Audio Science: Does it explain everything about how something sounds?

[stehno]

I think that applying numbers for the sake of argument is a bad practice: First because use of such numbers tends to cloak an argument in undeserved scienticity(!). Second because the conclusions drawn tend to be dependent on the values chosen.

First, hypotheticals are used routinely to illustrate points. I could have used letters from the alphabet but my point most likely would have been more lost. Are you saying you prefer to throw the baby out with the bathwater because of a bad practice? BTW, I’m a little confused about your tolerance of good / bad practices within the industry but I’ll save that for another time.

Second, is it possible that a non-science-minded type (me) was simply trying to put all the cookies on the lower shelf to illustrate a point that others often times attempt to overly complicate or prefer it remain a complex unknown?

Well argued but, imho, not proven.

What’s not proven?

- That much raised noise floors are far and away the single biggest culprit causing significant percentages of a given recording’s music info to remain inaudible at the speaker output?

- That noise floor levels are determined by the sum total of various distortions induced on our sensitive components?

- That music info below the noise floor remains inaudible?

- That a distortion is nothing more than adding to or subtracting from the original signal?

- That anything less than 100% must be a percentage of that 100%?

Not proven or not understood?

Frankly, Kal, I’m at a loss for words here as I thought I was stating the obvious even if on the surface some of it may not seem obvious. Nevertheless, I’d be interested to know, if in your opinion noise floor levels are not the greatest and final arbiter that determines levels of musicality in a given system, what do you suppose it might be?

 

[stehno]

You misunderstood me. I mentioned this as one possible example of many that could be used to set a standard. I was not suggesting that it was a good example, but rather as a possible example of what we could agree or not agree to use as a standard. This is one of many possibilities.

Got it. Sorry Peter, I misunderstood.

 

[Don Hills]

I am not implying anything about noise floor because I am not obsessed with it.

So you enjoy viewing blurry photographs?

You can't even get that right. A photograph can be blurry. It can be noisy. They are independant effects. Noise in a photograph is usually there as a deliberate artistic choice. Less so in audio.

 

[853guy]

Where did I say that was the "real art?"

You said it here in post #579:

"All that is happening here is gathering raw elements. Some elementary mixing is done to allow them to play the track but that is not at all what we are going to hear on the album. Tons more manipulations happen with the artist and label buyin before we get the real art."

I said that the recording is the second instantiation of art. The first one, the live version, is inaccessible to us. You only got to see it in that Youtube video.

Perhaps this is a difference of degrees. Yes, we weren’t there, so in that respect, the event itself is completely inaccessible to us, I agree. But no, you said “All this is happening here is gathering raw elements. Tons more manipulations happen with the artist and the label buyin (sic) before we get the real art.”

The finished version at highest fidelity ironically lands far more removed from the original art than that Youtube video! So if your goal is to hear the original art, then you should listen to that Youtube version, not the final record.

You have listened to both versions, right? Because I’ve listened to the Youtube video and the finished version, and you know what? They’re the exact same mix. We are listening to the final version in the video, which is why it’s crazy to me you’ve chosen this particular example. (Milli Vanilli’s Girl You Know It’s True - now that would be a great example.) Why? Because George Massenburg tracks in a way that’s as close to what he wants to put out on the record as possible. That’s what he’s known for as a producer. (It’s difficult to say categorically that the CD version has been mastered in a way the Youtube video hasn’t but I’m not going to make definitive statements based on a 240p Youtube video vs. anything, anyway.)

So the point is not that the original art is unimportant. It certainly is not. It simply is the case that you cannot ever make a reference to reproduction in your system as being a version of a live event. That event was not presented to you in the recording to mimic. You were given an alternative version and all you can do is be faithful to that.

Well then I’m not sure what it is you’re listening to. Again, you chose the Youtube video, of a one-take, ready-mixed, everyone-in-the-same-room playing music together in which the final version is the version of the live event.

Oh, he is part of creating the art. There is a reason top recording engineers are in such high demand by artists. They are part and parcel of making great music, lest you tell me Keith Johnson won his Grammy for no reason:

Dude, the Grammy’s? Really? The awards ceremony that has awarded Pat Metheny 20 times over and given Grammy’s to the Baha Men for Who Let the Dogs Out and Macklemore for “The Heist”?

