Vinyl and Digital and the AES

There is a current puzzle in audio since the introduction of CD.

While the measurements of digital are in many cases an order of magnitude better than vinyl (and audio Skeptics are fond of gleefully pointing this out), yet in the end, vinyl tends to sound more natural and more real most of the time (which subjectivist use to rub the Skeptics nose in it).

While we don't need to rehash the millions of words lost in the ongoing argument, I think it is easily summarized by three statements below:

• The audio Skeptics have made a basic assumption that digital audio reproduction is essentially perfect, and have stubbornly maintained this idea
• The audio Subjectivists have not adequately defined high resolution well enough to inform study.
• Neither Skeptics nor Subjectivists have been able to acknowledge that our understanding of sound perception is incomplete and its study and revisions are ongoing.  This is either through ignorance or stubbornness.

In a perusal of AES papers (yeah, we're members, too) a couple popped up that are pretty important to this idea.  I'm putting the link here, but usually only members will have access.  But you can at least see the thinking and direction of some lines of research:

     A conference paper on October 8, 2014 [LINK] (they don't have the same weight as a journal paper, but they do have the benefit of being peer reviewed, and are given in front of peers, too, so you have to know your stuff).  They found that some sounds cannot be transparently recorded in CD, that the differences in various filters that are used these days are small but are audible using the Double Blind Test.  But even more remarkable for these types of studies, they assert that in order to hear this, the chain has to be capable of high resolution sound reproduction.  This is remarkable to hear from the AES, since most audiophile subjectivists main encounter with these sorts of people are angry ill informed posts on forums, or in the over the top somewhat slimy Amazing Randi performances.  

     But also just a few days ago (LINK.  November 5, 2015), an article that cited the conference from a year ago amongst others openly says and asks (you can download and read this overview.  Its an easy read, and a breath of fresh air): 

Up to now, High Resolution in audio hasn’t been usefully defined, which is a pity because without a secure bridge between auditory science and audio engineering, development can be haphazard.

And

If a reproducing system is to be flawless for the human listener, then its errors should be both natural and plausible. Although we tend to define errors in terms of our measuring instruments, we need to move away from the poor proxy of escalating sample-rates and bit-depths. Perhaps an even more fundamental definition can help?

For me this is tremendously good news.  While it is somewhat exacting and painful to assemble a really high quality stereo, it feels more expensive and difficult that it ought to be to get everything just right.  And like in a camera, there is more to the solution to better pictures than more megapixels, if vinyl is capable of a more convincing and transparent playback than digital despite measurements, maybe we haven't been asking the right questions, or understanding what our measurements are telling us?