[johncarm]

Tremendous amount of information , so I can't respond directly to all of it. I'll hit the highlights.
 
My statements are generating disagreement here, but there may be points of agreement. For instance, I understand that orchestras and recording engineers already check the sound, or hire a producer to check the sound, and we probably all agree that's a good thing.
 
And, where we do disagree, there are some matters of taste. If we are talking "preferred sounds" (rather than "accurate sound") and an engineer or producer knows what they want, then who am I to contradict that?
 
But if we are going to talk about accuracy, then I am making a case that it's a musical judgment, and that it doesn't make sense to divide the phenomena of recorded sound into two categories (the "sound" and the "music").
 
I am not saying that all art or all music has the same patterns. It is not "universal" in that sense. I am just saying there are compelling reasons to make a study of art if you are a cognitive scientist.
 
I will limit my scope here to live classical music.
 
So, how much of art is "touchy-feely" and how much can be analyzed? Some concepts can be defined in way to make them amenable to investigation. I do not know how much cognitive science has done so, but I like to use these concepts as a place to evaluate the accuracy of recordings that is a little more objective than saying "it's all about what you dig, man."
 
For instance, most art seems to have a concept of foreground and background objects. I will take gregorio's caution and make clear this is not about staging.
 
If you consider visual layout (of photos, painting, or even newspapers) it seems that the visual perceptual system likes hierarchical grouping. In the layout, there is some point of major focus (or a few points) and the other objects are arranged in a subordinate way to them. Also, critically, if there are many objects, some of them are grouped and appear more as a single object when they are in your peripheral vision.
 
The artists seems to be trying to balance the idea of having richness (a lot of objects) together with a way of easily perceiving a whole and an easy way to move one's attention around.
 
I am not a cognitive scientist but this seems quite amenable to investigation. Object shapes and positions can be analyzed. We can use an eye tracking device to watch how people move their attention around.
 
So, there's something analogous in sound. The orchestra blends, yet there is some clear point of focus and clear groupings.
 
At the few recording sessions I was at, it seemed obvious that the musicians had this way of structuring the sound in mind. (It was chamber music, not an orchestra, but the idea is pretty much the same.) So I would listen to the live sound, then go into the control room and ask a question: does this music work?
 
What do I mean by "it works"? I had a sense that the musicians were balancing themselves in a way that fit the needs of the structure of the music. The question of "who is foreground", "who is blending", etc. was a musical decision that related to the form as stretched over time.
 
So given two mic positions, A and B, and let's say we are talking only about balance and blending, my question is: in which mic position does the music "work" better?
 
Now, about this idea that musical concerns are different than recording concerns.
 
On this thread there has been the statement "musicians don't know what they sound like." That raises the question, what exactly is their goal when they are practicing if they don't know what they "sound" like? The only way I can make sense of this is to say, "Musicians don't know the tonal balance or reverb field out in the audience." That may be true, but "tonal balance" and "reverb field" is only a sliver of the properties of the final sound. So a recording engineer who is great at capturing that but knows nothing of other patterns in music, is not in a good position to evaluate the recording.
 
I am not saying all engineers have this problem, or even that any engineers have this problem. I am illustrating a point.

 

[johncarm]

   
Are recording teams often in the habit of not being at all familiar with the sound of the venue they are recording? At least in classical I can think of a plenty of conductor/orchestra/engineer "teams" that produced consistently good sound, and today it seems we're going even a step further to the big orchestras starting up their own labels. Perhaps you don't like the sound of SFS+MTT, but there are plenty of other examples to be had. This of course brings up another point in that classical is one genre where they pay a guy to stand in front of all the people actually playing and make judgements about sound, and I can't imagine he completely ignores the work in the studio (not that the conductor's musical point-of-view is exactly like the audience's, of course).

 
Of course there is someone evaluating the sound. The question is what they are listening for.
 
I love MTT (as a conductor) but this one recording didn't have good differentiation between the strings and woodwinds. I can only speculate what happened.
 
Here is what I know. Mahler orchestrated his music in a certain way for specific reasons, many of which can probably be analyzed fairly well (i.e. they are not all "preference" or "touchy-feely"). One goal of orchestration is to differentiate successive sections of the composition--to make A sound one way, and the following section B sound some other way.
 
