The 2011 Richard C. Heyser Memorial Lecture: "Where Did the Negative Frequencies Go?" Case Study 2: Loudspeakers

Case Study 2: Loudspeakers
During a visit to Canada's National Research Council many years ago, I noticed, stuck to the wall of the prototype IEC listening room, a page of results from one of Floyd Toole's seminal papers on the blind testing of loudspeakers. The scoring system was the one that Floyd developed and that I subsequently used for blind tests at Stereophile: "0" represents the worst sound that could possibly exist, "10" the perfection of live sound. On this scale, a telephone, for example, rates a "2." The speakers in Floyd's test pretty much covered the range of possible performance, yet their normalized scoring spread, from the worst to the best, was just 1.9 points. Other than some pathological designs, the audiophile speakers I test for Stereophile probably cover a range of 0.5 point on Floyd's scale; but, as I pointed out in my 1997 AES convention paper, our reviewers' ratings generally follow Floyd's findings: that people tended to prefer speakers with flat response and controlled dispersion.

But not always.

There are vocal advocates of high-sensitivity horn speakers. There are equally vocal advocates of low-sensitivity, large-panel speakers. The horn advocates, among whom was Stereophile's founder, the late J. Gordon Holt, enthuse about these speakers' "jump factor." I used to ask Gordon what he meant by that: was it dynamic range? No, it wasn't. Was it low distortion? No. His reply was always that there was something about the presentation of these speakers that more closely resembled what he experienced from live classical music.

What about panel speakers? Being dipoles, their interaction with the room is very different, so maybe that's why their advocates love them, not even considering the fact that, with a large panel, the listener is almost never far enough away from the speaker to be in the farfield, the result being more bass—which everyone likes, at least in two-channel audio. Of course, there are no cabinet resonances because there is no cabinet, and perhaps people's preference is due just to the fact that he very large radiating area inherently ensures more linear behavior when moving the same mass of air as a much smaller cone driver. But panel speakers' impulse responses look terrible. Well, I say "impulse response"; what I actually publish and use for my reviews is not the response of a loudspeaker to an impulse, but the impulse response calculated by cross-correlating an MLS signal or a chirp signal, based on the loudspeaker being a perfectly linear system. (John Vanderkooy has investigated the shortcomings of this assumption.)

Perhaps, then, the problem with measuring panel speakers is that the calculated impulse response does not accurately reflect the behavior of a large panel in the same manner as it does a conventional speaker's. At the 1989 AES Convention in New York, mathematician Manfred Schroeder discussed Chaotic systems in audio and mentioned, in passing, that a Chaotic system tended to produce subharmonics. And indeed, when I have measured panel speakers I have found that they produce subharmonics. You feed the panel speaker a 1kHz tone and in its output you find a 500Hz tone as well as the usual distortion harmonics at 2, 3, 4kHz, etc.

So in a large panel speaker, while the average position of the diaphragm follows the driving signal, the motions of the individual elements of that diaphragm are Chaotic. The speaker sounds much better than its measured behavior suggests, as all that randomness is integrated by human hearing's latency.


Live vs recorded; Steinway D vs Vivid GiyaG1 speakers

Or do people like the sound of panel speakers because the intensity of the sound they produce—the sound power per unit area of the radiating surface—is closer to what people hear live? This question was triggered by a live vs recorded demo in which I took part in 2009, in Maryland. Two pianists performed for an hour on a well-prepared Steinway D grand piano, which I recorded at 24/96. Then, after a break, the audience heard the recital again, this time played back through a system. (This meant miking the piano far more closely than I would do for a conventional recording, to avoid a double dose of the room acoustic when the recording was played back.)

The system was full-range, with low coloration and sufficient dynamic-range capability that I was able to match sound-pressure levels. But even though the SPL and the sound of the piano matched closely, listeners felt that the reproduced sound just didn't have the "bigness" of the real thing. I wasn't sure what they meant by "bigness." Yes, the piano's dispersion would excite the room very differently than would a pair of speakers, but perhaps these listeners were suggesting that they were aware that the intensity of the sound was not correct, even when the SPL and response were correct.

So when people say they like panel speakers that measure poorly, perhaps they're responding to those speakers' more accurate intensity, at least when it comes to large sources.

Food for thought?

Share | |
Comments
JohnnyR's picture
Misguided?

