Setting up the gear chain for a review got me thinking about generation loss, the cumulative effects of multiple transition stages, long a part of audio production and replication in many forms. Consider vinyl. You no doubt know that the first stage in the process of creating mass-produced vinyl releases is dragging a heated rock (stylus) over the surface of an aluminum disc coated in lacquer (dried paint), bouncing the stylus up and down and wiggling it back and forth at an audio rate. You may have heard folks tell of the glorious fidelity of playing a freshly cut lacquer master. Indeed, there is no better stylus in groove experience, though each play rapidly degraded the soft material.
The process for vinyl record production is a plating of the lacquer master (yielding bumps instead of grooves), a plating of that metal master to get a metal mother (now with grooves again). The mother is successively plated to pull metal stampers (bumps again). The stampers are used to squish hot plastic into the final product. That’s one generation of transition from source to lacquer and then four more generations from initial plating to pressed vinyl, a total of five generations. Yes, the generation loss is audible.
Generation loss isn’t confined to physical media. We have some loss each time we convert energy from one form to another throughout an audio chain. Acoustic energy is transferred to mechanical energy (moving a microphone diaphragm). Mechanical energy is transferred through magnetism (dynamic mics) or by modifying a capacitive circuit (condenser mics) to an electrical state. That electrical signal may be transferred to yet another state either for storage (analog tape [magnetism again], or some digital form and media) or simply to digital to pass through a digital console, a distribution network, and so on. At some point along the way, be it a recorded experience or live, the energy conversions start again — electrical to magnetic (speaker voice coils) to mechanical (speaker cones and diaphragms) to acoustic, finally reaching the end-user’s ear.
It’s actually pretty amazing that quality can be maintained throughout such a chain. There is some inevitable loss at each transition point, and often, significant character change. A long litany of audible effects can be attributed to analog media like vinyl and tape. Their longevity and popularity are based on those effects being euphonic.
In the earliest days of digital audio, much attention was focused on the avoidance of successive stages of A-D and D-A conversion. To be sure, the initial implementations of dither and clocking and filtering were far short of what we now enjoy, and the cumulative effects of multiple conversions could be audible and detrimental. Digital done right is all but transparent to the source. Conversion has improved enough that most engineers avoid successive conversions as good engineering practice, but don’t fret too much at an extra conversion or two.
In general, energy and electrical conversions are getting continually more accurate. A few generations of technological progress and refinement are building better bridges between the generations of transitions.