Sections A and B may be different in both pitch and orchestration. A widespread principle in both composition and performance is the goal to make (1) similar things sound similar, and (2) different things sound well differentiated. So if A and B are similar in pitch content, then the composer may want to use similar timbre to emphasize their similarity. This will contribute to the overall musical effect. On the other hand, if A and B are different in pitch content, the composer may want to mark that distinction in a clear way by using different timbre.
 
So there is some kind of musical effect going on, and timbre contributes to it. I hear MTT trying to bring this out. I can't get inside his head but I have a feeling what he is trying to do. In this particular recording, the woodwinds and the strings, however, did not sound well differentiated in color. So the desired musical effect fell flat. It just didn't happen. It didn't sound like the powerful Mahler I know.
 
If you were to say I'm talking about my personal preference, then you would be right. Actually in a very deep way all judgments of accuracy are subjective. But notice that my judgment has something to do with (1) understanding desired musical patterns in a way that is more analytical, and (2) judging the effect of the whole piece of music over time. It seems to me that if we want to stand a fighting chance of producing a recording that captures the musical intentions of the performer, we need to do something like that.

 

[spruce music]

  Tremendous amount of information , so I can't respond directly to all of it. I'll hit the highlights.
 
My statements are generating disagreement here, but there may be points of agreement. For instance, I understand that orchestras and recording engineers already check the sound, or hire a producer to check the sound, and we probably all agree that's a good thing.
 
And, where we do disagree, there are some matters of taste. If we are talking "preferred sounds" (rather than "accurate sound") and an engineer or producer knows what they want, then who am I to contradict that?
 
But if we are going to talk about accuracy, then I am making a case that it's a musical judgment, and that it doesn't make sense to divide the phenomena of recorded sound into two categories (the "sound" and the "music").
 
I am not saying that all art or all music has the same patterns. It is not "universal" in that sense. I am just saying there are compelling reasons to make a study of art if you are a cognitive scientist.
 
I will limit my scope here to live classical music.
 
So, how much of art is "touchy-feely" and how much can be analyzed? Some concepts can be defined in way to make them amenable to investigation. I do not know how much cognitive science has done so, but I like to use these concepts as a place to evaluate the accuracy of recordings that is a little more objective than saying "it's all about what you dig, man."
 
For instance, most art seems to have a concept of foreground and background objects. I will take gregorio's caution and make clear this is not about staging.
 
If you consider visual layout (of photos, painting, or even newspapers) it seems that the visual perceptual system likes hierarchical grouping. In the layout, there is some point of major focus (or a few points) and the other objects are arranged in a subordinate way to them. Also, critically, if there are many objects, some of them are grouped and appear more as a single object when they are in your peripheral vision.
 
The artists seems to be trying to balance the idea of having richness (a lot of objects) together with a way of easily perceiving a whole and an easy way to move one's attention around.
 
I am not a cognitive scientist but this seems quite amenable to investigation. Object shapes and positions can be analyzed. We can use an eye tracking device to watch how people move their attention around.
 
So, there's something analogous in sound. The orchestra blends, yet there is some clear point of focus and clear groupings.
 
At the few recording sessions I was at, it seemed obvious that the musicians had this way of structuring the sound in mind. (It was chamber music, not an orchestra, but the idea is pretty much the same.) So I would listen to the live sound, then go into the control room and ask a question: does this music work?
 
What do I mean by "it works"? I had a sense that the musicians were balancing themselves in a way that fit the needs of the structure of the music. The question of "who is foreground", "who is blending", etc. was a musical decision that related to the form as stretched over time.
 
So given two mic positions, A and B, and let's say we are talking only about balance and blending, my question is: in which mic position does the music "work" better?
 
Now, about this idea that musical concerns are different than recording concerns.
 
On this thread there has been the statement "musicians don't know what they sound like." That raises the question, what exactly is their goal when they are practicing if they don't know what they "sound" like? The only way I can make sense of this is to say, "Musicians don't know the tonal balance or reverb field out in the audience." That may be true, but "tonal balance" and "reverb field" is only a sliver of the properties of the final sound. So a recording engineer who is great at capturing that but knows nothing of other patterns in music, is not in a good position to evaluate the recording.
 
