the one clear problem in the objectivist position

[johncarm]

Never mind whether we can tell the difference between DACs, or 320 kbps MP3 and Redbook, there remains the issue of judging between sounds that are clearly different, such as different speakers, different mic'ing technique in the recording, or even vinyl and digital.

So this post is about accuracy, specifically the view that accuracy is subjective.

I'm into recordings of acoustic classical music. So there really is an original event to compare the reproduction to. This may not matter for pure studio recordings. My whole post may not apply to any recording for which there is not an original acoustic event.

I think that no recording of classical compares to live music in a good hall, so the recording that gets closer to the live experience is more enjoyable.

Have an engineer make the most best recording they can. Now have two people judge it. They may disagree about which one is closer to live. That's fine, because accuracy is subjective.

This contrasts with the view that accuracy is something measurable.

I've talked to several musicians and conductors who believe that analog recordings sound more like live music and get closer to the sound of the instruments in a way that is useful for judging performances. Yet some scientists will disagree. The common objectivist position is to say that the musicians must like the euphonic distortion. I think this is a cop out. The musicians are not stating they prefer the sound, they are stating that the sound is more accurate.

 

[VNandor]

It is super easy to tell the difference between different DACs. In fact, I am so good at it I can do it without even listening to them.

 

[castleofargh]

First, we need to clear the confusion between perception of fidelity and fidelity. You're mostly talking about subjective opinions on what feels more correct to some listeners. That might not have anything to do with fidelity of the sound, for various reasons:
- The most obvious being that taste and opinions will inevitably have a subjective component, meaning you'll never get everybody to agree on much of anything. Finding some guys who say this or that and feel this or that, I do not believe it constitutes evidence for or against fidelity or an objective approach to it. It just means some people have that opinion.
-Another obvious reason for the rupture between fidelity and perception of it, is that typically, nobody was at the scene when an album was recorded. So instead of a reference, we use our imagination for what we think it should have sounded like. And even if someone was there, memories aren't perfect.
-Another fairly obvious aspect for fidelity is how a classical orchestra, to take your anecdote, will rarely be recorded from the conductor's position. Even if his memory of the experience was extremely accurate, chances are that many mics were used, some straight on the instruments, some overhead or whatever scheme was used for a given recording session. Do you imagine listening to a recording done with 2 mics inside the conductor's ears? After 2mn, I'd leave to go puke with all the movements of his head. But of course when you're him conducting, the experience is completely different. Sound alone cannot replicate that experience. And nobody would enjoy the playback of that sound. Case closed.
-That naturally brings us to making an album, all those mics(not in the conductor's ears), will be mixed and the result mastered. AFAIK, the sound engineers trying to reproduce the sound of the event as exactly as possible, represent only a tiny minority of sound engineers. And again, what position would they try to imitate/reconstruct? The third row, 25th sit?
Then there is the matter of our playback system. When we decide to listen to headphones, what a sonic departure it is from speakers or the actual event at a given position.
And then there is you. If the sound was identical in every way, you would still never experience the same thing you did at the opera house or whatever live concert that day. Nothing we do the second time feels exactly like the first. And of course, because beside the sound, everything else would be different. Our experience of a moment is constructed with a mix of everything we can sense and think of. To even ask subjectively if the sound feels the same, you would, at the very least, need to listen in the same place with the same other spectators around you the same orchestra moving the same way, and then the sound being added the way you consume it at home. That would be a good start at giving a fair chance to sound from a subjective point of view.

So overall:
-Fidelity has many occasions to lose first place in the list of priorities when you consider the entire process from the event to your house.
-For each issue you will imagine or rightfully identify about an objective approach to fidelity, you can probably find 10 times more for any subjective approach you thought was more reliable.

Saying that accuracy is subjective means turning accuracy into a feeling or an opinion. The different people will evaluate accuracy differently, and what value will be left to the word? If it pleases you, you could discuss your subjective opinion of accuracy, the one you think is right. But that has meaning for you, and just like any opinion, other people will tend to have their own, so we're back to square one. Subjectivity is subjective. While an objective approach values repeatability regardless of the observer(up to a point).


