Audio reproduction is a matter of taste?

[Jud]

2 hours ago, Iving said:

I do think it would be rich, to say the least, for a non-musician to attest that a musician couldn't discern a plausible sound for instruments. Someone on this Forum has a great reputation for this. As @Jud reminded us not an eon ago, @bluesman was able to discern the size of a Grand Piano from a recording. Something like that. I bet he didn't need an absolute reference to do it. Just lots and lots of experience playing in bands/with others.

 

And yet, in another informal little blind test here at AS, when presented with three musical excerpts that had different levels of simulated jitter and asked to pick the best, he, as did I, picked the one with the highest simulated jitter.

 

The recording on which @bluesman  identified the size of the grand piano was absolutely excellent sonically, done very carefully to present the performance as it happened, and he is a practicing professional musician with many decades' experience (as well as a medical doctor).  I'm a great one for humility: We shouldn't be so quick to transpose that sort of feat, the result of thousands of hours of experience, to situations with which we are not so familiar (as witness @bluesman's failure to identify the "sound" of jitter; I - obviously, since I made the same mistake - don't know what it sounds like either, but I bet people like @Miska, @PeterSt and @damien78 who have thousands of hours of practice can do so).

[Jud]

1 minute ago, Iving said:

 

Thank you for all the extra context Jud. I'm struggling with the link between his Grand Piano feat and the medical doctor credential! No matter whether one "dimensionalises" the case for discerning instruments, the point is that we acquire experiential references for our judgements over our lifetimes, and that we don't need an absolute or exact reference to make a good judgement at a later time.

 

Isn't the jitter case a different one. The way you describe it - highly relevant to the Original Post. Didn't ancient discussions here relate this "taste" for jitter to the "taste" for valve amps etc. Some people like distortion. Some of those in turn may mistake it for "accuracy". Whatever that is!

 

All very good points.  I mentioned the medical credentials for two reasons: Here is a very careful, thoughtful person who can be expected to have considered the jitter question seriously.  I am assuming he, as I, picked the version that sounded most natural, though he had the benefit of vastly more general experience regarding what sounds natural in music.  Yet it didn't help in the specific context of a playback artifact he lacked great experience in identifying.  That's my point: Even people with vast general experience regarding what sounds like real instruments may lack specific familiarity with relevant aspects of the recording and playback chain.  One other interesting example is Mark Knopfler's recordings, reasonably famous for their good sound.  Yet if you read about his production process, he runs the music through various analog and digital decks to get specific colorations in the sound that he particularly likes.  And therefore it seems to me we ought to be duly humble about our abilities to discern what has accurately reproduced the natural sound of an instrument or vocals.

 

The other reason I mentioned he's a doctor and musician is that he's got to have great hands, and as a clumsy oaf I'm jealous.  😃

 

I do think your second point might help explain the result of the jitter test.  I picked the version in which the instruments sounded more "live," less "hi-fi."  Apparently this liveliness may have been due to distortion.

 

[Jud]

32 minutes ago, firedog said:

Austinpop is a very careful listener and comparer.

I would never setup a system like that, but he has spent countless hours evaluating different setups, and at the time of that graph, I assume it's what he found sounded best. 

None of us can say he's wrong with any degree of authority. We'd just be imposing our assumptions on his reality.

 

I don't get this whole discussion. It's obvious no 2 playback systems sound the same, and if both are fairly good, people are going to disagree about which one they want to listen to. None of us really knows which one is more accurate or real. We all have subconscious assumptions about how things should sound, and we base our evaluation on that. We don't all have the same assumptions, so we don't agree on evaluations of audio equipment. 

In that sense it is all taste. There is no absolute sound. Even 2 different masterings of the same recording can sound different. Which one is more accurate and real?

 

edit: by the way, I agree with using the term "convincing". My setup is the most "convincing" one I could find that fit my budget and not large listening room; and would sound "convincing" in that space. One of the things that "convinced" me to buy it was playing back large scale orchestral music like Mahler.

It clearly sounds nothing like a real orchestral performance of Mahler in a concert hall. But it sounds pretty "convincing" in my playback space. Much more than any other setup I've had or could reasonably presume to have in that space. 

I'm sure there are some that wouldn't like the sound of it or find it convincing. They have other systems. I probably wouldn't like the sound of their system in my space. 

 

Just wanted to add to this that we each may hear in quite different ways. See many of the examples on this page in which what you hear depends to a great extent on whether you are right- or left-handed:

 

https://deutsch.ucsd.edu/psychology/pages.php?i=201

 

 

[Jud]

41 minutes ago, hopkins said:

 

You are confusing things between accuracy of the recording/mastering and accuracy of the playback (to the recording/mastering). They are two different things. They obviously both contribute to the end result, but we are not talking about the recording here. We are talking about the equipment.