I am not saying all engineers have this problem, or even that any engineers have this problem. I am illustrating a point.

But if we are going to talk about accuracy, then I am making a case that it's a musical judgment, and that it doesn't make sense to divide the phenomena of recorded sound into two categories (the "sound" and the "music").
 
Here is a big part of your problem understanding.  Sound is the physical aspect of sound whether noise of machinery, people or musicians.  It can be measured and understood.  In a particular position with an orchestra or band it will be the same for you, me, a dog or a gerbil. We may all have different hearing abilities, but the physical sound is the same.
 
Music happens in the brain and will vary between you, me and may not occur at all to the dog or gerbil.   While I get that one mic position over another may alter the balance of instruments and that may have effects on the musical part of the recording you still have a case of it not being a property you can describe with any consistency.  Musicians can give their input as to whether the result is to their liking, with time them and a team of recording guys, conductors whatever may learn how to work together better.  In the end it is all someone's preference.  You might say recording engineer A gets Miles Davis' music and knows something about blending it better than engineer B.  You can't say engineer A knows music better has a better musical ear and can be used as some sort of reference to live.  Nor can you say one guy with more musical knowledge recording has a better chance at making a recording that compares better to reality or not.  It has become preference, it may involve hyper-reality for a desired effect, but saying one version is more real on that basis is going too far.  For some purposes processing away from accuracy can result in a better than real result.  Your version of musicality in the middle of how live and recorded sounds is getting in the way of you understanding this.  Nor would it aid one in making recordings or judging components. That is a myth and fantasy you seem unwilling to let go of. At that point it really is all about what you dig.
 
In regards to how you are trying to look at this, Hollywood movie sound is who works on that.  And they aren't so much concerned with reality as a directed possibly real sound with proper suspension of disbelief. This works with music, voice, foley sound effects or natural sounds

 

[johncarm]

  But if we are going to talk about accuracy, then I am making a case that it's a musical judgment, and that it doesn't make sense to divide the phenomena of recorded sound into two categories (the "sound" and the "music").
 
  [1] While I get that one mic position over another may alter the balance of instruments and that may have effects on the musical part of the recording you still have a case of it not being a property you can describe with any consistency. 
 
 
  [2] Your version of musicality in the middle of how live and recorded sounds is getting in the way of you understanding this.  Nor would it aid one in making recordings or judging components. That is a myth and fantasy you seem unwilling to let go of. At that point it really is all about what you dig.
 
In regards to how you are trying to look at this, Hollywood movie sound is who works on that.  And they aren't so much concerned with reality as a directed possibly real sound with proper suspension of disbelief. This works with music, voice, foley sound effects or natural sounds

 
A lot of what you are saying is not relevant to recording live classical music.
 
As far as [1], there are cognitive scientists who investigate music despite it not being a property you can describe with consistency. And they are learning something!
 
As far as [2], you say my "version" would not aid in making recordings or judging components. That sounds pretty abstract. What is the problem, specifically?

 

[Joe Bloggs]

Of course there is someone evaluating the sound. The question is what they are listening for.

I love MTT (as a conductor) but this one recording didn't have good differentiation between the strings and woodwinds. I can only speculate what happened.

Here is what I know. Mahler orchestrated his music in a certain way for specific reasons, many of which can probably be analyzed fairly well (i.e. they are not all "preference" or "touchy-feely"). One goal of orchestration is to differentiate successive sections of the composition--to make A sound one way, and the following section B sound some other way.

Sections A and B may be different in both pitch and orchestration. A widespread principle in both composition and performance is the goal to make (1) similar things sound similar, and (2) different things sound well differentiated. So if A and B are similar in pitch content, then the composer may want to use similar timbre to emphasize their similarity. This will contribute to the overall musical effect. On the other hand, if A and B are different in pitch content, the composer may want to mark that distinction in a clear way by using different timbre.

So there is some kind of musical effect going on, and timbre contributes to it. I hear MTT trying to bring this out. I can't get inside his head but I have a feeling what he is trying to do. In this particular recording, the woodwinds and the strings, however, did not sound well differentiated in color. So the desired musical effect fell flat. It just didn't happen. It didn't sound like the powerful Mahler I know.

