Whether our ears are ready for absolutely clean sound, our recording transducers are not, and the only way to dodge microphones is with synths really synthing, not playing back the samples derived from the flawed outputs of microphones. There's no "perfect" possible, it's simply a matter of choosing which faulty simulacra one prefers.