If you want to go apples to apples, get a pair of binaural mics, record the sound with them in your ears at a concert, or get some musician friend play something he knows by heart. Get that pressed on a vinyl and then listen to both the digital and the vinyl. The result is so very predictable that it's pointless to even try. Vinyl adds all sorts of changes, it just makes no sense to imagine more differences feel like fewer differences.
But once very many variables becomes dramatically different, who knows what may or may not help give a more coherent subjective experience, how often it happens and why? Maybe there is a tendency for vinyls to blend things together in a more subjectively coherent way that feels truer to the original, to you and some other people in a particular context? IDK I'm not inside your heads.
I grew up on vinyls and got most of my first audio love stories from that support, so at the very least there is a heavy emotional and nostalgia bonus added to vinyls when I accidentally find myself around some. I have never thought to myself that vinyls sounded closer to real sounds. Not ever. And when someone like you suggests it(you're not alone), it just confuses me. It's like when someone tells me his headphone sounds like he's right there with the band, sitting in the 5th row. To me, using a headphone without heavy DSP and head tracking has always given me an image going no further than maybe 30 or 40cm to the sides and only rarely getting anywhere past my skull at the front. I can't even agree or disagree, it's like we're living in different realities, and we are. Different subjective realities.

 

[gregorio]

I'm into recordings of acoustic classical music. So there really is an original event to compare the reproduction to.

No, there really is no original event to compare the reproduction to. Assuming it’s not a recording of a live performance, then there will be several “original events” (takes) and the best ones edited together. Additionally, there will not be one (or several) acoustic “events”, there will typically be around 20-50 (in the case of an orchestral recording for example), as each microphone will capture a different acoustic perspective/event.

I think that no recording of classical compares to live music in a good hall, so the recording that gets closer to the live experience is more enjoyable.

You’re confusing a couple of different principles here. Firstly, yes, obviously an audio recording does not compare to the live performance. An audio recording cannot capture and reproduce the sense of sitting in a concert hall, the sense of anticipation waiting for the performance to start or any of the various other sights and experiences of a live performance. Secondly, there was no “live experience” to “get closer to”, there were several takes in a studio or concert hall with no one (other than the musicians and conductor) there to experience it and even in the case of a live performance, who’s experience should the recording attempt to “get closer to”? An audience member in the front row of the auditorium, someone in the middle of the auditorium, the person looking at and focusing on the violins at a particular moment in time or the person focusing on the trumpets or woodwind?

Unfortunately, you’re repeating an ancient audiophile myth. The engineers/producer are not trying to replicate a live experience because that is impossible anyway, we’re trying to create a sort of new, idealised experience, essentially a better audio experience than anyone could have experienced.

Have an engineer make the most best recording they can. Now have two people judge it. They may disagree about which one is closer to live. That's fine, because accuracy is subjective.

The engineers do make the best recording they can but “best” does not mean closer to any one particular experience, it means to use a number of well chosen, well positioned microphones with mic pre-amps adjusted to an appropriate level, in order to facilitate producing the envisioned mix. There will be at least two people who judge the mix, the recording and mix engineer/s, the producer, typically one or more of the musicians and/or the conductor, the mastering engineer and probably the A&R or other representative of the record label. However, absolutely no one is judging whether the recording is “closer to live”, no one cares if it’s closer to live, the only thing anyone cares about is whether it’s a subjectively good recording! And incidentally, it’s never going to sound the same as the live performance to any of the musicians. A violin for example is deliberately never going to sound the same on a recording as it does to the violinist physically holding/playing that violin and hearing it from just a few inches away and, the musician would virtually never want it to, they want it to sound subjectively as good as possible from an audience perspective, not their own!

Your assertion about “accuracy” is therefore invalid/incorrect and in the case of rock and other “popular” genres, even more invalid/incorrect, because “closer to live”, the “original event”, is pretty much the very last thing anyone wants!

This contrasts with the view that accuracy is something measurable.

There is no (sane) view that accuracy to an experience of a live performance is something measurable. If you’re thinking of “fidelity” then that is something entirely different, which is both measurable wholly objective. “Fidelity refers to the correspondence of the output signal to the input signal” - Wikipedia. EG. How accurately an audio playback reproduces the recording.

