Can We / Should We "Balance" Our Way To Our Desired Sound?

[Ron Resnick]

"neutral Free from coloration."

"coloration An audible "signature" with which a reproducing system imbues all signals passing through it."

-- Sounds Like? An Audio Glossary (J. Gordon Holt, Stereophile, July 1993)

Whether or not a component has a coloration is subjective. Therefore, I think whether or not something is neutral also is subjective.

I agree it should be relatively easier to obtain a consensus on whether a component conveys more or less detail.

 

[DaveC]

Whether or not a component has a coloration is subjective.

That's not true, they are mostly due to easily measurable distortions. While it may be true there are some unknowns most of what you hear wrt coloration is cold, hard science and is easily measured and quantified.

 

[FrantzM]

That's not true, they are mostly due to easily measurable distortions. While it may be true there are some unknowns most of what you hear wrt coloration is cold, hard science and is easily measured and quantified.

+1

If an uninitiated person were to get to this board and read many posts he/she would conclude there is no objective reality to music reproduction that it is simply a matter of taste and everything goes...
They would also notice that Bose is not considered High End by us... Nor the AVRs some people use in HT to great effects ... Same for Polk as speakers goes .. or Axiom ... so ..

There is an objective reality, however much we claim and debate the contrary. There are ways to measure a portion of it not all of it IMO. UNfortunately there the space between what si being measured presently with what is not yet, is filled with the equivalent of noise: Lot of opinions, obfuscation and of misinformation that make our hobby almost unapproachable. For people to understand what our hobby should be about it is a matter of experiencing some really well put systems. Those systems give a glimpse of what is achievable. We, the experienced audiophiles, have come to think a good systems need to cost an arm and a leg and we go on spending appropriately. Most often a great system is the result of patience and care taken in assembling the speakers and what drives them and how those speakers interact with a given room. It is often as simple (and complicated ) as that for surprisingly good, believable sound. Listening recently to a la Klipsch Cornwall driven by a Peach Audio DAC/amp reminded of that.

 

[microstrip]

That's not true, they are mostly due to easily measurable distortions. While it may be true there are some unknowns most of what you hear wrt coloration is cold, hard science and is easily measured and quantified.

Please educate us on these easily measured and quantified distortions and their correlation to sound quality. Ralph Karsten of Atma-sphere has already pointed to some work in this direction, but mostly quantitative comments.

 

[bonzo75]

From what I understand some distortions in the k2, k3, and k4 make the sound more natural, and such distortions are caused by good valve amps as opposed to SS amps. Jadis is supposed to distort to such an extent that it makes the sound seem more natural. Any techies who can confirm this?

 

[morricab]

Within certain frequencies, yes... outside that range you won't hear a difference. Also, phase and time are the same thing.

No, phase and time coherence are not the same thing. A speaker can be phase coherent but have the drivers offset in such a way that the sound waves do not reach your ear at the same time and therefore is not time coherent. In Stereophile, they show the phase shift of the speaker over its frequency range and even if it looks really flat, if the drivers are not aligned properly the step response will not be correct and therefore the speaker will not be time coherent.

 

[the sound of Tao]

That's not true, they are mostly due to easily measurable distortions. While it may be true there are some unknowns most of what you hear wrt coloration is cold, hard science and is easily measured and quantified.

With respect Dave, most but not all is the point here. For anything to be truly neutral requires that it does this across the board. That to have real transparency there is no colouration in any part that could change in any way the original character. That these notions are all holistic and therefore require that we could assess all possible characteristic elements to test for any truth and not by just evaluating some or even many of the individual criteria.

The presence or absence of fine detail that you mentioned earlier as a test that is objective rather than subjective (not sure that this cannot be either type of assessment any way) is just one criteria but of itself does not even begin to cover the range of possible variables a component can imprint upon the character of the original signal.

Even a components capacity to accurately communicate the real nature of the original musical experience is a complex and holistic element that doesn't fit the notion of an objective measure.

We have great measures for many of the parts of assessing a components performance but holistic concepts like neutrality, natural, musical, balanced, realistic all require that we deal with them in the whole and this is where subjectivity becomes an essential way of evaluation.