If you were to say I'm talking about my personal preference, then you would be right. Actually in a very deep way all judgments of accuracy are subjective. But notice that my judgment has something to do with (1) understanding desired musical patterns in a way that is more analytical, and (2) judging the effect of the whole piece of music over time. It seems to me that if we want to stand a fighting chance of producing a recording that captures the musical intentions of the performer, we need to do something like that.

This is all quite whimsically depending on a series of unlikely assumptions:
1. That your interpretation of the musical score is the same as MTT's, and hence, MTT worked to segregate the woodwinds and strings in this particular soundscape (when he might just as well have been working to blend them)
2. That the performers then played out their parts as he desired
3. That, when all is said and done, that the woodwinds and strings actually sounded well separated at the positions where the mics were placed for the recording. Because from the context of your writing it's obvious that you weren't present at the performance where the recording was made, let alone sitting anywhere near the mics.

As for the rest, I have outlined a series of technical decisions that can actually affect the recorded outcome (as opposed to wishing how the outcome should be). I asked in my previous post and I'll ask again: is there anything in your theories that would allow the sound engineer to make those technical decisions better?

Suppose MTT listened to the recorded performance and quipped that "the woodwinds and strings don't sound well differentiated". THEN the SE could have done a dozen things to make the two jump out of each other, especially if the recording were redone or if the recording were a multitrack recording. Is there anything you can add to this?

 

[johncarm]

posted by gregorio
Quote:

   
1. I think you must be explaining poorly what you mean by clarity or misunderstood what I stated, if you really "mean something else". Because ...
 
2. How is separating (groups, sub-groups or individual) musicians into "foreground" and "background" not "separation"? Would you generally not want/expect greater clarity of those musical entities which are supposed to be in the foreground or put differently, a level of clarity which allows for a differentiation/separation between foreground and background?

 
Some objects group together with each other and "feel foreground," while some group together into the background. Too much separation will mean that the background objects don't blend into a single impression of "background." Too little separation and there will be no structure at all.
 
It is the pattern of "foreground/background" that I want to be clear. What you are calling clarity might be the "sharpness" of a visual object in a visual analogy. In orchestral music, sections take turns being foreground and background. The musicians adjust their "sharpness" as needed. That should be clear in the recording. That's one example of "clarity."
 

 I'm not sure I understand what you're saying, what is the right way to look at the problem? If I see a pig (A) and imagine that pig flying (B), I can accept that my imagination itself exists in reality but not that what I'm imagining does, I don't accept that flying pigs actually exist. If I want to understand something about the process of imagination, isn't comparing the difference between A and B a good place to start and then coming up with theories/experiments to explain the difference? What's the alternative, coming up with theories/experiments to explain how pigs fly?  
 

 
I think I misinterpreted your point previously, so I don't think we disagree here.
 
I don't disagree with most of what you say about the producer's and engineer's jobs. To some extent you are talking about the division of labor, but I'm talking about the criteria by which a recording is judged as accurate. The labor or technique depends on who's doing it and who has control of what. Does the musician need to change their technique? Does the engineer need to change the miking? To answer that question, you need an assessment of what's going on in the sound.
 
If we are talking about recording a live acoustic classical event, and if we are talking musicians who have honed their sound to a high degree of precision, then I think the recording should be judged by asking "How well does it work as music?"
 
I can hardly believe you are still saying a musician has a "vague" idea of what they sound like. What are they doing all those hours in the practice room? As nearly as I can tell, you are saying they don't know their tonal balance and their reverberation field from an audience perspective. But that's a tiny sliver of all the details they have practiced. For instance, all the musical shapes. Subtle variations in articulation, dynamics, timbre for instance. If Hilary Hahn has only a "vague" idea of what she sounds like, then how did she learn to sound so terrific from an audience perspective? I think the answer is that she has a very good idea of 99% of the detail she performs. Adjusting for an audience position is 1% of the technique, 1% of the practice time.
 
All those details of miking and phase and so forth, first let me say I understand you have a lot on your plate as the engineer and you need to do all that stuff well. I'm just saying that I think those choices should be judged by how they impact the music.
 