I've talked to several musicians and conductors who believe that analog recordings sound more like live music and get closer to the sound of the instruments in a way that is useful for judging performances. Yet some scientists will disagree. The common objectivist position is to say that the musicians must like the euphonic distortion. I think this is a cop out. The musicians are not stating they prefer the sound, they are stating that the sound is more accurate.

Yep, classical musicians and conductors typically have little/no more understanding of recording technology than most consumers. The facts, the scientists and engineers, dictate the fidelity of digital is higher than analogue. There is no rational argument with this fact. What you are claiming maybe true of some musicians/conductors but they are no more likely to be right than consumers. And to re-iterate, the sound a musician hears when performing their instrument is not the sound a member of an audience will hear and not the sound a musician wants accurately reproduced!

G

Edit: Just seen castleofargh’s post, who’s effectively stating the same thing.

 

[johncarm]

@castleofargh You're relying a lot on arguments from imaging or positioning.

(Just saw @gregorio .. you're also rely on arguments from positioning.)

Let's say you recognize the voice of a family member. You would recognize them from multiple distances, multiple rooms, outdoors, in an anechoic chamber, etc. In other words, your brain builds an abstraction or model of the sound that can pattern-match to different scenarios.

The hypothetical conductor has heard the sound of instruments from many positions, not just while conducting. They probably play an instrument. They have attended concerts in the audience.

Let me mention one of my favorites aspects of classical music - and this is something I experience whether hearing live or a reproduction: musical expression as carried by the microdynamics. (MEMD we'll call it.) I'm quite familiar with this in a general sense and how good it can be in live concerts, with good players, in good acoustics. I have a model of live MEMD.

Different systems reproduce MEMD in different ways. But the most startling reproduction I've ever heard was an LP playing a baroque orchestra. I recognized the MEMD as resembling live music more than any other recording I've heard. But that goes for most analog (whether tape or vinyl). Of course both analog and digital vary in quality and can be good or terrible, but in general moderate analog outpaces digital in reproducing the MEMD.

When I speak of accuracy, I'm talking about the musicians and recording engineers who strive to get the musical experience right in their recordings (of live acoustic events). Not imaging, but music: dynamics, phrasing, PRaT, blend of instruments, etc. There is no better word than accuracy for this group of people, because they aren't judging "what they like better" or "what sounds good."

 

[castleofargh]

@castleofargh You're relying a lot on arguments from imaging or positioning.

Because position has significant acoustic impact, and our brain makes significant use of those cues. They matter to our experience.
I don't get the idea of discussing accuracy and at the same time arguing that change is fine because we manage to interpret sounds even when they changed a lot through clever pattern recognition. I surely could recognize my mother or Schubert's "Unfinished" through a bad phone line from the 80's, but recognizable and accurate have different meanings.

 

[johncarm]

Because position has significant acoustic impact, and our brain makes significant use of those cues. They matter to our experience.
I don't get the idea of discussing accuracy and at the same time arguing that change is fine because we manage to interpret sounds even when they changed a lot through clever pattern recognition. I surely could recognize my mother or Schubert's "Unfinished" through a bad phone line from the 80's, but recognizable and accurate have different meanings.

I'm not saying "change is fine." You could probably also judge the fidelity of a family member's voice or whether they are live in the room. It's that changes in position or imaging are not crucial to the aspects of music that I and the musicians I know (some from Sheffield Lab back in the 80's) care about getting right. Read my post again and consider my statement about musical expression and microdynamics.

EDIT: to clarify, for example, the mental model is that of a "live trumpet", and what is judged is how closely the reproduction matches that mental model.

 

[2leftears]

I think that no recording of classical compares to live music in a good hall, so the recording that gets closer to the live experience is more enjoyable.

As others have already said, that's impossible anyway. That ought to be obvious the minute you are standing at the box office counter and need to pick a seat.

And even if you were referring to the live ambiance, that is also impossible to capture because that encompasses far more than just sound. I can guarantee you that if you sat at a concert with your eyes closed, your music experience would be quite different.

What I have learnt is that there simply is no good benchmark for what constitutes "closer to the live experience" when it comes to mixing for stereo or even surround sound reproduction.