 

In fact, it could be argued that accuracy of the recording/mastering is irrelevant to evaluate the equipment. Why? Simply because you can compare the playback of a same recording on different equipment and evaluate the accuracy of the equipment on a set of criteria. 

 

It's all common sense. 

 

You can evaluate input versus output for various kinds of distortion, and if that's satisfactory and complete for you, great.

 

Fallible as we all are, some of us like to listen. And where that's concerned, taste rules.

[Jud]

2 minutes ago, hopkins said:

 

Wow, I'm really impressed with the strength of that argument 😉. 

 

Analogy time:

 

You can certainly check how faithfully someone follows a recipe. But whether I like the food is purely up to me.

 

Same with audio. An example I like to use is my speakers. They use linear phase crossovers, which gives them hella imaging qualities but necessitates an unavoidable "presence hump" in the midrange.

 

I love them - have had this brand 30+ years. The imaging makes the music they provide sound "real" to my ear-brain. That presence hump, on the other hand, drives my dear friend @semente crazy.

 

You can't have a speaker with both linear phase crossovers and lovely flat frequency response through the midrange (in other words, no perfect speaker). So which choice is objectively correct? Who's right, me or @semente?

 

The answer of course is that each of us has chosen a system that sounds real to us, so we're both as right as it's possible to be in this imperfect world.

 

Of course if you had concentrated on understanding rather than a snide retort you might have grasped this without the necessity of a longer explanation.

[Jud]

14 minutes ago, hopkins said:

There is a huge difference, however, between saying that there are "compromises" and saying that the different versions of sound reproduction produced by different components are all valid alternate realities, and that there is no objective accurate reproduction of a recording. 

 

At the late lamented RMAF, I attended a seminar on different forms of distortion given by Audio Precision. I learned that I can't abide slew rate limiting distortion and am particularly sensitive to it. My amp has a tremendously fast slew rate. On the other hand, the very wide bandwidth this requires can potentially cause problems with harmonic and intermodulation distortions, to which other people are more sensitive than I am.  Which of us is objectively correct?

 

In filtering used in software players and DACs, time-based distortions and frequency-based distortions are what are called "conjugate variables." Mathematically, as you reduce one the other necessarily increases. What is objectively accurate - maximum time-based distortion, maximum frequency-based distortion, or something in between?

 

There are "compromises," as you call them, throughout the playback chain. Different components make these compromises in different ways for different people's listening preferences. Please do tell us which of these is objectively "accurate," since you are the one maintaining such a thing exists. Which specific components, and what are the specs that make them objectively accurate?

[Jud]

2 hours ago, hopkins said:

Not all digital solutions implement filtering

 

There are a tiny handful that don't, because if you don't use it you get lots of intermodulation distortion.

[Jud]

5 minutes ago, The Computer Audiophile said:

That’s only if you count the playback side. I’m unaware of an A to D without filtering. 

 

If it has no filtering, it's an A to A! 😀

[Jud]

11 minutes ago, hopkins said:

My objective in looking for an accurate system is to find one that reproduces the file/cd/LP while introducing as little distortion as possible.

 

Let's start with the fact that you've just told a professional musician, who is more familiar with the sound of live instruments than you will ever be, that he has "little to contribute" to this discussion.  Humility is in order.  How do you or any of us know for certain at all times what distortion is when we hear it?  As others have said throughout the thread, of course gross distortion is evident.  But what about when we are getting closer to the edge of what is possible, when the choice is not between distortion and no distortion, but between different forms of distortion short of perfection?  I pointed out to you in an earlier comment that such choices are unavoidable in speakers, amplifiers, and DACs.  (You were in error when you disagreed about DACs, again pointing to the need for a little humility.)

 

So now let me pose to you what will hopefully be a fun little challenge that may give you a notion of what people like @bluesman, @The Computer Audiophile and I have been talking about.  I'm hoping you have a music streaming service.  If you do, please bring up the Gillian Welch album The Harrow and The Harvest.  Now no fair peeking: In the 4th track, The Way It Goes, please tell me whether the guitar in the left channel is acoustic or electric, with effects or without, and how long into the track it took you to decide on your answer.

[Jud]

Just now, hopkins said:

And if I fail I will never be able to discuss accuracy of systems? 

 

Heh, no, of course not.  It really isn't a pass/fail sort of thing. (You may understand that remark a little better after you've had a listen to the track.) The point is that when we get down to very fine distinctions, it starts to become extremely difficult to tell whether what you are hearing is a true reproduction of the live performance, something that went into the production of the album, or some very slight form of distortion in the playback system.

