Why we believe

[damonrpayne]

I found another Ethan Winer article I like:

http://www.ethanwiner.com/believe.html

[DrWho]

Great article [Y]

I loved his cassette tape analogy at the end too...someone needs to show Ethan how to do it without introducing tape hiss. []

[damonrpayne]

Remember my post in 2ch where I said that's why everyone likes tubes? It "colors" the sound in a way that's pleasing. That's what Bose does too [], pumping up the midrange and ignoring the extremes.

[DrWho]

You so should have posted this in the 2-Channel forum []

[Arky]

You so should have posted this in the 2-Channel forum []

Yeah the spkr wire thread is running out of steam.[6]

[gaspr]

So when you are listening to music, and you move your head from side to

side, you should be able to here these anomalies, no?? I tried

but for the life of me, I can't discern any differences at all. I

think maybe Parrot is right in that our brains are much better than we

them credit for at analyzing and filtering out these room induced

glitches?? Someone (John Warren?) on another thread was talking

about some software that can do this, why can't our brains do it

as well or even better??

[mas]

So when you are listening to music, and you move your head from side to
side, you should be able to here these anomalies, no?? I tried
but for the life of me, I can't discern any differences at all. I
think maybe Parrot is right in that our brains are much better than we
them credit for at analyzing and filtering out these room induced
glitches?? Someone (John Warren?) on another thread was talking
about some software that can do this, why can't our brains do it
as well or even better??

No - it is not that simple.

Comb filters less than 1/3 octave in width are 'smoothed' out by the ear/brain. Additionally, you need to understand the relationship of the time component to the comb filter and the associated polar lobing that accompanies comb filtering...You can't separate them - they are all aspects (simply different viewpoints) of the same phenomena. Also, an understanding of the various aspects of the Henry precedence/Haas psychoacoustic effects will be of benefit.

You might take a look at the last several post in the small room acoustics thread regarding time and frequency relationships.

But be careful, you might start to appreciate the value of the time domain perspective! [6]

Yeah the spkr wire thread is running out of steam.[6]

...running out of steam?

Give it a little more credit!...It was dead before it began! []

[Arky]

You are so right mas & for some reason the topic is continually resurected.

JT

[WMcD]

It is a very illuminating article. But I believe his logic is flawed. More on that below.

It does show just how difficult it is to reach a conclusion about the performance of speaker plus a room. Every location of measurement produces very different result. No "one" is more true than another, some are bad or good.

There are averaging techniques.

PWK ran many curves in a room on the same chart paper, and then took tops on the many ink traces. Maybe that is as valid as any.

- - - -

Ethan's argument seems to be that someone starts with Tweek Zero (no tweek) and there is a sound curve for that ear location. Then they make a Tweek One in the hardware. Then they sit down in a slightly different location and "hear" a change.

He says, Tweek One DOES sound different BUT it is because the location is different and the curve has changed, even if altering from Tweek Zero to Tweek One in hardware has accomplished nothing significant in the hardware sending out the signal.

He argues that hardware changes are co-incidental to different listening locations. And listening location changes are the real cause.

Note how very close this argument is to the fact that moving the measuring microphone by a few inches or feet makes a good room and speaker combination into a bad one; when in fact, little has changed when we move our ears a few inches or feet.

- - -

Ethan's approach is lacking in contol groups in many ways. He is certainly correct, in my view, about self deception, though.

First. Per his thesis, sitting down after installing the tweek results in a REAL perceived acoustic change (resuling not actually hardware, just position). Fine. How come this slight alteration in listening position was not perceived before the Tweek? We've been experiencing that non-Tweek (one inch alteration) situation for years, Yet we didn't hear it before.

To belabor the point. The Tweeker ACTUALLY perceived something different after the Tweek, per Ethan. But if Ethan's curves are correct, it has always been going on. It is as if Ethan is saying the Tweeker has misinterpreted the Tweek as part of of the constantly changing scenry. But how come the Tweeker failed to see the constantly changing scenry before? Did the listener suddenly become acute to room effects?

He seems to argue that the Tweek is equivalant to, say, two inches of head displacement, and is preceived. But, in fact, we don't perceive two inches of head movement for year before.

- - -

Second. This is more my argument. Just about all Tweeks are reported as being positive. People will say, of course they are. "Better caps" always perform better, all better amps perform better. I don't believe that. But let us assume that.

But Ethan seems to be arguing that perceptions of "better" hardware is misperceived as better sound curves because the listener sat down in a better spot. I have to ask, how did the Tweeker accomplish that? Realistically, half of his sound curves are worse than the others. Which ones are the starting point of our considerations? Four inches off the sweet spot may well be the other guy's sweet spot.