I agree that you've done a great job if you accurately reproduced a terrible performance! Of course what we hope for is an accurate reproduction of a great performance, and no recording I've ever heard even remotely approaches the clarity of sitting in the hall. So it seems that an engineer needs to do the "least damage" to the result.
 
Regarding using spot mics to highlight soloists--- you know, there's plenty for cognitive science to investigate in just the sound. And musicians do things with their sound to bring stuff to the foreground. If we are talking a great conductor, he/she will be balancing the sections and soloists. So just don't mess with it. If the use of spot mics gets closer to the SOUND in the hall and closer to the INTENTIONS of the conductor, then I'm all for it. It does seem like it adds a complicated extra layer, however.
 

 

[Joe Bloggs]

On this thread there has been the statement "musicians don't know what they sound like." That raises the question, what exactly is their goal when they are practicing if they don't know what they "sound" like? The only way I can make sense of this is to say, "Musicians don't know the tonal balance or reverb field out in the audience." That may be true, but "tonal balance" and "reverb field" is only a sliver of the properties of the final sound. So a recording engineer who is great at capturing that but knows nothing of other patterns in music, is not in a good position to evaluate the recording.

I can hardly believe you are still saying a musician has a "vague" idea of what they sound like. What are they doing all those hours in the practice room? As nearly as I can tell, you are saying they don't know their tonal balance and their reverberation field from an audience perspective. But that's a tiny sliver of all the details they have practiced. For instance, all the musical shapes. Subtle variations in articulation, dynamics, timbre for instance. If Hilary Hahn has only a "vague" idea of what she sounds like, then how did she learn to sound so terrific from an audience perspective? I think the answer is that she has a very good idea of 99% of the detail she performs. Adjusting for an audience position is 1% of the technique, 1% of the practice time.

So you would have us believe that the "tonal balance" and "reverb field" is only a "tiny sliver" of the sound properties, yet when we pointed out the properties of the recording only affects those properties and hence does not affect one's perception of the performers' skill, you throw your own assertions out the window to argue with us...

 

[Joe Bloggs]

I agree that you've done a great job if you accurately reproduced a terrible performance! Of course what we hope for is an accurate reproduction of a great performance, and no recording I've ever heard even remotely approaches the clarity of sitting in the hall. So it seems that an engineer needs to do the "least damage" to the result.

Or you just need to improve the tuning of whatever sound systems you're using to listen to said recordings...

 

[johncarm]

This is all quite whimsically depending on a series of unlikely assumptions:
1. That your interpretation of the musical score is the same as MTT's, and hence, MTT worked to segregate the woodwinds and strings in this particular soundscape (when he might just as well have been working to blend them)
2. That the performers then played out their parts as he desired
3. That, when all is said and done, that the woodwinds and strings actually sounded well separated at the positions where the mics were placed for the recording. Because from the context of your writing it's obvious that you weren't present at the performance where the recording was made, let alone sitting anywhere near the mics.

As for the rest, I have outlined a series of technical decisions that can actually affect the recorded outcome (as opposed to wishing how the outcome should be). I asked in my previous post and I'll ask again: is there anything in your theories that would allow the sound engineer to make those technical decisions better?

Suppose MTT listened to the recorded performance and quipped that "the woodwinds and strings don't sound well differentiated". THEN the SE could have done a dozen things to make the two jump out of each other, especially if the recording were redone or if the recording were a multitrack recording. Is there anything you can add to this?

 
 
 
 

So you would have us believe that the "tonal balance" and "reverb field" is only a "tiny sliver" of the sound properties, yet when we pointed out the properties of the recording only affects those properties and hence does not affect one's perception of the performers' skill, you throw your own assertions out the window to argue with us...

 
I'm making some observations. You are the one with theories.
 
It's very easy to observe the following:
 

1. When listening to a good musician up close, it is obvious that their sound has many nuances that contribute to a musical effect
2. 99% of those nuances were learned while practicing alone
3. If you move around the hall, the nuances don't disappear just because you changed locations. In a good hall, they will be mostly audible
4. However, in any recording I've heard, the nuances are considerably less clear, a much greater effect than moving around the hall
5. The problems in recordings affect whether the music "works" properly

 
You are the one with theories about why these things occur, or maybe you use your theories to deny that some of them occur. I stand by these observations, but I don't have a full explanation.
 
