by Eric Flint
Even if the majority of people were still listening to them on acoustic gramophones, the range of what could be recorded had gone up enormously. Glenn Miller could play some mezzo pianos, rather than being stuck in forte and above. We, by using the modern electronics we have available, can use two microphones, one for the band and one for the vocal. Politicians speeches can be retained for posterity (no advance can be all positive). Radio broadcasts can be recorded during the transmission, and rebroadcast ad nauseam without having to pay the performers again, a golden age!
What will Grantville build?
What follows is one man's attempt to design a record mastering facility for Grantville. It presumes that the mechanical gramophones described above are being produced, and that a market for the classic 10" and 12" 78 rpm disc recordings exists. This design concentrates on the cutting room, since the earliest records will mainly be reissuing pre-existing recordings from up-time recordings. Our recordings of up-time CD's are going to be the best recordings we produce. They will be better in all ways than down-time reproductions of those same songs, simply because of the quality of the sound. Eventually, though, down-time composers will arrive in Grantville, eager to have their music be distributed by the new network of record stores and dealers. Even when producing records of down-time compositions, they will be recorded to tape or hard disk and then cut rather than cut live as long as our tape recorders and computers continue to function. The cutting studio will be equipped with the most reliable amps and speakers we can get. There is no reason that the up-time components of our cutting studio shouldn't function for thirty years or more.
The Record Cutting Lathe
The record cutting lathe is going to be expensive, and will take some machining, but it's a one off job (maybe two or three if this really takes off, but minor, anyway).
Start with the turntable. This is what will hold the record blank we will be cutting. It needs to be heavy, perfectly balanced and dead flat. It can be specially machined if necessary, but we should be able to find something. Perhaps a brake drum from something large? In any event, there is a 14 inch flat circular plate that will hold our record blank. It needs to be heavy so that it acts as a flywheel. As the cutting head cuts, the heavy turntable keeps the speed constant while the pressures change. Also, heaviness helps control any changes caused by variations in power reaching the turning motor.
Directly below the turntable is the bearing that supports it. We need one really good bearing, as rumble free as can be managed. We're recording in mono, so rumble problems are reduced, but the cutting head is pressing down forcefully on one side of the rig, with nothing balancing it on the other. If the bearing isn't perfectly smooth, as it grinds and rumbles, those sounds would be cut into the record. It's very important to avoid this.
To the right of the turntable and its bearing is the motor. The motor is going to be very oversized. This will ensure that it will run at a constant speed because the strain of cutting is small compared to the mass of the motor. Up-time, the motors for cutting lathes are hysteresis synchronous or feedback controlled. Although those are better, a big, powerful, simple three phase motor is more likely to be findable than anything sophisticated, and it saves us trying to design the fancy controls to go with the fancy motor we don't have.
The geeks are going to ask: How do we measure the speed of the turntable? The tuned reed technique is good for checking the frequency of the motor drive current, but short of sticking a series of fridge magnets round the rim (not, I suppose, impossible) it doesn't work for the turntable. There is another simple answer. Strobe markings round the edge combined with a light source driven by a known frequency (perhaps a tuning fork) should work as long as the lathe is not in sunlight.
The cutting head has to be held just over the blank record, and moved to the side as the record spins so that it cuts a spiral groove. The mechanism holding the cutting head must be very rigid and accurate. What we do is suspend the cutting head from an arm, and run a threaded rod through the carrier. The carrier is machined to engage with the threaded rod, and an overhead bar improves rigidity and carries the flexible cabling and cooling fluid to the cutter head. A vertical shaft and gears carries the rotation of the motor up to the rod so that everything moves together. All this stuff is heavy, carefully machined and fussy, but as we said, we only have to make one.
Everything is open for ease of cleaning and maintenance, rather than enclosed for protection, which means that if the operator has long hair, he or she wears a hair net. There will be no baggy clothes, nothing that can get caught in the mechanics. This thing will fail any health and safety inspection, so operators have to be careful. We don't want any scalps clogging up this mechanism.
The fanciest bit is going to be the cutting head and drive electronics. The easiest thing to modify into a cutting head we have is the speaker assembly from an old telephone handset. It has a moving iron driver that is simple to attach a needle to. The telephone handset has some problems; it's somewhat low power for the purpose, has higher impedance than would have been optimal, and it doesn't heat the cutter as much as I would like, while overheating itself too easily, finally because we've (oh, all right—I've) selected a moving iron design, very little of the heat is conducted to the cutting point where it would do some good.