What we can set a benchmark for is how accurate an individual audio component reproduces the recorded ( & mixed) sound, which is the fidelity we are talking about. "Fidelity" in terms of reproducing the "live experience" is an impossible goal (except maybe for very specialised setups designed specifically for that purpose such as an anechoic chamber, but a band playing in an anechoic chamber is not something I think anyone would consider to be a great live experience.)

To be fair, I myself used to be in the camp that thought the "live experience" could be captured better by just placing one good stereo microphone in the most desirable position (no other members in the audience ). But I have listened to what sound engineers and other experts had to say on the matter, and have seen what good-sounding "one-microphone" recordings actually look like in terms of positioning of the band members during those recording, and thus I have learnt some valuable lessons. Reproducing "the life experience" through proper recoding techniques and audio equipment sounds like the holy grail (for some music), but it is simply unattainable.

I'm not saying "change is fine." You could probably also judge the fidelity of a family member's voice or whether they are live in the room. It's that changes in position or imaging are not crucial to the aspects of music that I and the musicians I know (some from Sheffield Lab back in the 80's) care about getting right. Read my post again and consider my statement about musical expression and microdynamics.

EDIT: to clarify, for example, the mental model is that of a "live trumpet", and what is judged is how closely the reproduction matches that mental model.

You can recognise you family member's voice through a 60's landline telephone (better than through a modern cellphone connection, ironically).

If an audio component messes up the sound to a degree that it clearly differs from your mental model of what it should sound like, than those are distortions from the ideal sound reproduction that should be easily measurable and (to some extent) quantifiable.

 

[gregorio]

Let's say you recognize the voice of a family member. You would recognize them from multiple distances, multiple rooms, outdoors, in an anechoic chamber, etc. In other words, your brain builds an abstraction or model of the sound that can pattern-match to different scenarios.
The hypothetical conductor has heard the sound of instruments from many positions, not just while conducting. They probably play an instrument. They have attended concerts in the audience.

I don’t see how any of this is relevant unless you’re claiming that you can’t recognise a digital recording of a trumpet as a trumpet but could if it were an analogue recording. If you’re talking about recordings, the conductor obviously has not “heard the sound of the instruments from many positions” for any particular event/recording, they obviously cannot be in many positions at the same time, they do not have numerous ears which are spread throughout the orchestra and recording venue. They’ve certainly got a good idea of the sound of the various instruments in general but not of the specific details of particular events “from many positions”, almost without exception it will be from the position of the conductor’s podium and of course it will always be from just one position at a time.

Just saw @gregorio .. you're also rely on arguments from positioning.

I’m not sure exactly what you mean but in a literal sense, yes, of course. Positioning defines levels, balance, frequency response and room acoustics, EG. Everything that is recordable!

But the most startling reproduction I've ever heard was an LP playing a baroque orchestra. I recognized the MEMD as resembling live music more than any other recording I've heard. But that goes for most analog (whether tape or vinyl).

All that tells us is about your personal perception, experience and knowledge but your personal perception, experience and knowledge do not define the actual facts.

Of course both analog and digital vary in quality and can be good or terrible, but in general moderate analog outpaces digital in reproducing the MEMD.

You are extrapolating your personal perception/experience as an actual fact, which unfortunately is contrary to the actual facts. Digital has a very significantly lower noise floor than analogue and therefore more tiny/“micro” detail can be recorded/resolved with digital.

When I speak of accuracy, I'm talking about the musicians and recording engineers who strive to get the musical experience right in their recordings (of live acoustic events).

Then you’re again “speaking of accuracy” incorrectly/invalidly. The musicians and engineer are striving for the subjectively best sounding recording they can achieve and there is no “accuracy” of an arbitrary subjective imagination of “best” that is not intended to be an accurate reproduction “of live acoustic events”.

You could probably also judge the fidelity of a family member's voice or whether they are live in the room.

You could not judge the fidelity of a family member’s voice, there is no matching of an input and output signal, didn’t you read the definition of fidelity?

It's that changes in position or imaging are not crucial to the aspects of music that I and the musicians I know (some from Sheffield Lab back in the 80's) care about getting right.

That is a false assertion. Changes in relative position cause a massive change to aspects of music. You think a violin from say 1m away sounds the same as from say the back of a concert hall, where you wouldn’t even hear any of the fine details and that not hearing any of the finer details wouldn’t affect “the aspects of the music”? Additionally, classical musicians certainly do consider position (and therefore acoustics) as crucial and Sheffield Labs were very particular indeed about the positioning of the microphones!