If real neutrality could be seen as the absence of any modification to the nature and intent as well as the innate character of the original signal or indeed more importantly the original performance of the music we are going to have to get out of the notion that any partial measure can give us any more than a glimpse of the whole story.

 

[DaveC]

No, phase and time coherence are not the same thing. A speaker can be phase coherent but have the drivers offset in such a way that the sound waves do not reach your ear at the same time and therefore is not time coherent. In Stereophile, they show the phase shift of the speaker over its frequency range and even if it looks really flat, if the drivers are not aligned properly the step response will not be correct and therefore the speaker will not be time coherent.

It is the same thing, phase and time are directly related. But it is true a speaker driver can be 360 degrees off and phase aligns but it's still off by 360 degrees... such as combining a full range driver + subwoofer with a 24 dB LR filter, if the drivers are aligned the phase will be off 360 degrees... at those frequencies it really doesn't matter much though.

 

[microstrip]

+1

If an uninitiated person were to get to this board and read many posts he/she would conclude there is no objective reality to music reproduction that it is simply a matter of taste and everything goes...
They would also notice that Bose is not considered High End by us... Nor the AVRs some people use in HT to great effects ... Same for Polk as speakers goes .. or Axiom ... so ..

There is an objective reality, however much we claim and debate the contrary. There are ways to measure a portion of it not all of it IMO. UNfortunately there the space between what si being measured presently with what is not yet, is filled with the equivalent of noise: Lot of opinions, obfuscation and of misinformation that make our hobby almost unapproachable. For people to understand what our hobby should be about it is a matter of experiencing some really well put systems. Those systems give a glimpse of what is achievable. We, the experienced audiophiles, have come to think a good systems need to cost an arm and a leg and we go on spending appropriately. Most often a great system is the result of patience and care taken in assembling the speakers and what drives them and how those speakers interact with a given room. It is often as simple (and complicated ) as that for surprisingly good, believable sound. Listening recently to a la Klipsch Cornwall driven by a Peach Audio DAC/amp reminded of that.

Frantz,

IMHO people come to forums such as WBT mostly for the not ALL. They have learned from experience that the available measurements are insufficient to guide them.

There were many threads on great systems that did not cost an arm and a leg. And yes, we have many great system that are the result of patience and care taken in assembling the speakers and what drives them and how those speakers interact with a given room in WBF. IMHO arriving people must also read old threads, not just the "recent activity".

Each of us presents and debates his favorite subjects - there are many great threads going on besides this one. And, sorry to say, systems that cost an an arm and a leg can have performance matching its cost ...

 

[morricab]

That's not true, they are mostly due to easily measurable distortions. While it may be true there are some unknowns most of what you hear wrt coloration is cold, hard science and is easily measured and quantified.

Yes, you are right that a lot of it IS avaiable in measurments but how those are interpreted with regard to human psychological response is still an area where science is still catching up. Several researchers have looked into the psyhchoacoustics of distortions from electronics. One of the earlier ones was D.E.L. Shorter from the BBC. He came up with a metric where the weighting of importance of a given harmonic was the nsquared/4 where n is the harmonic order number.

Cheever later came up with his T.A.D (Total Aural Dissonance) and Geddes came up with the GedLee number.

The upshot of all of these attempts to quantify psychological preference all come to basically the same conclusion: High order harmonics are very destructive to realisitic reproduction of music...even at extremely low levels...levels that many claim are inaudible.

What Crowhurst found out was that the application of negative feedback contributes to the production of high order harmonics and in such a proliferation that the result is a signal modulate "noise" floor that also destroys realisitic reproduction.

Jena Hiraga came to the conclusion that the most natural distortion pattern in an amplifier was monotonic...an exponential decay from low order to high order harmonics.

Cheever found basically the same thing as the low order harmonics are largely masked by the ear/brains own distortion and tone masking. Howver, he went even further to determine that this is not a static thing and that the correct ratios of harmonics is actually SPL dependent, so the interaction of the amp and speaker therefore becomes important so that the amplifier is working in a range that gives the most correct harmonic pattern at a pleasant listening level.