[Jud]

4 minutes ago, hopkins said:

 

And why would I want to do that given what I have just explained? Who cares about the model of microphone when you compare the same recording on two systems? 

 

Your insistance is really baffling. You spend your time comparing systems, equipment, looking for "realistic sound" and all of a sudden you claim it's all BS? Really? 

 

It really isn't one extreme or the other, and I think you'd probably agree.  We all try to achieve something that sounds "real" within our budgets, but we can never be certain we've achieved it - or for that matter, that a different system wouldn't sound more real to someone sensitive to different forms of distortion than ourselves.  There's a reason there are lots of different brands of audio equipment at any given price point.

[Jud]

1 minute ago, hopkins said:

But there are some things that we can do, I think to minimize distortion.

 

Yes, but (and I apologize for the repetition) - since as I've noted there are multiple unavoidable places in the system where minimizing one form of distortion increases another form, it really must come down to which forms of distortion bother you the most.  And that is where I think what we call "taste" unavoidably comes in.  This is apart from the issue of being able to identify whether quite low levels of distortion are present, whether it was something in the recording, or whether whatever instrument it is really sounds like that.

[Jud]

I'm still curious to read your reaction to the Gillian Welch track (I don't know if it's to your taste, but I like it).  I know you understand the point, but there is nothing like an actual demonstration.

[Jud]

1 minute ago, hopkins said:

At least one person (Salvatore) seems to support the same ideas - and he is certainly experienced. I was curious to know how others felt about all this. There have been some interesting points made. Sorry to those I have offended. 

 

It's a very interesting discussion, and because of the level of interest becomes a little heated at times.  But I doubt you have really offended anyone, and I hope none of us has done so to you.

 

By the way - Salvatore is experienced, but he is not uncontroversial! 🙂

[Jud]

4 minutes ago, opus101 said:

 

Yes, but if we compare two components in the signal chain its reasonable to say which is the more accurate. For example if I'm listening on YT to Schubert's Octet and one DAC gives me distinct images of the first and second violins and allows me to follow their contributions at will but a second tends to blur those together then I'd say its reasonable to claim the first is more accurate (or more convincing).

What if the producer's intent was to meld the two, but your speaker's crossover happens to separate the frequencies at which they're playing?

[Jud]

2 minutes ago, opus101 said:

 

How could the producer meld the two? I'm curious. What does 'separate the frequencies' mean and how could that impact the way perception decodes the components of the recording?

 

By "meld the two," I mean emphasize the point/counterpoint interplay and harmonies so the two instruments are heard almost as one.

 

Separate the frequencies simply means the instruments are playing very near a crossover point so that depending on what frequency each instrument is playing at, it jumps to one driver or another.

[Jud]

7 minutes ago, fas42 said:

Jumbling of the sound of the instruments, in the mix, is a giveaway that the resolution of the system is not adequate

 

In my little town we have quite a good chamber concert series. (We've had a Van Cliburn winner and other award winning artists.) I've been lucky enough to have very good seats. If you close your eyes you often cannot pinpoint the locations of the violins in a quartet. So if that were presented to me from a recording, it would be a good representation of what I have heard in live performance.

 

My speakers, like many, have the bass drivers at the bottom, midrange higher, and tweeters at the top.  Instruments or vocals at higher frequencies literally form images that are vertically higher than those at lower frequencies. This is one (artificial) way of creating a sense of vertical dimension from a stereo recording.

[Jud]

4 minutes ago, opus101 said:

 

I have visited concerts too and my experience concurs with yours. The 'instrumental separation' beloved of some reviewers isn't primarily about physical separation in the image, rather its about cognitive separation in our mind. The ability to follow the different musical strands (in ASA I think they're called 'streams') and switch focus from one to the other effortlessly. 

 

BTW - is there a missing negative in your last sentence - 'it wouldn't' ?

 

Sorry if I was unclear, I meant the sense that the physical locations of the instruments weren't separate pinpoints would match my live experience.

 

I agree that ability to follow instrumental lines is very often a reliable indicator. There are times, however, when the sense of two or more instruments or voices as one dominates. (One example is early Beatles recordings, where John and Paul's harmony vocals blended so well.)

[Jud]

Someone wrote over at ASR about the speakers I own:

 

"I don't have Vandersteen 3A sigs, but I am very familiar with them, my brother owns a pair. They can sound quite wonderful or be blah and disappointing. It all depends where you sit. They have a narrow sweet spot, and if you move slightly to the left or right of that spot, they loose that wonderfulness.

"They have 1st order crossovers (measured acoustically). Those 6 dB/octave crossover slopes mean that drivers on each side of a crossover frequency will both contribute to the sound you hear for more than an octave above and below the crossover. That can cause some additions and cancellations that change with your seating position. Both vertical and horizontal changes (in stereo) in position make a difference."

