Real or Surreal. Do you throw accuracy out the window for "better" sound?

If your speakers are anyone of Thiels, Dunlavys, Vandersteens, etc.,  with first-order crossovers and time/phase aligned, then you cannot throw away accuracy. These speakers give an accurate reproduction of music. So, you have to throw away the speakers. If you are going to do such a thing, please contact me ahead of time, I have a dumpster near my house that will accept any one of the above speakers. But do please call me ahead of time (LOL).

The perennial question for audiophiles (assuming by "surreal" the OP meant something like "super-real"; I doubt any but the oddest of us would prefer our systems to sound "unreal," "bizarre," "freakish" or any of the other synonyms of "surreal," which originally designates visual art like that of Salvador Dali). No, you don't want reproduced music to be the acoustic equivalent of clocks melting over table tops! So: do you want your system to sound like "real instruments at a live (acoustic) concert"...or do you want MORE than that? This, I presume, is what the OP meant to ask.

Mahgister always brings us back to the fact that perception of sound is a complicated thing involving a lot more than the straightforwardly measurable (that is, the "undistorted"). If that were not true, we'd all like the same equipment—and the same music, too, probably. If measurable accuracy were the gold standard, everyone would prefer solid state to tubes; tubes add distortion. But we often like the "right" kinds of distortion; don't be misled by the seemingly pejorative character of the word. 

Having said that, I'll admit to being fond of various kinds of "distortion," despite the fact that I play cello and guitar, my wife plays piano, my daughter violin, and we all play here in our home in the same environment where I also listen to piano, violin and cello as recorded music on my system. Yes, getting timbre right is important. Yes, conveying the scale of the instrument, and its position in space in relation to other instruments in an ensemble—all that is important. But finally, a kind of "super-realism" is often desirable in reproduced sound. Perhaps it compensates somehow for the displacement effect created by the domestic space, which inevitably reminds the brain that it is not actually listening to live music.

Here's a possible analogy to make this point. I used to be a photographer, back in the pre-digital days. I've won awards at juried shows, had my photographs published, etc. I knew what I was doing with a camera. Now, however, I find that I rarely can resist using one or another post-production retouching program for my digital images. I can not only correct for an out-of-true horizon, or crop the image easily; I can actually enhance the color contrast in ways that make the image "pop." Whether you know it or not, most, if not all, published images have been manipulated in such ways. Is that "realistic," "true" to the "original"? Strictly speaking: No. But we often like it better. There's nothing wrong with that. A photograph of, say, an Alpine vista simply cannot capture all the features that make the "original experience" so compelling: the freshness of the air, the sense of grandeur that comes with the sheer physical scale, and so on. So tweaking the photo a bit may trick the brain into supplying some of that missing visceral excitement. The photo is a simulacrum, not a substitute. So with reproduced music.

Despite all this, I agree with tvrgeek about the relative importance of the different elements in the audio chain, no matter what final effect one is striving for: "In order:  Source ( fixed, stuck with it). Room (we can do [adjust] within limits). Speakers (pay to play [not sure what this means here]). Electronics (small differences, even ss to tube is small in relation). Tweaks (tiny tiny tiny)."

I've got two systems, both built over many years, both excellent (to my ears), both in acoustically sympathetic rooms. One is "more accurate": I've had PSB Synchrony Ones in there, which measure extremely well; then Von Schweikerts, which sounded a little "better"; now Magneplanar 1.6 QRs, which are the "best" yet. That's my second system. My favorite rig has speakers you've probably never heard of (Scientific Fidelity "Teslas"), which were very badly reviewed by Stereophile when they were made in the 1990s, which pretty much killed them on the market. So be it. They create a more compelling simulacrum of piano, violin, cello—to my ears, which hear these same real instruments in this same acoustic environment daily. They also are more exciting for jazz and rock: better imaging (more than "realistic"), more bass punch, etc. Are they more "accurate"? Draw your own conclusions....

@mahgister 
The original recording is an acoustic perspective or take resulting from trade-off choices  conveyed by the gear system to another acoustic perspective ,where they will be translated  in our room  for specific ears/head...

@mteetank 
The subjectivity and variables in recorded music is inherent in the entire creation process from the capture to the audio engineer’s bias to the equipment used to listen to it to the ears and brain of the listener. 

I'm with these folks.

The word "real" should be abolished from everyone's lexicon. 