Just to dispel the notion of euphonic distortions as claimed by the "neutral" crowd about tube and SET afficianados. There was an experiment published by Keith Howard in Stereophile where he added distortion to music files. He found that adding any distortion to the original file degraded the sound. This dispels the "Euphonic" claim that people say about high 2nd order harmonics etc. The important side point from this experiment is that he also found that the "Hiraga" monotonic pattern was the least offensive of the distortion patterns he explored.

So, where does that leave us? There are clear preferences for electronics which will minimize the most destructive harmonic distortion. It is no wonder to me that the Lampizator DAC has taken the audio world by storm...using the most linear amplification devices is a good way to get a more truly neutral reproduction with low and low order harmonic distortion (unlike an amp a line level tube amplifier can have very low distortion without any negative feedback and it will then be primarily low order harmonics as well). Most output stages of DACs are simply ruining the sound in the analog output stage.

I am a objectivist who has used the science that is out there to direct my subjective choices. I still ultimately decide by listening because even among gear that has more or less proper designing there are still plenty of audible differences that can lead to preferrence.

What I can clearly hear now as well though is how electronics that haven't taken psychoacoustics into account (most do not but get kind of "lucky" and do it more or less correctly) get it really wrong and it doesn't take long with careful listening to find out that is the case. It has reduced my "mistake" buying drastically.

 

[DaveC]

The presence or absence of fine detail that you mentioned earlier as a test that is objective rather than subjective (not sure that this cannot be either type of assessment any way) is just one criteria but of itself does not even begin to cover the range of possible variables a component can imprint upon the character of the original signal.

I never said it did, but I am saying that listening for fine detail is a good test to determine neutrality as warmth smooths over detail and harshness hides it. The further you go from neutral the less resolving the component will be.

Earlier, I did question whether some kinds of distortion makes tone/timbre more realistic (because it does), and nobody answered... I asked this because this is one area that is quite subjective... so there is definitely some amount of subjectivity, quite a bit in fact, but it is also true there is quite a bit of science and measurement behind the distortions we hear as well, and the claim this is all subjective is simply not true.

One example is harmonic distortion and the relative levels of the orders of distortion. It's very well known that even and odd order distortion sound very different and higher modes are more objectionable vs 2nd and 3rd... In fact they have been scored and weighted long ago, with higher order distortions getting much more weighting. Nobody actually uses this (in audio), but it's been thought about, measured and quantified long, long ago and is used in other industries.

 

[NorthStar]

I've read the thread since the beginning, three days ago, when Ron started it.

Of course we try to balance the sound to our ears in our rooms. I agree that we all have our own journey. Some members here are more advanced in that journey than others.
And other members are even more advanced than the members posting; they don't post...they remain quiet and only read. Plus they use some great tips too.

What can I contribute positively that wasn't already?
I won't post a youtube music video (I could though), and I won't provide a link (or maybe).

'Balance' is a great word in audio; balance of everything that matters...the artist/singer/musician's performances, the quality music recordings by the master recording/mixing professionals, their recording techniques, their studio's acoustics, their recording gear (analog, digital, microphones, ...), and us, the main listeners/music lovers sitting comfortably in our rooms with our gear and respect for everything we love in life.
That's the balance to me. ...Simplicity.
_______

Bonus :b : http://lifehacker.com/5883665/how-to-be-an-audiophile-on-the-cheap

 

[microstrip]

From what I understand some distortions in the k2, k3, and k4 make the sound more natural, and such distortions are caused by good valve amps as opposed to SS amps. Jadis is supposed to distort to such an extent that it makes the sound seem more natural. Any techies who can confirm this?

Distortion and distortion and noise of a Jadis JPL preamplifier taken two days ago. Except for the 50Hz peak, all components are bellow -100 dB. Should we consider that it can't sound natural? BTW, unfortunately the ECC83 tubes have no marking at all - it is one of less noisier ECC83 I have ever measured.

 

[morricab]

Frantz,

IMHO people come to forums such as WBT mostly for the not ALL. They have learned from experience that the available measurements are insufficient to guide them.

There were many threads on great systems that did not cost an arm and a leg. And yes, we have many great system that are the result of patience and care taken in assembling the speakers and what drives them and how those speakers interact with a given room in WBF. IMHO arriving people must also read old threads, not just the "recent activity".