I haven't seen one iota of explanation from yourself yet as to why both of us and Harman Kardon and the other links George posted to are wrong. Still waiting ChrisSy.

ChrisS's picture
Georgie & JRusskie = Harman Kardon?

Has HK hired you guys as DBT consultants?

 

Hey JRusskie,

Can you answer this one?

If A=B and C=B, then A=?

If you pass the test, then perhaps someone will hire you... But you and Georgie might have to fight over the job.

JohnnyR's picture
Duhhhhhhhhhh........

"i think THESE are the sort of differences between individuals that make DBT difficult: everyone hears differently. there is no absolute sound."

The sole purpose of DBT is to see if the person listening can distinguish between A and B. If they can't then for all practical purposes there is no difference in the sound from A and B. You and ChrisSy seem to think it's all about what the person "likes". It's a straight forward test method and "likes" has nothing to do with it.

Please explain to us all how Harman Kardon manages to use DBT all the time and do it well? I will be awaiting your reply Ariel.

ChrisS's picture
Twas the night...

So every household has a Harman Kardon product? And you and Georgie have living rooms that look like anechoic chambers? No fireplaces, of course....

ChrisS's picture
And, and...

Hearing a difference between a Harman Kardon product and another product in a anechoic chamber means what to you, Georgie and JRusskie?

Do you know that Ford makes the best trucks in the world?

ChrisS's picture
That's the Pepsi....No, it's...

Are you sure I didn't say 'licks". You know maybe tasting an audio product will yield just as useful results in a DBT.

ChrisS's picture
It's So Easy, Even A 3 Year Old...

JRusskie,

Now run out to your nearest Boris' Convenience store and get yourself a can each of Pepsi (do you even have Pepsi in the Former-USSR?) and Coca-Cola and set up your own Pepsi (or whatever passes for cola in Russia) Challenge.

Wiki has a nice explanation of how to do a DBT...

Once you've done your very own Peps(k)i Challenge, please send us your conclusion. We're curious...

The next step now is to get everyone in your subsidized housing project to participate in your Pepski Challenge.

Gather up that data, compare it your own conclusion and let us know how useful that information is.

I'm sure you'll enjoy the challenge of your doing your very own DBT's! (You won't even have to ask Harman Kardon to use their anechoic chamber!)

John Atkinson's picture
Some knowledge of statistics required

JohnnyR wrote:
The sole purpose of DBT is to see if the person listening can distinguish between A and B. If they can't then for all practical purposes there is no difference in the sound from A and B.

And that's the problem with these tests. If a formal blind test gives results that are indistinguishable from what would be given by chance, formal statistical analysis tells us that this result does _not_ "prove" there was no difference in the stimulus being tested, only that if there _was_ a difference, it was _not_ detectable under the conditions of the test. No more general conclusion can be drawn from the results. And as I have said, it is very difficult to arrange so that those conditions don't themselves become interfering variables. Even the fact that it is a test at all can be an interfering variable, as I explain in this lecture preprint.

JohnnyR wrote:
Please explain to us all how Harman Kardon manages to use DBT all the time and do it well?

I have visited Harman's facility in Northridge and their blind testing set-up is impressive. They have worked hard to eliminate interfering variables and their testing is time- and resource-consuming and painstaking. Even so, they have to make compromises. Blind testing of loudspeakers, for example, is almots always performed in mono. And despite the rigor of their testing, you still have anomalous results, like the Mark Levinson No.53 amplifier, which was designed with such testing but fared poorly in the Stereophile review.

While formal blind tests are prone to false negative results - not detecting a difference when one exists - sighted listening is prone to false positives, ie, it detects a difference when none exists or perhaps exaggerates the degree of difference. As neither methodology is perfect, we go with the one that is manageable with our resources. We therefore offer our opinions for readers to reject or accept in the context of their own experience and I believe Stereophile does  a better job of that than any other review magazine or webzine.

If you are uncomfortable with that policy, then you should not read the magazine. And if I remember correctly, JohnnyR, you admitted in earlier discussions on this sute that you neither subscribe to Stereophile, nor do you buy the magazine on the newsstand. So why should anyone pay attention to your opinions on how the magazine conducts itself?

