On the relativity of audio phenomenon...

[Boomzilla]

From a disagreement with @tllfkacn (whom I respect & admire) in another thread, the question has (again) come up about the relativity of audio observations. The scenario under discussion is as follows:

An electronic component is replaced in an otherwise unchanged system. If differences are heard, I contend that they must be from the changed component. Chuck responds:

1. "That is presuming that differences between two high quality stereo preamps and two high quality/high power power amps can actually be clearly heard at reasonable levels. That is most times questionable."

I respond: In some cases, yes, it could be questionable. If the remainder of the system lacked sufficient resolution to reveal the differences, then you'll never hear a difference. However - if the remainder of the system DOES have sufficient resolution to reveal small changes, then this is no longer "questionable" - it's a slam-dunk. Last night I listened to three different power amplifiers driving a pair of Martin Logan electrostatic speakers. The characteristics of the amps were so different that even the tedious "double blind test" would have clearly revealed them. My "normal speakers" are significantly lower resolution than the ML ones, but even there, changes are obvious.

2. "However, for the moment let's say that you in your room do make a change in the preamp and now you hear a difference as you describe as a slightly "polite" treble (I presume you mean the treble level has decreased slightly). How do you know that the preamp before the change was perfectly flat? That answer is you don't! The same for the power amp and the power amp you had before the change. You are making presumption that your evaluation was based on a perfect neutral system in your room in the first place."

I respond: This is the crux of the argument - is there an objective standard to measure against? I'll readily concede that without significant and extensive measurements, then no - there is no objective standard. But there IS a subjective one. And it has nothing to do with the differences between components. Every one of us has the "standard" of live music that we know and measure against. Now the majority of us have never heard live music in our own living rooms (but I have, since my daughter is a musician). But even so, we know live from recorded. We've all had the experience of being in a location where a real band is playing, but out of our immediate sight. We know within seconds whether the music is live or recorded. The brain can tell.

We use that subjective standard to compare against musical reproduction in our homes. Yes - our memories are imperfect. Yes - the venue alters the expectation. But ANY listener can decide in short order "more like real music? or less?" And the imperfect vocabulary we have to apply to the experience leads to comments about a "polite" treble. Unless specified, this isn't a comment about one component vs. another - it's a comment about the component vs. our best expectation of what live music would sound like.

So I claim that audio commentary is independent of the room - independent of the previous components (unless specifically noted as a comparison) - and independent of whether the component is flat or colored. Audio commentary is comparative to the "reference standard" of our own ears' experience with live music.

The advantage of this reference is that it is already (and absolutely) pre-calibrated to our own hearing. How YOU hear a live performance is not exactly how I hear the same performance. Our ears and hearing are different. But once we know what live music sounds like (to ourselves), then our reference is established for any future comparisons with reproduced music.

Let the philosophical tennis match begin!

Boomzilla

[monkumonku]

Oct 4, 2016 4:39:14 GMT -5 Boomzilla said:

From a disagreement with @tllfkacn (whom I respect & admire) in another thread, the question has (again) come up about the relativity of audio observations. The scenario under discussion is as follows:

An electronic component is replaced in an otherwise unchanged system. If differences are heard, I contend that they must be from the changed component.
Boomzilla

That seems logical, because with all other components constant except the different one you introduced, it must be the different component causing the difference in sound.

However, perhaps something about this different component has an effect on one of the original components, which causes that original component to sound different. Like an impedance mismatch. So while it is the new component that has brought about a change in sound, the bottom line is it is the reaction of one of the original components to it that is actually responsible for the difference in sound. Or maybe you can say they are co-responsible.

You take that same new component and put it into a different chain of other components and there may be no effect on the sound whatsoever.

In other words, it could be the synergy of the entire population of components that makes the system sound the way it does.

On the other hand, I'm a believer that aside from speakers, if you operate components of the same type within their intended specs/ranges then you ain't gonna hear much if any difference between them unless they are deliberately designed to not be neutral.

[garbulky]

Well we don't have the measurements and we don't know what the measurements tells us compared to what we hear.
For instance...we don't listen to a 20-20khz sweep. We don't listen to THD or SNR.