You don’t seem to have read my previous response and don’t appear to know much about the practicalities of being a classical musician or recording engineer and are just parroting old audiophile myths. Incidentally, the Sheffield Labs story is cautionary tale.

G

 

[johncarm]

@gregorio I did read your two posts, but I’m trying to address some misconceptions before I can address everything in everyone’s posts.

I’m not talking about every occurrence of the word “positioning”: I never said mic positioning isn’t important, or musician positioning in the hall isn’t important, or that moving to a different location doesn’t change the sound, or that a final recording with natural imaging isn’t important.

What I’m referring to are two strange ideas that tend to show up when I talk about recreating live music:

• that to recreate a live sound, you have to create the impression of being in row D, seat 15 or whatever, or even put the microphones there
• that if I get to know a trumpet from 5 feet away in room A, that I’ve learned nothing about the sound of a trumpet from 20 feet away in room B, or in a concert hall

Here’s are some qualities of music that are present in live music (to varying degrees, depending on player, room, etc.):

• Ease of macrodynamics
• High-resolution microdynamics
• Clear phrasing
• Top end extension (Walt Disney concert hall is amazing for this)
• Bass impact
• The feeling created by the rhythm (i.e. PRaT as it would show up in the recording)

None of these qualities require maintaining precise imaging to row D seat 15 to recreate (although of course a final recording with natural, not necessarily original, imaging is important).

If two recordings which are audibly different A and B are compared for fidelity to factors such as these, it’s up to the listener to determine which is more faithful. So if a conductor says that analog recordings are more useful for evaluating a performance in factors such as these, they are right: given what they listen for.

Maybe some other musician or recording engineer listens for other factors, and finds digital recordings to be more accurate relative to whatever other factors they listen for.

This is why I’m saying that fidelity to an original is subjective. For the pursuit of recreating live acoustic music, I think this is the more fundamental meaning of “fidelity” or accuracy than measurements. It’s the entire point of the affair. (Again, for those who think live acoustic, unamplified music sounds better than any recording, and so it is the standard.)

 

[gregorio]

@gregorio I did read your two posts, but I’m trying to address some misconceptions before I can address everything in everyone’s posts.

There is no evidence that you read my posts, if you did read them then you’re ignoring them. And, you are apparently “addressing some misconceptions” by promoting misconceptions rather than correcting them. So, I’m unfortunately going to have to re-phrase but effectively repeat what I’ve already posted.

What I’m referring to are two strange ideas that tend to show up when I talk about recreating live music:

• that to recreate a live sound, you have to create the impression of being in row D, seat 15 or whatever, or even put the microphones there
• that if I get to know a trumpet from 5 feet away in room A, that I’ve learned nothing about the sound of a trumpet from 20 feet away in room B, or in a concert hall

First point: A. You keep repeating “recreate a live sound” when it’s been explained to you that is not what a recording does or attempts to do. If your assertion were true there would be no such thing as different “takes” and no editing, and that’s just for starters. B. Your “strange idea that tends to show up” has not shown up, YOU are the only one promoting the strange idea of “being in row D, seat 15” (for example). That would only ever be the case with a true binaural recording but I’ve already explained that a recording is an idealised concept, the completed mix could not have existed or been heard at the live event, it does NOT correspond to a specific location and we do NOT “put the microphones there”. So your point is entirely a strawman argument!

Second point: Another strawman argument! Again, no one except you has “showed up” the “strange idea” that “if you get to know a trumpet from 5ft away in room A, you learn nothing about the sound of a trumpet from 20ft away in room B”. No one is suggesting you wouldn’t learn the sound of a trumpet in general, enough to easily identify a trumpet in many other acoustic spaces. What you obviously could NOT learn from listening to the trumpet from 5 ft away in room A is the details of any other specific trumpet performance in room B, the phrasing, dynamics, micro-details, top end extension, in fact all the “qualities of music” you yourself listed. You don’t seem to realise you’ve effectively contradicted yourself!