 

There are multiple pages in the owner's manual devoted to positioning of the speakers. I can tell you that both horizontal and vertical relationships between speakers and listener are important.

 

There are many, many people who justifiably feel such fussing with how one sits isn't at all what they want in a speaker (or what a high end speaker should be). On the other hand, for those willing to put up with it, magic can happen, enough so that the Vandersteen 2 series is the best selling high end speaker ever (when I last bothered to look several years ago, over 60,000 pairs IIRC).

[Jud]

4 hours ago, Rexp said:

Which produces the more realistic soundstage, the well set up 3A's or poorly set up? 

 

Hah! That's a very good question actually, and raises the issue of what's meant by "realistic."

 

Relative to other speakers on the market, the Vandersteens when properly set up produce more of a sense of a sound field or "soundstage" that contains singers and players at individual, consistent locations within it, so it is more realistic in that sense. Horizontally the soundstage is quite realistic in terms of the intent of the artist and producer, which may or may not reflect actual positions during recording.

 

Like many (but not all) other speakers, with the Vandersteens the bass drivers are low, the midrange in the middle and the tweeters on top. This produces a purely artificial soundstage height from a stereo recording, vocals coming from a higher apparent position than kick drums and bass for example. That isn't necessarily realistic in the sense of reflecting actual positions (though a kick drum and a bass guitar will likely be closer to the floor in reality than the singer's mouth), but because sounds come from different vertical positions in real life, this does instinctively feel more "real" than speakers that present everything at the same vertical level or don't have such a well defined vertical soundstage.

 

In fewer words: Horizontally, they reflect the reality the artist and producer want. Vertically, they *feel* more realistic, though they aren't actually.

[Jud]

2 hours ago, kumakuma said:

 

In my opinion, much of the contention in this thread stems from the conflation of these two concepts on the part of some. 

 

In my mind, at least, these are two very different things.

 

Accuracy refers to whether or not a system accurately reproduces what's on the recording.

 

Realism refers to whether or not a recording sounds like live music. This will be determined for the most part by the recording itself, not the equipment being used for audio reproduction.

 

Well, but here's the thing: It's a fact that in many of the components of our systems, making one aspect of the sound less distorted will make another aspect of the sound more distorted.  To my mind, this turns "accurately reproduces what's on the recording" into "reproduces what's on the recording with the forms of distortion that are less obnoxious to me."  And that's the way we go from absolute accuracy to taste.

[Jud]

26 minutes ago, Allan F said:

 

Not if the aspect of the sound that is made more distorted more accurately portrays what is actually on the recording. Then the result would be sound most likely to be perceived as more obnoxious, as is often the case with higher fidelity playback of poor recordings.

 

Let's move from the theoretical to the specific:

 

- Does better imaging or slightly flatter frequency response throughout the audio band sound more realistic to you in a speaker?

 

- Does a faster (higher slew rate) amplifier or one with more limited bandwidth and thus lower IMD sound more realistic to you?

 

- Does a DAC with lower IMD or one with less group delay sound more realistic to you?

 

Of course appended to each of these questions are two sub-questions: (1) If you know, and (2) How much of these sorts of distortion are we talking about?

[Jud]

1 hour ago, hopkins said:

 

Theoretical question: if you did hear a system with less distortions (no tradeoffs) than others you may have experienced so far, do you think you would even know ? 

 

I know there are systems I've heard that have knocked me out, one at the last RMAF I'm thinking of in particular. Was that because it had levels of distortion so low they were inaudible to me, even subconsciously, fronted by speakers good enough that I didn't miss my Vandersteens? No idea - it could have been. 🙂 That's always a question for me: Do I think it sounds great to me because of low distortion, distortion that doesn't bother me, or even euphonic distortion? Since I don't have lab grade equipment to measure, I don't know the answer.

[Jud]

18 minutes ago, fas42 said:

 

That's how it seems a lot of the time ... luckily, there is an escape clause 😁 - a standard of accuracy exists which subverts that 'rule'; very, very few people reach it, which is why this principle is often mentioned. And the really good news is that the closer one gets to true, "absolute accuracy", the more that type of rule can be tossed out the window ... 🙂.

 

Umm, not unless 2+2=5, because all I'm talking about here is math. Are there people very talented at working with the math? Sure. But no one's exempt from it.

[Jud]

Thank you so much, @hopkins, very much appreciated.  I will be busy for the next little while and back later with more to say, but for now let me clear up the mystery. (Note: I've seen David Rawlings playing this in concert, and it looked smaller to me than it appears from this picture.)