We listen to music. We do hear a soundstage. We do hear tonal balance. We do hear realism. We can hear the same treble pitch but just a slight change can make that same treble pitch sound "lacking in snap" rolled off, not airy etc. None of these are real things but it's constructed by what we hear.
The reality is in our home we don't have the measurements nor the ability to intepret it to the accuracy our ears and brain can do with talking about things like what I just described like soundstage, tonal balance, microdynamics.

And even if we did....would it matter? If it measures a certain good way but sounds worse then is it a better piece of equipment or not? The answer is....it's not.

Now the reasonable question is....how do you know that the other peices of gear weren't basically "balancing each other out" while all sounding imperfect. Introduce a neutral unit and now you have unbalanced the other gear. So how do you know which is better?

Well that's a reasonable question. I have made that same mistake myself.

Let's say you have some bookshelf speakers. These are - without you realizing - not that great in bass. Now you purchase a preamp that's bass heavy. Now suddenly the recordings are sounding way more balanced. The top to bottom balance is better. So you turn around and blame your previous preamp for being lacking in bass. When in reality, the current preamp was not that great. But your speakers were the the real problem.

So it has some validity. But you then have to consider two allowances.
First - most gear on the market already spec out to inaudibility with measurements available to us. So you have to allow that DIFFERENCES EXIST AND THEY ARE AUDIBLE if you go down this route about imperfect upchain electronics.
Second - If they are audible then would it be so extreme to think that more likely than not a better piece of gear does indeed sound better. Isn't that the more likely scenario than you have a perfect storm of imperfection that some how makes things sound better and a neutral unit sounding worse.
It's possible but is it the more likely of the two?

If somebody was hearing increased detail and more realistic sounds, in their setup, wouldn't it be better to assume that this was the case?
Now we can say - well it's in their setup. But then we can turn it back and say okay....give me the huge number of measurements - which we don't have and that we can't REALLY intepret anyhow - and tell me how that correlates to this gear sounding better. Is that the better solution?


There's drawbacks either way.
Hearing - you can actually intepret it to the only person that matters - you - in the only place that matters - your setup. Then you tell it to someone else. They or their setup can make things sound different. But it's only his experience that will _____ yours. Not the measurements he doesn't have or the interpretation of measurements he doesn't have.

But I argue that it's more useful than saying all things MUST sound the same because it's objective though we don't really have any proof for that. And yes double blind testing negatives are not proof. It's proof in THAT specific setup of POSITIVES - not negatives.

I acknowledge there are lots of errors with subjectivity. But I argue that it's more useful than being objective and just assuming the thing sounds the same. And if it doesn't, I don't think we should tip toe about mentioning what we hear with our ears and perceived with our brains. Because they are important. We would be foolish to dismiss an instrument that can describe in huge amounts of very accurate detail what it can hear just because it on ocassion fools itself.

[garbulky]

Side note: Differing hearing and setups that are more resolving than another can make differences in perception. It sounds snobbish. But it's not. It's a real thing. I've had setups that are not as resolving as others. They couldn't distinguish the increased detail. Poor room acoustics can muddy up the sound. Less resolving gear cna also muddy up the sound. Poor speaker placement can do the same thing.
I've known people that could not hear certain subtle details. That's hearing loss/damage. Or simply they cannot hear as well. It's inevitable to most of us. It doesn't mean I have golden ears. I certainly don't think I do. But it did mean I could hear things that some people couldn't because of hearing issues. There are times when I can't hear things that others can. So like it or not, there are differences in hearing and setups that can make a difference in perceiving a higher performing component.

I also don't think people have to be especially trained to be able to hear some of the more obvious stuff. I know I am not. But they have to be willing to listen.

[dcg44s]

Oct 4, 2016 9:15:20 GMT -5 monkumonku said:

Oct 4, 2016 4:39:14 GMT -5 Boomzilla said:

From a disagreement with @tllfkacn (whom I respect & admire) in another thread, the question has (again) come up about the relativity of audio observations. The scenario under discussion is as follows:

An electronic component is replaced in an otherwise unchanged system. If differences are heard, I contend that they must be from the changed component.
Boomzilla

That seems logical, because with all other components constant except the different one you introduced, it must be the different component causing the difference in sound.