I may be over interpreting Ethan's thesis and I don't mean to be too harsh.. He may be saying that it is only "different". Still, he seems to argue that hardware improvements and sitting down have some co relation. But arguably, it goes the other way, every day.

Smile,

Gil

.

[sredmyer]

Ethan's argument seems to be that someone starts with Tweek Zero (no tweek) and there is a sound curve for that ear location. Then they make a Tweek One in the hardware. Then they sit down in a slightly different location and "hear" a change.

He says, Tweek One DOES sound different BUT it is because the location is different and the curve has changed, even if altering from Tweek Zero to Tweek One in hardware has accomplished nothing significant in the hardware sending out the signal.

He argues that hardware changes are co-incidental to different listening locations. And listening location changes are the real cause.

Note how very close this argument is to the fact that moving the measuring microphone by a few inches or feet makes a good room and speaker combination into a bad one; when in fact, little has changed when we move our ears a few inches or feet.

- - -

Ethan's approach is lacking in contol groups in many ways. He is certainly correct, in my view, about self deception, though.

First. Per his thesis, sitting down after installing the tweek results in a REAL perceived acoustic change (resuling not actually hardware, just position). Fine. How come this slight alteration in listening position was not perceived before the Tweek? We've been experiencing that non-Tweek (one inch alteration) situation for years, Yet we didn't hear it before.

To belabor the point. The Tweeker ACTUALLY perceived something different after the Tweek, per Ethan. But if Ethan's curves are correct, it has always been going on. It is as if Ethan is saying the Tweeker has misinterpreted the Tweek as part of of the constantly changing scenry. But how come the Tweeker failed to see the constantly changing scenry before? Did the listener suddenly become acute to room effects?

He seems to argue that the Tweek is equivalant to, say, two inches of head displacement, and is preceived. But, in fact, we don't perceive two inches of head movement for year before.

- - -

Second. This is more my argument. Just about all Tweeks are reported as being positive. People will say, of course they are. "Better caps" always perform better, all better amps perform better. I don't believe that. But let us assume that.

But Ethan seems to be arguing that perceptions of "better" hardware is misperceived as better sound curves because the listener sat down in a better spot. I have to ask, how did the Tweeker accomplish that? Realistically, half of his sound curves are worse than the others. Which ones are the starting point of our considerations? Four inches off the sweet spot may well be the other guy's sweet spot.

I may be over interpreting Ethan's thesis and I don't mean to be too harsh.. He may be saying that it is only "different". Still, he seems to argue that hardware improvements and sitting down have some co relation. But arguably, it goes the other way, every day.

Smile,

Gil

I think that we all listen much more criticaly after tweaking. We are trying to decern the impact of the tweak. Normal listening does not (usually) carry such a requirement...we are not looking (listening) for a change so we do not hear one.

[mas]

A slightly different take.

I agree with the Gil and the idea that expectations play a huge role in one's conclusions.

But one thing that I am struck by, and I hope no one is offended by, is the extent to which so many are enamored with the frequency response!

Why do I say this? Most still do not understand the fundamental relationships between multisource comb-filtering and the concurrent change in polar response (lobing)! Many still treat the frequency response as some static monolithic measurement! The result is that many are shocked to see it vary.

My suggestion is to please take that shock and amazement and use that as an impetus to learn more about the relationship of the views and affects!

The change in much of the frequency response in the mid-upper frequencies is due in large measure precisely to the variation in polar lobing! And a simple thing to remember is that as the frequency increases, the lobes become more numerous and narrower (higher Q). And we haven't even begun to address the affect of relative source gain on comb filter Q! And yes, for those who have been familiar with this, the subtle changes can be recognized - within the limits of one's hearing capabilities.

An area where this has been addressed for years - with varying degrees of success, but with an increasing awareness with time - has been in the pro sound area and the use of EQ. I know quite a few folks who are literally walking RTAs. They can literally identify a frequency anomaly with uncanny accuracy. Traditionally, the result would be to run for the EQ. That topic of debate was popular during the late 1980s and early 1990s. In that market the use of EQ to address superposition anomalies has largely died out. But not in home audio use! It is amazing what one can learn by measuring the acoustical effects of 2 sources with one modified with a delay unit and gain control.

Are many of the effects real that many here dismiss? Yes. Are most folks aware of them in everyday experience? NO. By moving ones head, I suspect that many are expecting a radical change - it 'ain't' going to happen! Can one train themselves to become aware of the very subtle differences within the limits of human hearing? Yes. Does one need to? No.