Cutting heads on late generation stereo LP lathes were frequently cooled with liquid nitrogen, which gives an idea of how much heat is generated. So I'm going to damp out the cavity with some high temperature lubricating oil. I'll install a thermostat and a pump to move the oil around when the temperature goes above reasonable. Also mounted on the cutting head will be a sensor wire (for the temperature), a pair of flexible plastic pipes carrying heating/cooling oil, a vacuum cleaner tube to suck off the swarf, power wiring for the needle heater, and hopefully a microscope to check the groove (nothing too esoteric, a child's first biology set, maybe even a reading lens—I get enough magnification by taking my glasses off). None of these, nor the second lightweight arm for checking cuttings are marked on the diagram, to avoid it becoming too muddled—a photograph showing the final setup would be virtually incomprehensible.
One of the hardest things for a recording engineer cutting records is to keep the signal to the cutting head adjusted so that the groove doesn't cross over itself. If the needle wiggles too far, the grooves can touch, allowing the record to skip. Limiting the dynamic range of a modern digital master to what is available on our somewhat less sophisticated final medium is going to have to be manual at first. The cutting engineer will be doing three or four rehearsals with current running through a meter, but the cutting head will not be in contact with the disc. The engineer will be raising and lowering the volume control to compensate for the differences in level of the original. This is a major advantage with pre-recorded material. If this were a direct recording with musicians, we would have to cut lower levels so as not to risk the groove crossing over into the adjacent one
Making Records
The cut lacquer master disc is brittle and would wear out if played much, so the next stage is to make a metal master from it. A thin finish of graphite or silver is applied to the master and used as a conductive element to allow electroplating. The electroplating is done slowly to avoid softness or sponginess. The longer you run the electroplating system, the thicker the plate. In this case plating continues to an unusual thickness, which will take hours to weeks. Then you remove the lacquer master and you have a stamp. Put the stamp on a hot iron plate and you can stamp out records like a printing press.
Now let's make records. We take our copper negative and attach it to a rigid iron plate, so it doesn't deform under pressure (These plates used to be mass produced. I once met a crate of them and thought it would require antigravity to shift it) Since we want to hot press, we mount a good thermostatic hot plate and a thick copper plate to get an even distribution of the heat with a short spike mounted in the middle to keep everything centered
I propose that we m
ake our records single sided on a turned hardwood disc. The disc should be about a quarter inch thick. I suggest the hardwood, first, because wood is easily and rapidly machined and quality woodworkers are easier to find than, say, plastics people. Second, the wood is tougher weight for weight than anything else we're likely to find and transports well. I suggest that the hardwood disc will be less expensive than many other materials. If the hardwood is well seasoned and coated with the same lacquer we're using to press to, the discs shouldn't warp.
So, we've got a nearly disc size lump of lacquer, preheated till it's soft, and a slightly oversized varnished wooden disc. We squeeze the lump onto the plate, and it forms into the grooves in the copper master, and bingo! We've got a record.
For testing, we need a decent playback system on hand. This will probably not be a mechanical transport—one of our standard export turntables—but a light modern pickup arm and cartridge, equipped with an oversize sapphire needle (if available—if not we screw in a metal needle) so we can play finished discs with minimum wear. Certainly the first of any series of records will be played end to end, as soon as it is sufficiently cool and before we start pressing the series. We'll want to compare it with the original recording, to make sure it sounds right. From then on, a certain percentage will be tested, just to know when the mold is wearing out and needs replacing, or the lac is going bubbly, or the lubrication needs to be more frequent or any one of a thousand other problems.
I've glossed over a hundred problems with the process I described. There will be problems of cleanliness (medieval clean room conditions), of finish, homogeneity of mixes of ingredients, labeling and packaging, of getting surfaces into decent thermal contact (electroplating gives a rough, lumpy finish which will have to be smoothed before mounting in the press—but without deforming the grooves on the other side), of waste material that needs cutting off without either deforming or dirtying the disc. All of this needs fastidious care, but is fairly obvious. Since individual discs can carry about five to seven minutes of music, or perhaps nine or ten minutes of speech, the labeling and transport of a symphony or a Shakespeare play is a non-trivial exercise, but it's all standard problems, nothing special.
I find I have filled my allotted space with transfer of already recorded material. Recording new original material will have to come in a later article (if anyone is still interested).
Finally, I cheated in this article. I wrote it entirely from memory, as if I were there, but discovered certain gaps in my knowledge, which I filled by looking up details on the web. The following sites I found particularly interesting.
www.shellac.org/
www.recordcollectorsguild.org 1
www.recording-history.org/HTML
OTL
1857—Leon Scot in France demonstrates the Phonoautograph system for recording sounds. It uses a diaphragm flexible enough to respond to strong sound waves, attached to a fine stylus, which presses against a glass plate, cylinder or disc. In the cylinder form of the device, the glass cylinder was coated with black carbon (smoke) and rotated, recorded sound as a wavering line.
1877—In July, Edison files his first patent in Great Britain on a sound recording and reproduction of sound. A full specification for the phonograph was filed in April, 1878.