Here’s are some qualities of music that are present in live music (to varying degrees, depending on player, room, etc.):

• Ease of macrodynamics
• High-resolution microdynamics
• Clear phrasing
• Top end extension (Walt Disney concert hall is amazing for this)
• Bass impact
• The feeling created by the rhythm (i.e. PRaT as it would show up in the recording

A. Ease of macrodynamics is a nonsense phrase, there’s no such thing, macro dynamics are not easy.
B. There are not high-resolution microdynamics at say a live orchestral concert, unless you somehow break the laws of physics. Although microdynamics is another invented term that is poorly defined.
C. Clear phrasing, yes to an extent but it depends where you are in the audience. A quiet phrase, if sitting near the back of the auditorium would not typically be clear, it may in fact be barely audible, let alone clear.
D. Top end extension. No, again, not unless you break the laws of physics. High frequencies are absorbed by air, 5ft away will always contain very significantly more HF content than in the middle of a concert hall, no matter how amazing the hall is.
E. As PRaT is another audiophile invented nonsense term, it would NOT “show up in the recording” (only potentially in the perception of it).

None of these qualities require maintaining precise imaging to row D seat 15 to recreate (although of course a final recording with natural, not necessarily original, imaging is important).

Just repeating the same falsehood does not eventually make it true. That your assertion is false is incontrovertible, unless you have some reliable evidence which demonstrates that the inverse square law is somehow not generally applicable to music and neither is HF absorption in air. Are you really claiming that air and the laws of physics know when sound is specifically a live classical music performance and suspend their proven behaviour?

If two recordings which are audibly different A and B are compared for fidelity to factors such as these, it’s up to the listener to determine which is more faithful.

Repeating your misconception does NOT “address misconceptions”, it just promotes your misconception. Two different recordings cannot be “compared for fidelity to factors such as these” because those factors are NOT the factors that fidelity compares, the factors compared are an input signal with an output signal (as ALREADY explained and cited to you). You cannot simply redefine an already well defined scientific/engineering term (fidelity) to mean something entirely different, especially in an actual science discussion subforum! Additionally, while it could be up to a listener “to determine which is more faithful”, that is a different question to “fidelity” and would be a pointless thing to try to determine anyway, because as also already explained, professional classical music recordings are not attempting to be faithful to the original event.

So if a conductor says that analog recordings are more useful for evaluating a performance in factors such as these, they are right: given what they listen for.

If a conductor (or anyone else) were to use the term “fidelity” for “evaluating a performance in factors such as these”, they would NOT be right, they would be wrong and there’s little reason why a conductor would be any more right than a consumer because in general they have little/limited knowledge of engineering/scientific terminology. Two additional points: A conductor saying “analogue recordings are more useful” is meaningless if the vast majority of other conductors say the opposite, and in general they have done for many decades, hence why classical music was the first music genre to fully adopt digital recording! And secondly (and again!), conductors, virtually without exception are after the best sounding recording, not the most faithful to the original live event.

Maybe some other musician or recording engineer listens for other factors, and finds digital recordings to be more accurate relative to whatever other factors they listen for.

The factors they are listening for is the best captured recording possible, the tonal quality, the dynamics, details, tuning/intonation, phrasing, etc. For a musician, particularly a classical musician, that would be in terms of their performance rather than the capture of it, for an engineer it’s both. And, digital recordings are provably more accurate for ALL these factors without exception, although only marginally in comparison with the very best analogue recording technology (however, that analogue technology was never available to audiophiles or other consumers).

This is why I’m saying that fidelity to an original is subjective. For the pursuit of recreating live acoustic music, I think this is the more fundamental meaning of “fidelity” or accuracy than measurements. It’s the entire point of the affair.

Again, you appear to have little/no knowledge or experience of being either a professional classical musician, conductor or engineer and therefore little/no idea of “the entire point of the affair” (which again, is NOT perfectly recreating an original live event), nor do you apparently know the meaning of the word “fidelity” despite it being explained several times and even cited to you, and, you can think whatever you like but you do not get to unilaterally redefine it’s “fundamental meaning” for anyone else. The only thing you appear to have in abundance is misconceptions based on decades old audiophile myths/marketing.

