormous economic flexibility, adaptable to all sorts of technical and economic requirements in the record field.

The percentage of vinylite in a record is not the final measure of its value to you. The actual materials used to extend the pure plastic are of very great importance, as is the uniform fineness of grain of those materials. A small addition of carefully chosen materials, including the ultra-fine carbon black, actually improves the basic vinylite's properties for record making, and that is why most high quality plastic records are opaque black. Beyond this, while sound quality and/or wearing quality may decrease, the variations are enormous. Each brand must be judged for itself on these bases, and of course, in relation to the price charged and the value of the record's sound. At least one expensive brand of small-company records was known to use a material with very decidedly sub-standard wear properties, though the associated hiss when the records were new was barely audible.


After the half dozen or so processing steps that occur between the original recording and the final metal negative stamper disk comes the final operation, the making of the actual record.

The pressing process is surprisingly simple in operation. The kitchen waffle maker comes to mind, and indeed the pressing machine's working parts are not so very much larger than a good, solid waffle iron, though the accompanying complications are something else again.

Let's take the waffle maker as a beginning and do a mental conversion job on it. The basic structure is precisely what is needed for record making... two similar molds, both heated, mounted face to face with a hinge at the rear so that the machine opens up facing you. Enlarge these two molds to record size; provide, instead of the waffle pattern, means to hold two record stamper disks, one below and one above (fastened in by their centers and around the edges), and you have the beginnings of a record press. Waffles cook by electric heat; here we need a major substitution in the form of an elaborate channel system behind each stamper to allow first a sudden heat, using super-heated steam at three hundred degrees, then quick cooling by cold water, all of which must be controlled by the opening and shutting of appropriate valves - and automatically, since no human operator could maintain the exact desired cycle of hot and cold that produces the perfect record. With this we approach the complete, if hypothetical, conversion of waffle iron to record maker.

The simplest record to make and the most common is the solid disk, of one homogeneous material all the way through, though the more complex records, such as Columbia's old laminated disk, go through the same presses. Record and label are bonded together in the pressing. The operator of a pressing machine has beside him a hot tray, precisely like a hamburger grill in a fast food restaurant, on which is placed a dozen or so "biscuits" - rectangular blocks of material, (shellac or vinylite) about half the area and two or three times the thickness of the final record. The biscuits soften up on the hot plate until they are of the consistency of a soggy piece of fried mush, just about movable in one piece, and no more. With the press open like a waffle iron, a sandwich, first a label, then a biscuit, then another label is placed in the press, and the top lowered, waffle-wise. The automatic system then takes over unobtrusively; steam heats and flows the plastic material into every tiny groove; at the predetermined moment it is replaced by water, and the record is instantly hardened. The record is lifted out, and the next one is ready to go in.

The occasional reversal of labels on records can easily be understood, since the operator of a pressing machine must put the labels and the biscuit into the press by hand. Nothing is easier than to reach for the wrong pile first, or to put the right pile in the wrong spot in the first place. My wonder in watching these operations was aroused by the positively staggering opportunities for confusion that were present - what with hundreds of piles of labels, biscuits of various sorts, unattached stampers, loaded presses, all looking more or less alike, and easily interchangeable. Yet in almost every case, the right pair of stampers connects with the proper pile of labels and the correct biscuits to produce an acceptable record that reproduces the music it says it will. Drastic mistakes are altogether rare.


After coming out of the hot press the newly made record has its ragged edge neatly trimmed by a circular cutting device (something like a revolving can opener) and is then passed on for inspection.

If you think that perhaps record companies are less careful than they should be, witness the inspection process for yourself. Records are rejected by visual inspection and by actual playing checks. It's enough to say that a significant quantity are rejected - large bins of rejects, returns and cut-outs wait to be fed to the elaborate machinery that reduces these masses of unsatisfactory disks to chunks, and then to powder for recycling into more records. Visual inspection is done by workers who look closely at every disk that comes to them on conveyer belts, sorting out the duds. The supplementary playing-out-loud of a sample record every so often catches most faults that may have developed in a stamper before it has pressed too many bad disks. Considering the numerous possibilities for trouble in this whole mass-production operation, I'd say that consumers were reasonably well protected.


If you always assumed that LP and 45 records would have some special plant unto themselves, utterly removed from the old-fashioned processes that turned out "shellac" records, guess again. The record pressing machine is truly a versatile instrument. The required adaptation is all made quite simply (though no doubt the original research and experiment were anything but simple) with existing machinery, and thus 33s, 45s, and 78s can be fabricated side-by-side all in the same room. The same goes for the "shellac" and plastic records. Just a matter of a different type of biscuit and a different pressing cycle, different heats, and different timings. The LP pressing process is more carefully managed than the old 78's, being a somewhat slower process per record. In terms of playing time, however, the LP is far faster in the processing than its equivalent shellac forerunner. Six 78 records that's twelve stampers-to one LP with two stampers.


The record pressing department is actually a rather quiet and unobtrusive corner of any large record-making plant. The big noise and the big machinery and the big (and dirty) storage and conveyer space are parts of the pre-pressing operations that end in the biscuit. Record material is, except for the infrequent pure vinylite, a mixture of ingredients. The mixing cannot, unfortunately, be done merely by pouring materials together or dissolving them conveniently in a solvent. Grinding is a major operation here - grinding of recycled material, grinding and mixing of new raw material. Heat, huge furnaces, prodigious rollers squashing out tons of half-melted ooze, great cutter wheels that slice acres of hot ooze into convenient rolls or into the ultimate small biscuits, immense heaps of dusty bags, whole bins of smashed and crushed reject disks, huge hoppers full of black powder, enormous mills that grind tons of coarse material down to fine... these elements of a mighty manufacturing process made up the typical large-scale pre-pressing operation. The modest, hand-operated individual record presses are utterly dwarfed.

Rejected new records and assorted old ones brought in from outside are a major raw material. Ground up whole, labels and all, they were used for the insides of laminated records. But for other use the paper labels must be removed, and since this is impractical, the centers are simply stamped out in a big press, the remaining outside doughnut chopped up and then ground down to feed back into the endless production lines. Little seems to be lost.

...End of article.

Graham Newton

1998-present the Record Collectors Guild - Original material may be copied
with permission or credit and link back applied (the Terms and Conditions must be adhered to).