The same awards ceremony that’s denied Bob Marley, The Velvet Underground, Jimi Hendrix, Nas, Bjork, Queen, Patti Smith, Public Enemy, Ramones, Talking Heads, The Who, Morrisey and The Doors though nominated many times over and only awarded Led Zeppelin last year with an actual Grammy?

Pfff. C’mon…

(By the way, Keith Johnson has three Grammy’s - George Massenburg has four. Does that tell us anything meaningful? No.)

I have the hugest respect for producers/engineers. My personal heroes include Terry Date, Michael Beinhorn, Bob Ezrin, Mr. Massenburg, Brian Eno, Daniel Lanois, Rick Rubin, Bob Rock, Steve Albini, Toby Wright, Trevor Horn, Mutt Lange, Roy Thomas Baker, Steven Lipson, Bob Clearmountain, T.Bone Burnett, Joe Henry, David Bottrill, Gyln Johns, Ethan Johns, Nigel Godrich, Peter Collins, Rubert Hine, Brendan O'Brien, Bill Laswell, Alan Parsons, Bob Thiele, Butch Vig, etc, etc.

They've all contributed significantly to the art form we know as music, and I tried to emulate many of them in my short-lived but highly educational career. Their ability to help shape the performance, content, arrangements, and sonics have been the defining features of their respective careers. And some of them have made their careers by being incredibly hands-on (Trevor Horn comes to mind). But only a few of the above would be so bold to consider themselves artists in their own right, most of the ones above try and serve the intention of what's already there, by allowing the artist to fulfil the potential of each creative act and idea. Some, but not all.

It is and only exists in the Youtube video. Not the CD.

(I can't work out how to quote me in your quote but I said)



So, to come back to this point again…

You're saying that the video of the session contains “live music” but the CD of the same session - even though they are the exact same session and the exact same mix - does not contain “live music”?

Live music happened on that day, we can agree on that. Someone captured the event with some mics/pres/comps/EQs and through a desk (sonically) and someone else captured the same event with a video camera (visually, duh) using the exact same feed from the exact same session but the video is somehow superior to the CD?

Again, since it’s possible I am a terrible communicator and horrible writer of my thought processes, let me try and be more clear:

Live music happens in studios everyday around the world. It’s what Keith Johnson records and what George Massenburg records. At some point in time, all music is live, because there is no music made without human intention, whether it be by vocal chord, Telecaster or MPC. Whether that intention survives the process of being recorded/mixed/mastered and played back it part of why I think this thread was started. My take - and I believe I’m not alone in this - is that some components are much more able at passing on that intention to the listener than others, and measurements cannot quantify that aspect of it.

All true but not related to the argument I made which is the inaccessibility of original art to us as buyers. Clearly it is absurd to say there was no art until the recording was made, a point which I did not make.

Well if the original art survived the recording/mixing/mastering/stamping process (and can survive intact onto the internet via a Youtube video at 240p, no less), then hopefully that’s what we buy as consumers. Music is the most democratised art form there is. The fact that the original art is copied thousands of times and mass distributed doesn’t demean the art form, nor lessen its potential impact on the consumer. But I think we might all agree that certain things can impede the art and lessen its impact, some but not all of which have to do with the hi-fi gear we play it through.

Measurements actually tell us why it is impossible for us to even hear the art that was created in the studio, let alone upstream in the live session.

Can you clarify what you mean by this?

Genelec is a common brand of professional speaker used in pro world. Here is a sample measurement of it:

And here is a Wilson:

No way would these two loudspeakers produce the same anything. And we can use measurements to show that.

Yep, I’ve used Genelec’s in the studio before. I’ve also used ProAc, Dynaudio and ATC. They all measure differently.

Can we use those measurements to show how they alter the physiological and neurological state of the listener when listening to music? No.

Hypothetically, could we create a test in which subjects are subjected to a hearing test, drug test, health check, MRI and put on the same diet in order to minimise the biological, physiological and neurological variables, monitor their heart rate, pupil dilation and brain activity, and sit them down in front of a bunch of components (whose measurements have been independently verified under the same conditions and are hidden from sight), and take meaningful measurements about how we as human beings respond to those components when playing back music, while attempting to draw conclusions about the electrical measurements of the gear playing back steady-state signals vis-à-vis the measurements of our body/brain when listening to those same components playing back music?