Here's an observation about music. The twin concepts of integration and differentiation are integral to composition, orchestration, performance, and conducting.
 
The MTT example is only an illustration of the problem. Of course the answer is not in my playback system, but in the original recording.
 
If cognitive science is going to investigate music, I think it should include an investigation of integration and differentiation and other concepts that are deeply embedded in music.
 
By the way, yes, sometimes the composer uses woodwinds and strings at the same time. That's just one more kind of tone color that needs to be differentiated.
 
What use are my observations? It's about where to fix your gaze, where to set your destination. There are many, many possible results to changing how you define your goal. That should be obvious. I am not an engineer so I don't have a list of equipment or microphones. But do we agree that where you set your destination affects the result? Also, do we agree that there is no one right mic or technique for all situations?

 

[Joe Bloggs]

I'm making some observations. You are the one with theories.

It's very easy to observe the following:

• If you move around the hall, the nuances don't disappear just because you changed locations. In a good hall, they will be mostly audible
• However, in any recording I've heard, the nuances are considerably less clear, a much greater effect than moving around the hall

These are your personal observations and they don't jive with mine!

But then if you're as dismissive about the importance of your playback system as this line indicates

The MTT example is only an illustration of the problem. Of course the answer is not in my playback system, but in the original recording.

How good could your playback system possibly be, to lead you to an absolutist declaration of playback vs live quality? "Not very good" would be my hunch.

 

[johncarm]

These are your personal observations and they don't jive with mine!

But then if you're as dismissive about the importance of your playback system as this line indicates
How good could your playback system possibly be, to lead you to an absolutist declaration of playback vs live quality? "Not very good" would be my hunch.

And your observations are your personal observations.
 
You didn't read me correctly. The kind of problem I was describing with the MTT recording (I mean in general, not necessarily that recording as I wasn't there to check), is a recording problem.
 
You know what's kind of funny? Objectivists don't like it when an audiophile says "you must have lousy ears." You have basically come up with the inverse insult. If an audiophile thinks live music is vastly superior to playback, "you must have a lousy system."

 

[icebear]

These are your personal observations and they don't jive with mine!

But then if you're as dismissive about the importance of your playback system as this line indicates
How good could your playback system possibly be, to lead you to an absolutist declaration of playback vs live quality? "Not very good" would be my hunch.

Here you go

:
[Headphone Inventory
Audeze LCD-2.2, LCD-X, Hifiman HE-500

Headphone Amp Inventory
Schiit Lyr2

Source Inventory
Schiit Gungnir]
edit : strikethrough, not up to date ...
 

 

[johncarm]

  Here you go

:
[Headphone Inventory
Audeze LCD-2.2, LCD-X, Hifiman HE-500

Headphone Amp Inventory
Schiit Lyr2

Source Inventory
Schiit Gungnir]
 

 
That's out of date, but anyway I've heard many more systems that I have ever owned. All sound playback is lacking something by comparison to live. There's a pretty huge gap. Many people agree with me. Many people don't. Either observation is worth investigating.

 

[icebear]

   
That's out of date, but anyway I've heard many more systems that I have ever owned. All sound playback is lacking something by comparison to live. There's a pretty huge gap. Many people agree with me. Many people don't. Either observation is worth investigating.

Maybe it's a good idea to:
 
1. update your profile to show what you are listening with right now, so we get an idea of your point of reference in playback
2. give an example of a recording that you think is of examplary clarity and is as good as it gets acc to your standards

 

[johncarm]

 
Maybe it's a good idea to:
 
1. update your profile to show what you are listening with right now, so we get an idea of your point of reference in playback
2. give an example of a recording that you think is of examplary clarity and is as good as it gets acc to your standards

 
The only good equipment I own is headphone equipment, but if the question is about how close a system comes to live sound, the best speaker systems I've heard are much better than any of my headphones. The Wilson Audio Alexandria speakers made me sit up and take notice. All audiophiles electronics, of course, including tube amps. Qualities like I'm talking about (differentiation, dynamic resolution) were pretty good in that system.
 
The audiophile label Sheffield Lab makes consistently musical recordings, and that includes the L.A. Phil.