Each of us presents and debates his favorite subjects - there are many great threads going on besides this one. And, sorry to say, systems that cost an an arm and a leg can have performance matching its cost ...

Actually most people are simply lacking the interest or ability to understand what the data that is available is telling them. This goes for engineers and designers as well.

 

[morricab]

I never said it did, but I am saying that listening for fine detail is a good test to determine neutrality as warmth smooths over detail and harshness hides it. The further you go from neutral the less resolving the component will be.

Earlier, I did question whether some kinds of distortion makes tone/timbre more realistic (because it does), and nobody answered... I asked this because this is one area that is quite subjective... so there is definitely some amount of subjectivity, quite a bit in fact, but it is also true there is quite a bit of science and measurement behind the distortions we hear as well, and the claim this is all subjective is simply not true.

One example is harmonic distortion and the relative levels of the orders of distortion. It's very well known that even and odd order distortion sound very different and higher modes are more objectionable vs 2nd and 3rd... In fact they have been scored and weighted long ago, with higher order distortions getting much more weighting. Nobody actually uses this (in audio), but it's been thought about, measured and quantified long, long ago and is used in other industries.

As a consumer, you can apply this research to narrow choices as long as measurements are available for the FFT at 1Khz, distortion vs. level, distortion vs. frequency and the IM distortion harmonics. You can potentially run these numbers through an old model, which is easy but limited (Shorter) or a newer one that is probably a lot more accurate but probably difficult to implement (Cheever and Geddes). There are enough designers out there with enough products that you can use data to "sift" out the ones that would fall short psychoacoustically.

Not everyone is analytically equipped to take this approach but I suspect it would not be a problem for you. Put your own gear to the test and see where it would fall. Maybe you are in a good way or maybe you should reconsider your choices...

 

[DaveC]

Yes, you are right that a lot of it IS avaiable in measurments but how those are interpreted with regard to human psychological response is still an area where science is still catching up. Several researchers have looked into the psyhchoacoustics of distortions from electronics. One of the earlier ones was D.E.L. Shorter from the BBC. He came up with a metric where the weighting of importance of a given harmonic was the nsquared/4 where n is the harmonic order number.

Cheever later came up with his T.A.D (Total Aural Dissonance) and Geddes came up with the GedLee number.

The upshot of all of these attempts to quantify psychological preference all come to basically the same conclusion: High order harmonics are very destructive to realisitic reproduction of music...even at extremely low levels...levels that many claim are inaudible.

What Crowhurst found out was that the application of negative feedback contributes to the production of high order harmonics and in such a proliferation that the result is a signal modulate "noise" floor that also destroys realisitic reproduction.

Jena Hiraga came to the conclusion that the most natural distortion pattern in an amplifier was monotonic...an exponential decay from low order to high order harmonics.

Cheever found basically the same thing as the low order harmonics are largely masked by the ear/brains own distortion and tone masking. Howver, he went even further to determine that this is not a static thing and that the correct ratios of harmonics is actually SPL dependent, so the interaction of the amp and speaker therefore becomes important so that the amplifier is working in a range that gives the most correct harmonic pattern at a pleasant listening level.

Just to dispel the notion of euphonic distortions as claimed by the "neutral" crowd about tube and SET afficianados. There was an experiment published by Keith Howard in Stereophile where he added distortion to music files. He found that adding any distortion to the original file degraded the sound. This dispels the "Euphonic" claim that people say about high 2nd order harmonics etc. The important side point from this experiment is that he also found that the "Hiraga" monotonic pattern was the least offensive of the distortion patterns he explored.

So, where does that leave us? There are clear preferences for electronics which will minimize the most destructive harmonic distortion. It is no wonder to me that the Lampizator DAC has taken the audio world by storm...using the most linear amplification devices is a good way to get a more truly neutral reproduction with low and low order harmonic distortion (unlike an amp a line level tube amplifier can have very low distortion without any negative feedback and it will then be primarily low order harmonics as well). Most output stages of DACs are simply ruining the sound in the analog output stage.

I am a objectivist who has used the science that is out there to direct my subjective choices. I still ultimately decide by listening because even among gear that has more or less proper designing there are still plenty of audible differences that can lead to preferrence.