However, perhaps something about this different component has an effect on one of the original components, which causes that original component to sound different. Like an impedance mismatch. So while it is the new component that has brought about a change in sound, the bottom line is it is the reaction of one of the original components to it that is actually responsible for the difference in sound. Or maybe you can say they are co-responsible.

You take that same new component and put it into a different chain of other components and there may be no effect on the sound whatsoever.

In other words, it could be the synergy of the entire population of components that makes the system sound the way it does.

On the other hand, I'm a believer that aside from speakers, if you operate components of the same type within their intended specs/ranges then you ain't gonna hear much if any difference between them unless they are deliberately designed to not be neutral.

This^  Not just with audio but with other things as well.You can take individual components that are all of very high quality themselves and combine them only to find that they just do not impress you as you had expected.Yet other times you can hit upon a combination that just flat kicks butt.It is that undefinable and unpredictable factor of synergy.

[repeetavx]

Many (many) years ago I tried an experiment. For some reason I had decided that I needed an external volume control between my CD player and my receiver. So being an electronics technician I grabbed a stereo potentiometer and wired it in. No problem varying the RCA output of my CD players volume, but something else happened. Not only did I vary the volume, but I was varying the tone (treble/Bass balance) as well. By varying the impedance that the CD player's output circuitry was seeing, it was varying its frequency response as well.

So whenever you have an analog signal, unless the circuit is engineered very carefully. Two amps with the same "specs", but different input impedance's, could have a different tonal balance. Like if you use a Y-cable to drive two channels when bi-amping your speakers.

Thankfully when I bi-amped my XPA-2 with a Y-cable, I experienced no tonal shift. But it should still be something to watch out for.

[Wideawake]

Oct 4, 2016 10:54:30 GMT -5 repeetavx said:

Many (many) years ago I tried an experiment. For some reason I had decided that I needed an external volume control between my CD player and my receiver. So being an electronics technician I grabbed a stereo potentiometer and wired it in. No problem varying the RCA output of my CD players volume, but something else happened. Not only did I vary the volume, but I was varying the tone (treble/Bass balance) as well. By varying the impedance that the CD player's output circuitry was seeing, it was varying its frequency response as well.

So whenever you have an analog signal, unless the circuit is engineered very carefully. Two amps with the same "specs", but different input impedance's, could have a different tonal balance. Like if you use a Y-cable to drive two channels when bi-amping your speakers.

Thankfully when I bi-amped my XPA-2 with a Y-cable, I experienced no tonal shift. But it should still be something to watch out for.

NOTE: I don't know anything about electronics. Just trying to understand what you've stated.

You state: "Not only did I vary the volume, but I was varying the tone (treble/Bass balance) as well. By varying the impedance that the CD player's output circuitry was seeing, it was varying its frequency response as well."

How does adding a volume control to the output of the CD player vary the impedance of the CD player's output circuitry? I would think it would vary the impedance of the input to the receiver, no? The CD player should be unaffected since the volume control is transparent to it. What am I missing?

[KeithL]

I have two comments here.....

First, yes, with some components there is in fact an objective standard.

Some components, like speakers, are complex. There are many different ways of measuring speakers, each measurement may itself be difficult to interpret, and a given speaker may sound and measure very differently in a different room. This means that you always have to qualify your measurements and observations with conditions - like "these speakers sound bright in this room when I'm sitting in that chair". However, some components, like preamps, aren't that complex. Most people agree that the job of an "ideal preamp" is to boost the level of the signal without changing anything else at all. And, while there are several things that a preamp can do"wrong", it's pretty simple to define the ideal situation. (If I set the gain of my preamp to x10, then divide its output by 10, the result should be indistinguishable from what I hear when I substitute a straight piece of wire for that preamp. ANY deviation from that result constitutes an error, and a variation from what it's supposed to do.) In that situation, while it's certainly possible for a preamp to "mess up" the signal in various ways, most of them can be objectively measured, and their variation from ideal specified. For example, the frequency response should be 20 Hz to 20 kHz +/- 0 dB; anything other than that is simply wrong.