But here many are still amazed to see these frequency response variations, and as a result, react in a manner far out of proportion (both more and less!) to their their causes and significance. (Again, please take a glance at the Time&FrequencyMeasurement post attachment the Small Room Acoustics thread. You will see two examples of radical comb filtering - each with radically different effects.) Unfortunately, instead of digging deeper and looking for more insight, the most common response is to run to find a bigger more expensive RTA! Oh well...(sung to the tune of the same name by a band most will not recognize called Fleetwood Mac - nope, NOT the band with SpaceCadet Nicks or Buckingham! ;-)

Some may even be amazed to hear me suggest that you put the @#%$* frequency response plots away. ...if for no other reason than the source and causes for the anomalies you 'see' in the frequency domain are obscured! There are better ways of looking at them that provide infinitely more useful information.

Oh, don't get me wrong... A frequency response can provide useful information, but not to the degree that many worship them. Ironically, most of the useful information is often to point to needing additional time based examination. And continuing to worship them simply delays one's becoming aware of other perspectives that are much more illuminating and useful.

And as a point of reference, we can assume that 'stuff' has contributed to the frequency response anomalies - but does anyone know exactly what the specific 'stuff' is that did so? I can guess and make assumptions rather accurately before any frequency measurements were taken. So I would go further and say...sure something is happening (of which I can provide a list of prime suspects), but we still are no closer to knowing precisely what!

He didn't use the stimulus of frequency plot anomallies to actually look and see what the actual causal factors are! The problem is that it stops prior to any real understanding! He is making a very general observation. It can explain that there may be variances in hearing (and I guess that is sufficient for his point), but little else. In other words, based upon an awareness of behavior in the time domain, I already know exactly what he states regarding the frequency response with regards to any conventional room at any time.

Think...when was the last time someone took the frequency response and asked just what contributed to a peak or a null in the response? And when was the last time anyone actually identified the source(s) of them? But most look at them and think that this is the actual response of a driver or of several drivers as if nothing else was occuring. Critical issues such as the synchronization of signals, superposition due to diffractive effects (at all levels), reflections due to environment, etc. etc. etc. are simply ignored - after all, how is one supposed to figure out that complex can of worms!?

The answer is that it is not difficult to do - given the right tools and the willingness to acknowledge their existence!! After all, ignore-ance of the factors sure seems to minimize their importance.

Has anyone wondered why those who actually examine rooms and systems spend so little time with frequency plots and spend so much more time with plots such as ETC?

For a start - as there is no way sufficient information can be shared in just a few posts, may I suggest:

Everyone needs to get the NEW 3rd edition to Sound System Engineering by Davis & Patronis. The added chapters on Signal Delay and Synchronization, Signal Processing, and especially Sound System Equalization are worth the price of the book alone, and IMO should be required reading before posting! Much of what often becomes the subject for fruitless argument is addressed here. And with an understanding of some of the fundamentals, it is much more interesting and fun to pursue more interesting topics.

And folks need to experience such a fundamental experiment as the one mentioned on page 334, fig 13-2, in which a varying offset signal provides a dramatic illustration of the dynamic nature of polar lobing and comb filtering. I first experienced this experiment in 1987 with Don, and it still ranks as perhaps the most simple and yet dramatic demonstration of the effects in this subject area.

And among the references to superposition and signal synchronization include some made as a result of hauling the complete Morningstar Sound SR rig used by Dan Seals up to Don and Carolyn's farm in Indiana - at least one of which are mentioned specifically in the book (e.g., p.420, fig.15-33). Doc might appreciate this, as dealing with the residual effects of summed reflections from hats and guitars on live sound can prove to be an unexpected challenge....But I wax nostalgic...as it is neat to be able to relate to so many of the events that subsequently led to some of the analysis and awareness of so much of what is now accepted and which is mentioned in the text. And for far more anecdotal history (including a lot more on PWK as well, the If Bad Sound Were Fatal, Audio Would be the Leading Cause of Death is hard to beat and a joy to read.

We are still 'suffering' from an large gulf in the awareness of technologies and the underlying science that makes it hard to discuss many subjects in a manner where everyone can contribute. Some are very aware, and others less so. Aside from some crazy workload requirements lately, it is this gulf that has complicated my plans to finish the small room acoustics thread - as so many are in so many different places - but I promise (for the 1 millionth time) to address this again in several weeks with a mid-level approach sure to bore some and lose others. ;-) At least we will introduce some to various fundamental aspects that may help them to pursue the subject further. From there each is encouraged to pursue more information in any of the aspects. But an understanding of the causes and effects of superposition, comb filtering and polar lobing and their relationship to the common frequency response and audibility are critical.

For now, in response to this post I suspect that there will be those who run out and object to my relegating the frequency response to such a low level. I would politely reply that these 'flatlanders' ;-) need to expand their horizons and to make the effort to learn of the complex factors that underly what they are seeing in the response, and to learn more about how it translates into what they actually hear. I think they will have many surprises! - and their understanding will benefit greatly. And I think that those who do will actually find it quite exciting!