What I can clearly hear now as well though is how electronics that haven't taken psychoacoustics into account (most do not but get kind of "lucky" and do it more or less correctly) get it really wrong and it doesn't take long with careful listening to find out that is the case. It has reduced my "mistake" buying drastically.

Great post, I agree 100%.

There has been a lot of work to measure and quantify the distortions we hear, so it's not at all subjective... but there are unknowns for sure, things we don't yet understand that make it seem as if it's just subjective/personal preference.

I also think understanding psychoacoustics is one of the most important things an audio designer can do, otherwise how do you assign priorities? I know some just go by numbers and this works, especially for studio monitors which are meant to do a certain job... but it doesn't make for the best sounding gear imo. Otherwise nobody would like some of the speaker designs that don't necessarily measure well, like single drivers and panels, but many do despite the technical flaws.

 

[morricab]

Distortion and distortion and noise of a Jadis JPL preamplifier taken two days ago. Except for the 50Hz peak, all components are bellow -100 dB. Should we consider that it can't sound natural? BTW, unfortunately the ECC83 tubes have no marking at all - it is one of less noisier ECC83 I have ever measured.

Can you measure the IM distortion as well? What about Distortion vs level? You should get the FFT at many different levels to see if it stays with the same relative harmonic pattern. However, seeing the third harmonic a bit higher than the 2nd possibly have an audible consequence. There is a nice lack of high order components but that is why it would be nice to see higher levels to see if the grow in.

 

[bonzo75]

Distortion and distortion and noise of a Jadis JPL preamplifier taken two days ago. Except for the 50Hz peak, all components are bellow -100 dB. Should we consider that it can't sound natural? BTW, unfortunately the ECC83 tubes have no marking at all - it is one of less noisier ECC83 I have ever measured.

Thanks I was referring to their power amps. I am Jadis fan, btw. I still have a JPL. Their JPS2 is outstanding.

 

[DaveC]

As a consumer, you can apply this research to narrow choices as long as measurements are available for the FFT at 1Khz, distortion vs. level, distortion vs. frequency and the IM distortion harmonics. You can potentially run these numbers through an old model, which is easy but limited (Shorter) or a newer one that is probably a lot more accurate but probably difficult to implement (Cheever and Geddes). There are enough designers out there with enough products that you can use data to "sift" out the ones that would fall short psychoacoustically.

Not everyone is analytically equipped to take this approach but I suspect it would not be a problem for you. Put your own gear to the test and see where it would fall. Maybe you are in a good way or maybe you should reconsider your choices...

My system consists of gear that I built myself, except for the source, and it has all been measured.

Personally, I do prefer some distortion in the form of tubes, my speakers are extremely efficient at about 105 dB and only uses 1st order xos in exchange for some minor dips in FR in the high treble so I'm no purist...

 

[microstrip]

I never said it did, but I am saying that listening for fine detail is a good test to determine neutrality as warmth smooths over detail and harshness hides it. The further you go from neutral the less resolving the component will be.

Earlier, I did question whether some kinds of distortion makes tone/timbre more realistic (because it does), and nobody answered... I asked this because this is one area that is quite subjective... so there is definitely some amount of subjectivity, quite a bit in fact, but it is also true there is quite a bit of science and measurement behind the distortions we hear as well, and the claim this is all subjective is simply not true.

One example is harmonic distortion and the relative levels of the orders of distortion. It's very well known that even and odd order distortion sound very different and higher modes are more objectionable vs 2nd and 3rd... In fact they have been scored and weighted long ago, with higher order distortions getting much more weighting. Nobody actually uses this (in audio), but it's been thought about, measured and quantified long, long ago and is used in other industries.

One of the reasons almost nobody actually uses it is because the values of the distortions of current equipment are well bellow the values that are referred in these studies. I said almost because a few manufacturers still use them. There are many other interesting scientific studies on coloration - thermal distortion effects, also called memory effects, the Vladimir Lamm models , some specific form of distortion, the Soulution model, just to refer a few. Surely all this people used instruments and measurements, but as they are proprietary and not public, we can not class them as confirmed science.

The point is that every time we are faced with direct questions asking for quantitative values and correlation with subjective sound quality we only get evasive answers with old arguments or the classical case of SS clipping versus tube clipping. Or someone complaining WBF is against science ...