Of course, if two preamps, neither perfect, vary from that ideal in different ways, then their errors may subjectively sound more or less unpleasant.... but we have an objective starting point or goal. There are any number of possible "wrongs" but only one "right" (and, while some wrongs can be more wrong than others, nothing can be "more right than right"). Another possibility is that, when we seem to be hearing things that disagree with our measurements, we may simply be misinterpreting what our senses seem to be telling us. For example, if I have two preamps, and one seems to be "brighter" than the other, yet measurements show that both are flat from 20 Hz to 20 kHz +/- 0.05 dB, then the one that seems brighter really isn't (sorry; measurements really don't lie). However, that doesn't mean that we're just imagining it either. Perhaps the one that seems brighter has a little more distortion at high frequencies, and we're misinterpreting that extra distortion as a boost in level at those frequencies. My point there is that the guy who thought that preamp was brighter wasn't entirely wrong, but there also wasn't some magical "subjective difference" that couldn't be measured - we had simply misinterpreted what we heard and so failed to measure the right thing. DACs often seem brighter, yet measure the same, due to time domain differences in their filters (our brains and/or ears seem prone to confuse the two); that's why Sabre DACs often sound bright even though measurements insist that they aren't. However, just because we haven't figured out what to measure yet doesn't mean that the difference isn't "objective".

Second, I would like to comment on the idea of "synergy".......

Let's assume that our starting point is a pair of speakers in a room - which combination we all agree sounds more or less neutral.
Now let's add an amplifier that also measures and sounds neutral, and a preamp with an obvious and measurable boost of +5 dB at 12 kHz.
Obviously that overall combination is going to sound "bright" or even "harsh".

So, what shall we do to fix it?
Here are a few ideas.....

1) replace the preamp with one that's actually flat
2) replace the amp with an amp that has a 5 dB dip at 12 kHz to cancel out the boost in the preamp
3) add some extra equipment, like an equalizer, to remove the boost
4) change something that's already there, like the speakers, to something with a dip that cancels out the boost

Now, many people who "believe in synergy" would say that any of those options would work just fine.....
Problem solved, harshness gone, all's good..... 

However, I disagree........
In fact I suggest that the ONLY real solution is #1.
Let's analyze the options.....

The first one actually locates and corrects the problem.
It will ALWAYS work.

All of the others assume that you can actually cancel out one error with another.
You can remove a boost with a dip, or vice versa, but there's no way to remove distortion, and no way to effectively correct many time errors caused by filters.
Also, you may not be able to find a second piece of equipment with an error that precisely compensates for the first error - resulting in an endless regress of corrections to corrections of corrections.
Therefore, the other options may or may not be possible in individual situations.

And, finally, you've created a system that has perfect synergy together (which is a positive spin on saying that "those components work poorly unless used in that specific combination").
This means that, if you ever want to change any one of those components, it's "back to square one" with finding other components that fit well with the ones you've kept.

Interestingly, this is just another variation of what Henry Ford is credited with figuring out (and I believe Winchester).
"If you have a bunch of standardized components that all perform as intended, then they will be interchangeable... and will work well in combination with other standardized components.
It is a lot easier to get consistent reliable results with standardized components than with components that must be custom hand fitted to each other."

[garbulky]

Exactly how would you interpret that result to say this preamp that measures identically is brighter in exactly this way? If we can't tell how it's going to sound subtly different with this type of distortion is it that useful? For instance what if the distortion was the more or less in the same place and same amount but it just looked slightly different on the scope?

[Wideawake]

Brilliant post Keith!

[Boomzilla]

OK, Keith - In theory I agree with what you've said. Now tell me how I measure imaging? Is it merely channel-to-channel separation, or is it more the interaction of the speakers with the room, or is it in the recording, or is it in the brain of the listener? Not everything is measurable. But our "inner reference" (is this more or less like real music?) never lies.