Probably. I think it could be fascinating. Though I don’t see anyone trying that anytime soon.

But you are right that measurements don't capture anything about the art. We measure the art with our brain.

I completely agree, Amir.

 

[853guy]

I must admit that I follow your line of thinking and for my ears and my music your beliefs are akin to mine.I enjoy reading your posts.

Thanks, Steve. Maybe it's just you and me then.

Thankfully no-one's died in the making of thread and no animals were harmed. Though it's possible some spouses and children have let out long sighs of bemusement and rolled their eyes.

Right, off to play tickle-town (with my children, not my wife, if that needs clarification).

 

[Groucho]

A photograph can be blurry. It can be noisy.

The analogy with photography makes an interesting contrast I think. Why do photography magazines and reviewers never home in small details in the image and wax lyrical about them ("They were so realistic, I thought I could reach out and touch those kittens")? Why do they seem to think that objective specifications are everything? It's almost as if they are saying that simple linearity is all that is required of a photographic system. Ditto for video.

I find that hi fi people are very hard to pin down on this. Do they think that the ideal audio system would simply be perfectly linear?

 

[microstrip]

Well, it is more difficult simply because of the increased number of variables consequent to the increased number of components. OTOH, it is not proportionally more difficult because those variables provide more tools. However, one cannot simply apply one's stereo experience directly to a MCH setup because the two demand a different interaction with the room. Stereo requires substantial support from room reflections to support the perception of space/ambiance. MCH requires, one might say demands, less room contribution as most of the necessary space/ambiance information along with their directional cues are provided by the source and system.

In theory, nothing. In practice, it is the choice between different compromises assuming that we do not have perfection. MCH music has an aesthetic shared with good stereo music which prioritizes tonal accuracy, clarity and neutral balance. HT should do the same but, often, there is a trade-off for extremely wide dynamic range, substantially extended bass range and power and dialog intelligibility (sometimes unnaturally in the context of the action).

Thanks - I would naively consider that a system needing less support from the room would be much easier to set. And I appreciate your frontal view considering theory versus practice.

 

[Kal Rubinson]

Thanks - I would naively consider that a system needing less support from the room would be much easier to set.

The task for MCH (and for stereo) is to have the room not impose its influence unduly.

 

[Kal Rubinson]

What’s not proven?

- That much raised noise floors are far and away the single biggest culprit causing significant percentages of a given recording’s music info to remain inaudible at the speaker output? ?

This.

 

[Kal Rubinson]

Also there is not as much choice of processors. You have the low end ones like Denon etc - and then a sudden jump to Meridian, Datasat, Trinnov, and Illusonic, all which are at a similar price and seem to be evolving much more than 2-ch preamps.

For MCH music (not HT), you do not need a processor at all and, at present, I do not have one in my main system. All my recordings are being converted to files on my NAS, a PC-based server is my interface, software handles bass management and roomEQ and an Audio Research MP1 analog preamp does the switching and level control.

Source can be an oppo, that's fine.

Yup. An Oppo 105 feeds the same preamp. It allows me to play physical discs including those that require decoding.

 

[stehno]

You can't even get that right. A photograph can be blurry. It can be noisy. They are independant effects. Noise in a photograph is usually there as a deliberate artistic choice. Less so in audio.

Last time I checked a noise is a distortion and I think I covered that when I said "a distortion is nothing more than adding to or subtracting from the original signal."

Perhaps injecting DSP into the mix is similar to a photographer intentionally injecting "noise" into a photograph. But your point is really outside the scope of the topic except perhaps for the fact that by the addition of a DSP component (or any component) the noise floor had to rise a bit more just by merely adding the component into the system.

But I stand by my claim that a much raised noise floor is far and away the primary silent killer of a playback system's level of musicality.

 

[Bobvin]

Thankfully no-one's died in the making of thread and no animals were harmed. Though it's possible some spouses and children have let out long sighs of bemusement and rolled their eyes.

Right, off to play tickle-town (with my children, not my wife, if that needs clarification).