G

 

[71 dB]

Not saying these things (How to record orchestras etc.) is uninteresting, but I find the way this goes uninteresting. This is like the parents trying to explain their young daughter the unicorns in a Disney animation movie aren't real and the daughter arguing back saying mom and dad are just too stubborn to believe in unicorns.

If this was a discussion forum for philosophy, the topic of problems in the objectivist position could be interesting (and it could be interesting to read what e.g. Slavoj Žižek has to say about it). Since this is a place for sound science, the premise becomes (became) nothing more than a futile attempt to muddy the waters.

So this post is about accuracy, specifically the view that accuracy is subjective.

There's both subjective and objective accuracy.

Example of subjective accuracy: "This Finnish salmon soup tastes exactly the same as the soup I ate last time I visited Finland."
Example of objective accuracy: "These two samples measure exactly the same height within our manufacturing tolerance of 0.1 mm."

I'm into recordings of acoustic classical music.

Same here, but acoustic classical music isn't the only kind of music I am into.

So there really is an original event to compare the reproduction to.

There is an original event, but only a very limited ways to compare it to reproductions of it. Only some people had the opportunity to experience the original event. We don't have time machines that allow us to go back in time into a parallel universe where WE bought the ticket to a particular concert instead of the people that did in the past of our timeline.

This may not matter for pure studio recordings. My whole post may not apply to any recording for which there is not an original acoustic event.

Our lives are full of original acoustic events. I just heard someone slamming a door close!

I think that no recording of classical compares to live music in a good hall, so the recording that gets closer to the live experience is more enjoyable.

Frankly I don't know if the recordings I enjoy the most are closer to the live experiences (in some subjective ways?). I have no means to tell. I have been to classical music concerts only a few times and I have not liked all aspects of them. They can be too stimulating for senses for autistic people. For me listening to music at home works better: I have so much control over what happens. For neurotypical people live concerts are probably superior experiences compared to listening to music at home. Apples and oranges.

Have an engineer make the most best recording they can. Now have two people judge it. They may disagree about which one is closer to live. That's fine, because accuracy is subjective.

As I wrote above, there is both subjective and objective accuracy. The engineers uses subjective accuracy to come up with the best possible recording and audibly transparent music formats use objective accuracy to make sure what you hear is close* to what the engineer heard.

* Only close because of transducers, room acoustics etc.

This contrasts with the view that accuracy is something measurable.

Objective accuracy is measurable, subjective accuracy less so (we can ask a lot of people what they think and have an average for a statistical result.)

I've talked to several musicians and conductors who believe that analog recordings sound more like live music and get closer to the sound of the instruments in a way that is useful for judging performances. Yet some scientists will disagree. The common objectivist position is to say that the musicians must like the euphonic distortion. I think this is a cop out. The musicians are not stating they prefer the sound, they are stating that the sound is more accurate.

Why do these several musicians and conductors think analog recordings sound more like live music? Do the objective inaccuracies (compared to objectively more accurate digital recording) compensate for what is lost when we record an event? If so, what kind of objective inaccuracies are best in order to achieve highest subjective accuracy? Perhaps there are something much better than what analog recording does (the inaccuracies of analog recording come from non-ideal mechanical and electronic operation. It would be a very strange coincidence if this produced optimal set of objective inaccuracies to compensate for what is lost subjectively whenever events are being recorded).

 

[redrol]

musicians are idiots, I know because I am one.

 

[2leftears]

Since this is a place for sound science, the premise becomes (became) nothing more than a futile attempt to muddy the waters.

I agree.

musicians are idiots, I know because I am one.

I don't agree; I know musicians who aren't idiots. I am not one of them.

 

[johncarm]

Not saying these things (How to record orchestras etc.) is uninteresting, but I find the way this goes uninteresting. This is like the parents trying to explain their young daughter the unicorns in a Disney animation movie aren't real and the daughter arguing back saying mom and dad are just too stubborn to believe in unicorns.

If this was a discussion forum for philosophy, the topic of problems in the objectivist position could be interesting (and it could be interesting to read what e.g. Slavoj Žižek has to say about it). Since this is a place for sound science, the premise becomes (became) nothing more than a futile attempt to muddy the waters.

Actually, cognitive SCIENCE and the SCIENCE of consciousness are interested in researching subjective phenomena.