1878—The first 600 or so tin foil phonographs are made by several small machine shops at Edison's request. These were distributed to demonstrate the principle of phonography
1886—Alexander Graham Bell and an assistant, C. S. Tainter, patent important improvements on Edison's original phonograph. They call their machine the Graphophone
1889—January—Columbia Phonograph Co. begins its commercial life, based on the patents of the earlier Graphophone Co. Before 1894, Columbia and the Edison company are part of North American Phonograph, but later they split to become rivals. Both companies begin selling phonographs for use as dictation machines
1894—Emile Berliner introduces the Gramophone in the US (1889 in Europe), using a disc instead of a cylinder and a groove cut from side-to-side (laterally cut) instead of Edison's "hill and dale" (vertically cut) method. [Edison had experimented with discs in 1878, but decided not to use them, and many believe that the vertical cut method resulted in better sounding records]. The Gramophone is aimed at the entertainment market, and home versions are not capable of making recordings.
Victor Talking Machine Company formed
1906—Victor Talking Machine Company, a relatively recent entrant, offers its first "Victrola," a disc phonograph [we use the term phonograph generically from this point on] which placed the horn inside the cabinet instead of outside it
1907—An early attempt at sound amplification, the Parson's Auxetophone, used compressed air to increase the action of the diaphragm in phonograph reproduction
1912—Edison, at long last, begins offering disc type phonographs and records for sale, in recognition of the large number of discs on the market. Cylinder machines and records, however, are still produced until the demise of Edison's Entertainment Phonograph division in 1929.
1924—In October, Columbia Records experiments with "electrical" recording equipment developed by Western Electric. Electrical recording employs electronic amplifiers, microphones, and electromagnetic record cutters.
1925—Victor releases its last phonograph disc made by the original acoustic process. Thereafter, it uses electrical recorders and releases the discs under the name Orthophonic.
1926—Edison announces a long-playing, 12 inch disc capable of holding 20 minutes of music per side. This format does not become a commercial success. In 1927, the company markets its first electrically-recorded discs.
1927—Edison offers a phonograph capable of reproducing either Edison vertical cut discs or his competitors' more popular lateral cut discs.
1929—Edison introduces a series of electrically recorded, lateral cut discs, a reflection of the increasingly marginalized market for acoustic and vertically-cut records.
1929—Edison ceases production of vertically-cut records and pulls out of the home phonograph business. Thomas A. Edison, Inc. continues to be a major force in the dictation equipment business, and continues to use cylinders through 1945
1931—RCA Victor introduces its version of the long playing disc, which is also a commercial failure (see entry for 1926)
1948—Columbia Records introduces the 33 1/3 rpm Long Playing record disc
1949—RCA Victor introduces its 45 rpm disc and a special record changer on which to play them.
Circa 1950—Introduction of 16 rpm discs for books-on-disc, recorded voice releases, etc.
ROF
1631—No energy available for non-survival based projects
1632—The press could be built in a week, with minimal up-time labor required. However, with nothing to press, and no machines to play it on anyway, this would be a bit futile. The electroplating baths need to be set up as fast as possible. It would be worth getting hold of some old 78 rpm discs, and start by duplicating these, before the cutting lathe is ready, as no-one has any experience in this field (and OTL many decent cuts were spoiled by inadequate electroplating, including a couple of mine). The room in which we're going to be cutting is chosen, and structural members mounted. Probably the old Victrola machine will be stripped, tested and measured (dimensions and performance). If several machines are available, all of them undergo this treatment, and third angle diagrams are copied onto tracing paper for comparisons of different models.
1633—The lathe is built into the room, and the electronics wired and tested. First test cuttings are done, listened to and adjustments made (adjustability—bolts rather than welds, variable heights, weights, pressures are to be built in everywhere. Even if this is a production facility, it's a test bed, too—perfectibility, not ultimate design is our watchword) Prototype turntables leave the machine shops, are assembled and tested, and a price for the production run worked out. Tests are done on bubble free production of shellac, both with and withou
t carbon filler. All of this goes on in parallel, and shouldn't require more than a couple of months. It does require rather a lot of up-time help, so there might be some big gaps in the process while essential contacts are otherwise occupied. I don't know what machines will be necessary for building the turntables, but having a set of these machines built and training a team of down-timers in their use is a high priority. Castings for turntables will be needed. Also, cabinet makers must be contacted. We'll have to explain why they have to use twice as much glue as for anyone else (a rattle or buzz in one of these units is intolerable). Where is felt available? The enthusiast who started all this is by now sleeping three hours a night, eating only socially, divorced from friends and family as well as spouse and offspring, and, in future years will look back on this period with nostalgia—which says something about humanity, but I'm not sure what.
Then the dead period. Everyone's working, things are happening. The first turntable has come off the production line, the first records have been pressed, there are no immediate disasters, but nothing much is being sold because, for the time being, there's nothing much to sell. After the previous hysteria the lack of adrenaline leaves everyone suspecting they've died, and just not noticed.