+1

Yesterday afternoon we were several pages shorter on this thread. This morning after taking the time to read additional posts I really know nothing new except several folks are splitting hairs and mincing words, parsing sentences, arguing this or that point. It is summertime, and on a Friday night y'all ought to be enjoying summertime, being with friends and family, and / or enjoying your music/system with a refreshment of choice. Noise floors and MCH vs 2 channel, has filled my head with noise this morning reading about it. Lets not dissect each post until we end determining what the meaning of 'is' is.

As for me, sometimes I hear improvements, other times not. I like it when science provides an explanation to why I heard an improvement, and I get frustrated that science does a poor job identifying, for example, why one set of cables sounds different from another. And even if I hear an improvement, if the mfg throws out things like "quantum coherance" I immediately rule out his products because of lack of understanding on my part, or my bullshit antenna start quivering uncontrollably because the mfg is using techno-babbel to convince me.

And wrt noise floor, yes I hear more from a recording when the floor is lowered, but significantly more to do with noise floor in my room which can be a large difference vs a fraction of a db due to a technical improvement.

 

[PeterA]

Thanks, Steve. Maybe it's just you and me then.

Thankfully no-one's died in the making of thread and no animals were harmed. Though it's possible some spouses and children have let out long sighs of bemusement and rolled their eyes.

Right, off to play tickle-town (with my children, not my wife, if that needs clarification).

No, I agree with 853guy also. So there are at least three of us, and I'm sure, many more.

 

[Orb]

Last time I checked a noise is a distortion and I think I covered that when I said "a distortion is nothing more than adding to or subtracting from the original signal."

Perhaps injecting DSP into the mix is similar to a photographer intentionally injecting "noise" into a photograph. But your point is really outside the scope of the topic except perhaps for the fact that by the addition of a DSP component (or any component) the noise floor had to rise a bit more just by merely adding the component into the system.

But I stand by my claim that a much raised noise floor is far and away the primary silent killer of a playback system's level of musicality.

Well not sure if noise could be classified as a distortion, it does not have the same or even comparable effect as say harmonic distortion, whether that be influence on the electronics or perception of the original signal (even test tones).
In digital one could show the differentiation between noise and quantisation distortion.
Just two quick and dirty examples.
Cheers
Orb

 

[stehno]

Well not sure if noise could be classified as a distortion, it does not have the same or even comparable effect as say harmonic distortion, whether that be influence on the electronics or perception of the original signal (even test tones).
In digital one could show the differentiation between noise and quantisation distortion.
Just two quick and dirty examples.
Cheers
Orb

Not to split hairs or go down unnecessary rabbit holes, but does a noise add to or in any way alter the original signal? If so, then by definition it could it not be classified as a distortion? I mean we are talking about "hi-fi" or high-fidelity here, right?

 

[stehno]

Quote Originally Posted by stehno View Post
What’s not proven?

- That much raised noise floors are far and away the single biggest culprit causing significant percentages of a given recording’s music info to remain inaudible at the speaker output? ?

This.

I'm not in a position to remotely demonstrate noise floor levels and how much of the music info remains audible above a noise floor and how much remains inaudible below a noise floor.

But if you had any interest in the subject, I suggest that you try placing one of your multichannel speakers on top of one of your sensitive components or on your racking system that sustains your sensitive components. If you do, I'll bet dollars-to-donuts you will be hard pressed to hear any sonic degradation in your system's level of musicality. The implication being your system's noise floor probably can't get much worse than it already is.

That in and of itself doesn't prove much, but such an experiment could open the door toward proving the significance of noise floors.

 

[amirm]

You said it here in post #579:

"All that is happening here is gathering raw elements. Some elementary mixing is done to allow them to play the track but that is not at all what we are going to hear on the album. Tons more manipulations happen with the artist and label buyin before we get the real art."

I will just comment on this part here as it seems to have triggered the misunderstanding. My use of the word "real" here was intended to mean that it is the finished art that was delivered to us. In that sense, it is the "real deal." It does not in anyway imply there was no art created in the gathering of the raw elements. Of course the talent creates the raw elements that go into the production of the final and "real" product that we eventually get in our hands.

 

[amirm]

Live music happens in studios everyday around the world. It’s what Keith Johnson records and what George Massenburg records. At some point in time, all music is live, because there is no music made without human intention, whether it be by vocal chord, Telecaster or MPC. Whether that intention survives the process of being recorded/mixed/mastered and played back it part of why I think this thread was started. My take - and I believe I’m not alone in this - is that some components are much more able at passing on that intention to the listener than others, and measurements cannot quantify that aspect of it.