[dcg44s]

I can't argue with Keith that ideally you should solve any problems that you have in your system by replacing the component that is introducing the issue rather than trying to compensate your way around it but sometimes in reality that just isn't the best choice.

I don't think that most people choose components based on measurements alone and how many components measure up perfectly on everything anyway?Heck,a lot of people won't even know what the issue truly is,they just know that their system isn't sounding as good as they like.

If you find a component such as a preamp that just rocks your world,it looks great and has all the features that you want but it has that +5db boost at 12 kHz you may still prefer it over another that measures flat but does not meet your needs and desires in other areas.We almost always have to play that game of give and take.

In a case like that if you come across a combination of components that have that synergy that makes your system just flat sing for you it just might be the better option even though you did not follow the ideal path in theory to get there.

Getting back to Boomzillas original post,the main point of bringing up the topic of synergy was to point out that when doing audio observations where you hear changes upon the substitution of a single component it is undoubtedly because of that change but might be based off of the interaction of components rather than directly due to the component alone.

[yves]

Oct 4, 2016 12:52:42 GMT -5 Boomzilla said:

OK, Keith - In theory I agree with what you've said. Now tell me how I measure imaging? Is it merely channel-to-channel separation, or is it more the interaction of the speakers with the room, or is it in the recording, or is it in the brain of the listener? Not everything is measurable. But our "inner reference" (is this more or less like real music?) never lies.

Our inner reference does lie, and double-blind listening tests can reliably constitute this fact. Only problem, double-blind listening tests can also constitute, also reliably, this "other" fact that those who claim that we know everything there is to know about how the human auditory system works are those who are telling the bigger lie.

[Wideawake]

Leaving the room out of the equation, and taking into consideration modern solid state equipment, IMO the two components that will have the largest effect on SQ are the DACs and the speakers.

This is presuming that the component chain in your system comprises of standards based components from reputable manufacturers who know what they are doing. That is, it is assumed that the Preamp provides the appropriate voltage to the amps; amps have a gain between 27dB and 29dB with good S/N and low crosstalk; components measure relatively flat with no crazy peaks or dips; etc. etc.

Given the above, if my goal was to create a killer system, I would build my system around my speakers since they will likely represent the largest cost. Audition some speakers (including subs) and equalize them for your room. Next, audition several DACs and pick one that you like most. All other equipment can be swapped out for incremental gains, if any, or features and this is when measurements may help support your hearing to guide you in selecting components.

However, things get complicated when you expand from 2.x channels to 5.x.x and up for HT. If HT SQ is important to you (it would be to me) then you might need to audition Pre-pros or transports to find suitable DACs. Most likely, it will be your Pre-pro since it may accept input from physical transports, USB, wireless etc.

It is a journey and you'll never be "done" since perfection can never be achieve, at least I don't foresee it in our lifetime.

[yves]

Oct 4, 2016 9:38:35 GMT -5 garbulky said:

because it's objective though we don't really have any proof for that.

I once met a guy who kept claiming he was being objective. I kept asking him to prove his claim, but replies I got from him were all subjective assumptions.

[yves]

Oct 4, 2016 14:54:03 GMT -5 Wideawake said:

Leaving the room out of the equation, and taking into consideration modern solid state equipment, IMO the two components that will have the largest effect on SQ are the DACs and the speakers.

This is presuming that the component chain in your system comprises of standards based components from reputable manufacturers who know what they are doing. That is, it is assumed that the Preamp provides the appropriate voltage to the amps; amps have a gain between 27dB and 29dB with good S/N and low crosstalk; components measure relatively flat with no crazy peaks or dips; etc. etc.

Given the above, if my goal was to create a killer system, I would build my system around my speakers since they will likely represent the largest cost. Audition some speakers (including subs) and equalize them for your room. Next, audition several DACs and pick one that you like most. All other equipment can be swapped out for incremental gains, if any, or features and this is when measurements may help support your hearing to guide you in selecting components.

