[lebuffon]

Old guy here. I brought a few masters into RCA in Toronto, just before dinosaurs went extinct.

At that time they used a Studer A80 (if memory serves) 1/2 track machine, modified, with an extra playback head that was placed before the head stack so it read the music on tape about 500mS before the playback head got it. The extra head sound was fed to the motor controller that controlled the speed of the cutting head feed motor that turned the screw that controlled the pitch depth of the grooves.

When the preview head sound was loud, the screw motor would slow down to make bigger grooves and then return to normal when the audio envelope was smaller.

That's how they optimized groove spacing before digital buffers. :-)

[ufocia]

Also https://en.wikipedia.org/wiki/RIAA_equalization

[gwbas1c]

I still wonder if the circuitry was analog or digital, though. That could have been an analog "computer".

[lebuffon]

In my case it was around 1975ish. There was no digital circuitry in signal paths then.

If you are asking how the thread pitch motor controller was working I have no in depth knowledge but it would likely be an analog envelope detector. (simple RC filter/integrator would suffice) That would be rectified to make a DC control voltage and the control voltage would be used to modulate the power to the motor in an inverse fashion. ( more audio signal -> less power )

No digits would be harmed in the process. :)

[lebuffon]

The RIAA equalization curve for recording and playback, was typically implemented with only resistors and capacitors. No DSP required.

I know I know. It's so hard to believe you can manipulate signals without a processor. ;-)