It is easy to imagine the discrete information of digital audio recording and playback being a bumpy version of an analogue waveform, but that is not what comes out of the process. Barring colossal failure of implementation, the waveform is perfectly reconstructed. Counterintuitive but true. Believing to the contrary is common in most circles, but ignores what actually occurs.
Vinyl, even direct to disk, has a great deal of distortion not found in digital systems. I know plenty of listeners who prefer the vinyl rendition; it sounds more pleasant to them. Same with tubes. That is a personal matter of taste, not a nod to greater accuracy or clarity. They simply like it better.
This is not a function of mysterious X-factors that are not being measured. If a human can hear it, it can be measured. Like I said, this was all figured out before most of you were born. What is new is transistors. (But not the idea; JFETs were theorized in 1925 but it took decades for the technology to build them to be developed. The team that first made a point contact transistor in 1947 was actually trying to make a JFET.) Circuit design has also grown with the new devices and applications of them. There is much new under the audio Sun, but sound waves, bits and bytes, and audio measurement were figured out before World War II. That war and the space race drove the technologies forward that could bring that understanding of audio to the practical devices we now enjoy, but don't confuse audio science with technology. Truthfully, it is appaling that some are still arguing the science here. It is akin to the Flat Earth Society. Most of its members dropped out after Sputnik, but some still cling to what they see in front of them. They see (or hear) what they see (hear), so that MUST be what's going on.
Quantum mechanics, relativity effects and the existence of tiny particles are also counterintuitive to us, but they accurately describe what is really happening. So do the nearly century old discoveries and developments that describe audio. So enough already. Sound Science is science and any debates about the truth of it were resolved long, long ago. If one wants to ponder a development that might actually matter, start a thread on PCM versus DSD. Which will ultimately yield a better implementation for our listening pleasure, thousands of multi bit words describing changes or millions of bits, each telling the recording engineer if the wave went up or down a really tiny amount? To simplify the issue, of course.
Edited by Clarkmc2 - 10/2/12 at 9:29pm