However, things get complicated when you expand from 2.x channels to 5.x.x and up for HT. If HT SQ is important to you (it would be to me) then you might need to audition Pre-pros or transports to find suitable DACs. Most likely, it will be your Pre-pro since it may accept input from physical transports, USB, wireless etc.

It is a journey and you'll never be "done" since perfection can never be achieve, at least I don't foresee it in our lifetime.

Well, my XPA-2 Gen 1 has an amp gain of 32dB, which, clearly, isn't between 27dB and 29dB so are you trying to suggest Emotiva engineers didn't know what they were doing?

[lehighvalleyjeff]

Absolutely brilliant post!

For what it's worth, speaking strictly about my experience with the XSP-1 and the DC-1 both being used as preamps in the same exact systems from the same exact source the results were constant. I preferred the XSP-1 by a significant margin versus the DC-1. I respectfully disagree that one is more "resolving" than the other. I also disagree with the notion that if the preamp specs are similar then they should sound "alike". Each one sounds a bit different and regardless of preference I don't think the cause for that is the amount of detail either or both can resolve. Just my $0.02

[melm]

Something to keep in mind. It's really all about the room, all things being equal. Let's try a thought experiment with a small string ensemble or jazz group. All acoustic instruments. We can agree that this is truly 'natural' sound.

Now let them play in two different, acoustically good rooms. Guess what? They will sound different. One may be better, but my take is they will simply sound 'different'. Add a third room. More difference.

In the final analysis, we need to get flat electronics, speakers with low distortion and wide dynamic range. and a decent room to house it all. That really is about the best we can do. You can swap high quality electronics to your hearts content, and you may hear different, but not necessarily better.

Which of the three rooms is best? I guess the one that sounds 'best' to you, but we are being subjective, not scientific.

I used to sell high end audio back in the day (the 70's). I heard some spectacular audio. I'm not sure anything today exceeds what I heard for realism, and the good stuff today is certainly no worse. The good products back then followed the same rules, (low distortion, wide dynamic range) and it sounded really excellent. Almost 'live'.

I feel I have an excellent system. XPA amps, XMC-1, Maggies and a solid subwoofer. Good sources (Oppo, PS Audio DAC, Aurender for HiRez playback.) My room is pretty good, and Dirac smooths it out. I feel pretty satisfied. I played in musical groups (non electronic) for years and every venue sounded different. From the stage, and the audience.

One can chase these variables, but at the end of it all, get the best you can afford, sit back and listen and enjoy. You'll never approach a 'live sound' because there is no standard for that, just differing methodologies. .

Mel

[Wideawake]

Oct 4, 2016 15:18:52 GMT -5 yves said:

Oct 4, 2016 14:54:03 GMT -5 Wideawake said:

This is presuming that the component chain in your system comprises of standards based components from reputable manufacturers who know what they are doing. That is, it is assumed that the Preamp provides the appropriate voltage to the amps; amps have a gain between 27dB and 29dB with good S/N and low crosstalk; components measure relatively flat with no crazy peaks or dips; etc. etc.

Well, my XPA-2 Gen 1 has an amp gain of 32dB, which, clearly, isn't between 27dB and 29dB so are you trying to suggest Emotiva engineers didn't know what they were doing?

I think all their current amp lines have 29dB gain. I believe they had made some 32dB gain amps for the entry-level novice market to accommodate less-than-capable receivers used as Pre-pros that did not supply the required voltage to outboard amps. This was likely an intentional and strategic move.

I use an IPS-1 which has 27dB gain. I think the XPR series may have been at 29dBs or adjustable (not sure). So they had made a mix of products to suit their market.

[repeetavx]

Oct 4, 2016 11:42:32 GMT -5 Wideawake said:

NOTE: I don't know anything about electronics. Just trying to understand what you've stated.

How does adding a volume control to the output of the CD player vary the impedance of the CD player's output circuitry? I would think it would vary the impedance of the input to the receiver, no? The CD player should be unaffected since the volume control is transparent to it. What am I missing?

By adding a resistor between the input and output capacitors of my CD player and receiver I created an RC Filter (you can read about it on the Wikipedia ).