John Atkinson

Editor, Stereophile

JohnnyR's picture
Silly Boy

So you are saying that if in a DBT the listeners could NOT tell a difference between two amps using music of their own choice then that doesn't prove the amps sound alike?  Funny stuff there Atkinson. For one who thinks you should trust your ears to evaluate components you just bashed the ONE single TRUE way to test by USING YOUR OWN EARS in a DBT.

"And that's the problem with these tests. If a formal blind test gives results that are indistinguishable from what would be given by chance, formal statistical analysis tells us that this result does _not_ "prove" there was no difference in the stimulus being tested, only that if there _was_ a difference, it was _not_ detectable under the conditions of the test."

Your own words are saying that "if there was a difference it was not detectable under the conditions of the test"........oh you mean like letting the listener use the music of their own choice and switch back and forth endless times between two amps and then guess wrongly enough times so that they can't tell which one was which? LMAO if that's not proof that both amps sound alike then what sort of test WOULD prove that they do?  Come on Atkinson you just don't like DBTs because they would show up so many components that people think sound "oh so better than the rest"

Opinions from you and your reviewers are the Gospel now folks. No need to test anything really just trust good ol'JA and his flunkies. Yay.

"If you are uncomfortable with that policy, then you should not read the magazine. And if I remember correctly, JohnnyR, you admitted in earlier discussions on this sute that you neither subscribe to Stereophile, nor do you buy the magazine on the newsstand. So why should anyone pay attention to your opinions on how the magazine conducts itself?"

Oh just maybe because  a lot of people care for this little thing called the TRUTH? When magazines like your's take liberties with the truth by having shoddy reviews instead of in depth testing, then it's everyone's and anyone's responsibility to speak up when crappy falsehoods are published and the readers are supposed to take it all on faith. That's why. I for one do not take your opinions on anything audio related as worthwhile at all for the simple reasons that you show so much promise when you measure speakers but fail to even bother with the snakeoli products that you let slide under the radar yet let your reviewers give them glowing reviews sans any testing what so ever. Maybe that's the sighted listening bias you just spoke about yet you fail to even try with those type of products to get to the real TRUTH.

ChrisS's picture
Inconclusive?

JRusskie,

If you like Harman Kardon marketing, but you're not sure if Ford makes the best trucks in the world, then get yourself an F-150 and whatever truck you used to rumble across Afghanistan with, do your DBT (just like  the Pepski Challenge) and let us know what you come up with...

You are marketing TRUTH now? How pure is it?

I know some construction workers who might be interested...

ChrisS's picture
A War In Your Head...

JRusskie, Just looking at your response to John's post and comparing word-for-word what John wrote and what you think he says, there's such a huge world of difference!! There's a war in your head!

[Flame deleted by John Atkinson].

GeorgeHolland's picture
"Even so, they have to make

"Even so, they have to make compromises. Blind testing of loudspeakers, for example, is almots always performed in mono."

Well Mr Atkinson the reasoning behind testing speakers in mono is to eliminate the dreaded comb filter affect that would otherwise show up if a stereo pair were auditioned and the listener moved their head even a couple of inces. I'm surprised you didn't mention that fact but then again you think DBTs are hard to do, so if you don't know how to do them then indeed they are hard to do. *Chuckle*  Any DBT done should be auditoned is such a manner. The rest of your "excuses" for not doing them is the same old same old from you, nothing surprising there.

ChrisS's picture
"Kiss the girls..."

So Georgie Porgie,

Let's say JRusskie is DBT'ing a $1500 speaker and a $500 speaker and can't hear a difference, and you are DBT'ing a $4500 speaker and a $4000 speaker and you happen to have enough working neurons to hear a difference... Which set of speakers should the ex-shepherd construction worker buy?

John Atkinson's picture
Mono vs Stereo testing

GeorgeHolland wrote:
John Atkinson wrote:
Even so, they have to make compromises. Blind testing of loudspeakers, for example, is almost always performed in mono.

Well Mr Atkinson the reasoning behind testing speakers in mono is to eliminate the dreaded comb filter affect that would otherwise show up if a stereo pair were auditioned and the listener moved their head even a couple of [inches].

That is a consideration, of course, but in my opinion a minor one. As I had understood from Floyd Toole back in the day, the additional complexity required  of  Harman's physical speaker shuffling apparatus to do blind speaker testing in stereo was not justified by the results, ie, they felt that the stereo performance could be predicted from the mono results.