It does no work, sheds no light, though it does permit a lot of flatus vocis*

* [A mere name, word, or sound without a corresponding objective reality —used by the nominalists of universals. -- Merriam-Webster's]

"Real" as "flatus vocis"; me like! But then, to some degree, ALL language is flatus vocis, since words are not the things they stand in for. On the other hand, Webster says that words should not be, but "correspond" to, some "objective reality" in order to be more than (or other than) mere flatus vocis. If that's the standard, then the word "real" used in the audio evaluation context surely is NOT a mere flatus vocis. The "real" here which is to serve as the "corresponding objective reality" is the sound of a piano, violin, cello, guitar, voice produced not by one's audio system, but by the things themselves.

Unless, of course, you want to insist (as Mahgister always does) that "perception" is not "reality." Granted; I'm a Kantian, too. But then, we're back to my first disclaimer here: that ALL language is flatus vocis! 

The point: if we can't use the word "real" in discussing a "reproduction" of something—if it is a "mere sound without a corresponding reality"—then there is no way to evaluate the "re" in "reproduction," and we must just shut up altogether. "Wovon man nicht sprechen kann, darüber muss man schweigen."

i am not nominalist...

Perception is not the capture of a non transformed nor participated reality outside an isolated subjectivity...

We dont perceive the world through a head set...

We "resonate" and "synchronize" with the phenomenon we focus on and we perceive them with some filtering through our specific species evolutive and personnally acquired filters...The fil;ters are not arbitrary either...

Words exist on two levels : conscious and unconscious... Each word exist as a Poetical metaphorical MUSICAL deep grounding on some layer in the body/world relation and history and on a prosaic conscious habit and conditioned surface manifesting level...The iceberg peak... Any meaning is the result and the cause at the same time of this partially unobserved dynamic...

Then when you use a word we condition our mind in some perspectival take on the phenomena but this conditioning is unconscious but never completely arbitrary ...

UNIVERSALLY ALL WORDS are born from a metaphor...Because the prosaic level is completely derived from the basic poetical-musical-metaphorical deepest level...

When Saussure established the belief in the COMPLETE and ABSOLUTE arbitrary of signs by commodity for his analysis he was wrong and his disciples make a dogma of this isolation tool principle ...

But going there and explain it will be too long here...

The choice of words matter...Yes...

When we say a sound is REAL , the only objective meaning come from experimental acoustics and psycho-acoustics ...

Timbre, transients, dynamic, tonality, etc all concepts are scientifically described in controlled acoustic settings as a laboratory ...

But when we dont know these concepts by experience , we used isolated conventional metaphors and we use the words as metaphors to translate our subjective experience with the gear disconnected from acoustics conditions and knowledge in many case...

Then using the word "real" to qualify our system experience has no clear meaning for someone else...This metaphor is then disconnected from the necessary acoustic concepts necessary to assign meaning to it... We say often our system experience is "real" because we are unable to relate our experience with acoustic concept to substantiate it ... It is not false because it is not even wrong ... We must live always on two language levels but we must distinguish them but not negate one at the exclusive profit of the other...Creative non conventional metaphor revived our dead prose but must be inspired by real experience to do so...

 

Here is my philosophical belief as an aside :

 

I dont believe philosophically in a world completely separated from consciousness ... I am a Perceian not a Saussurian and i am a disciple of the french linguist Gustave Guillaume whose works even if very different different from Peirce semiotics perspective, anyway as Peirce himself negate an absolute separation between the sign and the meaning, as between the soul and the Body separated in such a way in the Cartesian frame ...

Here what Peirce think :

«Peirce understood nominalism in the broad anti-realist sense usually attributed to William of Ockham, as the view that reality consists exclusively of concrete particulars and that universality and generality have to do only with names and their significations. This view relegates properties, abstract entities, kinds, relations, laws of nature, and so on, to a conceptual existence at most. Peirce believed nominalism (including what he referred to as "the daughters of nominalism": sensationalism, phenomenalism, individualism, and materialism) to be seriously flawed and a great threat to the advancement of science and civilization. His alternative was a nuanced realism that distinguished reality from existence and that could admit general and abstract entities as reals without attributing to them direct (efficient) causal powers. Peirce held that these non-existent reals could influence the course of events by means of final causation (conceived somewhat after Aristotle’s conception),[1]and that to banish them from ontology, as nominalists require, is virtually to eliminate the ground for scientific prediction as well as to underwrite a skeptical ethos unsupportive of moral agency.»

https://ndpr.nd.edu/reviews/peirce-and-the-threat-of-nominalism/

In a word, qualitative experience and subjective experience are not completely arbitrary meaningless isolated experience from a reality itself isolated from consciousness ... Husserl go deeper here after Goethe...And Peirce is philosophically right...It is the greatest American thinker...With Whitehead who is British anyway...