that the best you're gonna' do is a direct-to-disk recording - so at the absolute minimum you have 1 mike (that can't possibly capture a symphony) the mike amp, and the cutter head amp. So now you have the "perfect recording" except that every mike sounds different and so does every mike amp.But you can't have that recording, you get the one made from a stamper in some sort of material (like vinyl) that won't take the exact shape of the stamper, and in any case doesn't have anything like the dynamic range necessary. Then you run a needle through it that (to a very high degree of probability) isn't half as precise as the cutter, run up a millivolt or two through at least one equalizer (RIAA) and three stages of amplification (mV, line level, and amplifier). That signal travels through a couple of hundered discrete components (if you're lucky) of some clutch of ASICs to become a useful voltage signal.
Only now does the sound (really some simulation of the sound) in front of that unknown mike make it to the mike's analogue, the speaker. But the speaker isn't really the mike's analogue because instead of a single capsule it has at least (to a very high degree of probability) two drivers on each of two sides so the corespondence is one mike capsule = 4 dynamic drivers (and another bunch of signal transformng LCR components at a minimum or a bunch of A/D - processing - D/A ASICs again.
OOOOhhh how I love those Black Gate capacitors - they really let you hear how weathered and dry the the poor piccolo player's hands were last winter.