You are arguing a very different point. The debate is whether what we hear at home can approach what one would have heard in the real venue. It is not whether elements "survive" the transformation through production of finished music. Of course elements do that. A female singer doesn't become a male singer in the process . The question is if I magically did an A/B between you sitting in the live presentation and through a stereo system, would they be indistinguishable. The answer is that you can't remotely, in a million years, achieve that with a stereo system with music as is produced today. No microphone captures the room. Music is mixed, EQed, transformed, etc. To say nothing of the fact that tonally you have no idea how close your system is to the original one that was used in the production of music that the talent heard.

This is the concept that we need to understand and forever put behind us the #1 marketing technique used to sell us audio gear, dating back to the days of Edison claiming the same "live" music reproduction.

Well if the original art survived the recording/mixing/mastering/stamping process (and can survive intact onto the internet via a Youtube video at 240p, no less), then hopefully that’s what we buy as consumers.

That is what we buy as consumers. But no, that does not mean in any way or shape that if our equipment achieves nirvana, it somehow remotely matches what was heard in a live session. We are hearing a completely different presentation of the original art.

Can you clarify what you mean by this?

I was showing the frequency response of two loudspeakers: one that is a top brand in professional music production circles (Genelec) and one in consumer (Wilson). When fed the identical signal, both reproduce wildly different tonal response. That tonal response is a linear transformation that everyone, no matter how critical their hearing, can hear and differentiate. No way would the sound heard by the Genelec speaker be the same if you replaced it in the same production room with the Wilson. Therefore we have already radically changed what we hear versus what the talent heard. So no way can we say that we are replicating the "live" experience if in this case we forgive and forget the recording process that led to the final stereo mix.

In other words, there are two major barriers to us claiming that we are approximating the live experience. The first one is that no one attempted, nor is it possible to capture a live session with 100% transparency into two loudspeakers. And two, that the production and playback chains have no standards that would make the sound the same. These combined mean that no way, no how do you ever, no matter how good you think your system is, are hearing anything like what was heard either in the live recording session, or what final stereo production.

 

[Kal Rubinson]

I'm not in a position to remotely demonstrate noise floor levels and how much of the music info remains audible above a noise floor and how much remains inaudible below a noise floor.

But that's what is needed for proof.

But if you had any interest in the subject, I suggest that you try placing one of your multichannel speakers on top of one of your sensitive components or on your racking system that sustains your sensitive components. If you do, I'll bet dollars-to-donuts you will be hard pressed to hear any sonic degradation in your system's level of musicality. The implication being your system's noise floor probably can't get much worse than it already is.

It can imply that or that your presumption of sensitivity is mistaken or that the added weight reduces noise generation in that "sensitive" component.

That in and of itself doesn't prove much, but such an experiment could open the door toward proving the significance of noise floors.

I see no reason to make that effort. Look, as most would agree, reduction of noise in a system is important. Your contention that it is the most significant factor is not proven to me.

 

[Goliath]

As for me, sometimes I hear improvements, other times not. I like it when science provides an explanation to why I heard an improvement, and I get frustrated that science does a poor job identifying, for example, why one set of cables sounds different from another.

"Sounds different" may result from physical changes to the soundfield/sound waves impinging upon the ears, or ... the eyes, ears, knowledge, biases, imagination etc etc, mood, price, brand, etc, and so on and so forth, etc.

Any number of confounding variables may have resulted in "sounds different" including, but not limited to, nothing at all, but that also leaves the possibility open that the cable audibly sounded different. I don't think science does a poor job of identifying audible (as in, physically altering/affecting the soundfield/sound waves).

Measurements don't cover/address perceptual errors/psychological influences ie things that aren't in the soundfield/sound waves, so one has to weed out the culprit via process of elimination. Time consuming to be sure.

And even if I hear an improvement, if the mfg throws out things like "quantum coherance" I immediately rule out his products because of lack of understanding on my part, or my bullshit antenna start quivering uncontrollably because the mfg is using techno-babbel to convince me.

Well, "hearing", or perceiving a difference/improvement is all that matters. Doesn't matter what caused it, just that you "heard" it. It's all about listener satisfaction anyway.