For whatever it might be worth to any of you, I registered only for one point having to do with cutter head amplifiers. This is a story that started about 1961.
The most popular high-end cutterhead at the time was a Westrex 3C cutter head, at least in the United States. They were typically used by the big studios like in the Capitol Tower usually with Scully lathes.
The Westrex 3C cutter head had two coils on each of the two armatures. One coil was the high-power "drive" coil. The second coil on the same armature was a "feedback" coil. The clever concept was to wrap dynamic feedback from the coil around to the front-end (before the RIAA equalizer section) in the cutterhead amplifier, to flatten out distortion, improve linearity. The bobbin/mandrel of the coils was made of magnesium.
That, all by itself, informs that for the most-popular cutterhead of the era, the cutterhead driver was NOT just a PA amplifier.
The problem was that the "feedback" coil in the cutterhead armature, and therefore the dynamic feedback loop modeled as a servo-loop, had a serious phase-inversion. The phase inversion caused by magnetic flux coupling from the drive coil had a high-Q resonance between about 10kHz and 15kHz (by my ancient memory at least). The normal-mode of the feedback loop was supposed to be gain-negative, like any feedback compensation, directed to the amp's front end triode. The resonant phase inversion was a sharp spike of about 10dB (also my aging memory) GAIN.
When the feedback loop flipped phase 180 degrees, it went gain-positive, distorting the drive current spectral profile to the cutter head, and therefore to the "signal on the groove". If a record where being cut high such a +4VU or worse on the peg, and something in the higher spectrum crashed out, like a cymbal, the drive current into the cutter head would take off spontaneously, often blowing up the drive coils!!!! The repair could cost more than a thousand dollars - in 1960's terms. To avoid the damage, the mastering channel engineer would be compelled to attenuate the high end, reducing the "high" of high fidelity stereo, or cut at a reduced drive level, and accordingly groove velocity. The "cut" would be over-equalized, sharp, because the 10kHz to 15kHz band would be gain-high by definition.
A clever mastering engineer by the name of Howard Holtzer had an idea. He worked for Contemporary Records on Melrose place, and he was trying to get rid of the "peak" in the high band in the mastered acetate. The idea was to setup an equalization L-C network to "invert" the phase of the "feedback coil-inverted" phase with a Q and an attenuation notch that would match (invert) the resonant profile of the cutter-head feedback loop.
Then, he took the idea further: he set up a sequential series of equalizers, each about an octave in bandwidth, so that the spectrum "match" of the drive current to the cutterhead would be specifically tuned to each head very precisely. This was a VAST improvement over the simple equalizer-circuit to meet the RIAA curve.
To introduce the idea, Howard built a prototype amplifier at about 200 Watts per channel. He talked Bill Robinson, then recording chief at Capitol, into allowing us to set up a demonstration. Because of the precise tuning required to match a cutter head, hours of setup were required. Mastering channels are expensive to take off line during normal business hours, so Bill said we could work in Capitol all night long if necessary.
Bill allow Howard and myself access to the mastering room at Capitol one night. We set up the prototype, and by the approach of the morning work-shift hours, for the FIRST TIME IN HISTORY, we could cut a master acetate disc, and play it back, with a spectral pattern response that was within 1dB of flat to the RIAA curve.
By the way, the time-constant of the RIAA curve in the time domain was about 40 microseconds, about the same as a Ampex 351 of that era. If the RIAA equalization were removed and the amplified returned to cut the lacquer "flat", the time constant would be about 25 microseconds!
So, Howard got the agreement to produce his amplifiers, initally sold to Capitol, quickly followed by RCA that called the result the "Dynagroove" process, though in my opinion RCA over-equalized the cutter too much. Capitol had a different name, now lost in my memory.
Howard formed a company around the amplifier that we prototyped together. He called it "HAECO" and it changed the state of the art.
Also, by the way, I was looking at "dynagoove" today in Wikipedia, and the article there is not correct. It says that the Dynagroove process was the first to use computers to cut records. NOT TRUE. It was a very clever implementation, all analogue, to compensate for a flawed dynamic feedback loop servo design, then extended to "tunable" EQ over the whole range.
I know.
I was there.