I don't agree with that, but more importantly, this illustrates the thesis offered in my lecture, that when you move the testing situation a step away from how the product is going to be used, you can't be sure that the assumptions you make haven't invalidated the test. As I write in the abstract to the lecture, "perhaps some of things we discard as audio engineers bear further examination when it comes to the perception of music."

BTW, I am still waiting for you to acknowledge that the criticism you made of my lecture, that it was not about Richard Heyser, was incorrect.

John Atkinson

Editor, Stereophile

GeorgeHolland's picture
More excuses?  I see what

More excuses?  I see what Johnny meant. You are never wrong. I am pretty sure Haram Kardon knows what they are doing. Please address any criticisms to them not me.

I am afraid that comb filtering IS a big deal. That would explain why cables "sound" different. It's not the cable but the listener changing where their head is between "testing"

You will be waiting a long time for any ackowledgement about your "lecture". Stop being the primadonna already.

Regadude's picture
Prima donna supreme

Georgie wrote:

"Stop being the primadonna already."

Look in the mirror and repeat those words!!!! laugh

ChrisS's picture
Oops! My head moved...

So how does one differentiate speakers that sound differently, amplifiers that sound differently, pre-amps, turntables, tonearms, cartridges, DAC's, etc., if a turn of one's head makes that much difference?

Where's your reliability, Georgie? Doesn't science depend on reliability?

JohnnyR's picture
For The Lazy and Mentaly Deficient.......

I can see what George is up against in here with Tweedle Dum and Tweedle Dee tag teaming and showing their ignorance.

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

Golly look what he meant.I think Ethan was banned from here ages ago for showing up Fearless Leader and his cronies and out right showing how REAL science works. BWAHAHAHAHAHAH loser boys.

Regadude's picture
For Johnny, about Johnny...
John Atkinson's picture
Ain't that the truth

Just bookmarking the link for future reference.

John Atkinson

Editor, Stereophile

JohnnyR's picture
Atkinson..........

......when you start doing a single DBT or even a SBT then you can talk about the "truth". Have you EVER designed and built your own speakers? Nahhhhhhhh you are too lazy or too "busy". Still finding plenty of time though to post online all the time though strangely enough.cheekyTill then you aren't an engineer so take your own advice and don't comment on speaker design anymore.

If you are going to save the link then please also save this link where discussion about it unfolded.

http://forums.audioholics.com/forums/loudspeakers/83412-diy-loudspeakers...

As you can see the original post was just one of many OPINIONS about the topic that is if you bother to read it at all. There are various OPINIONS about the topic and notice just how many of the so called "hobbyists" ended up being professional speaker builders. If you just pick and choose certain OPINIONS from the thread then you are guilty of leaving out facts.

For starters read the sixth post down by Jinjuku regarding Jeff's post that pretty much sums up where DIY has progressed.

ChrisS's picture
You think, therefore it's true...

Not real science either...

GeorgeHolland's picture
Frick and Frack strike again.

Frick and Frack strike again. Regadude and ChrisS always come up with strawman replies and ignore the links posted."Not real science "? How pompus can you get? Mr Winer measured the effects of comb filtering, what did you measure ChrisS the length of your nose when you typed that reply? You dismiss anything people link to yet show us nothing in return. Regadude, posting opinions isn't real science just so you both understand. Now run along lil boys and study real hard, maybe in another 20 years you might be able to hold your own in a discussion.

ChrisS's picture
Georgie thinks....

Has Winer's results been verified?

Did you know that Harman Kardon makes the best audio products in the world? And Ford makes the best trucks, right?

JohnnyR's picture
Do Some Testing Yourself Genius

Or is that above your abilities like thinking?

Go ahead, put on a pink or white noise source and move your head about and tell me the sound doesn't change. You won't bother so forget it ChrisSy.laugh

ChrisS's picture
ADHD

When moving a microphone while recording a person's voice, the sound changes. Did the voice change?

GeorgeHolland's picture
You never answer a question ,

You never answer a question , you just put forth silly questions of your own. That's what people do when they don't know or are scared to try.

Moving a microphone while recording a person's voice? If that's how you do things then no wonder you don't know what Johnny was talking about. Yes the sound changes as recorded by the microphone so what?  Genius.angle

ChrisS's picture
Dear Georgie...

I'll answer you this one... You and JRusskie always answer your own questions that you pose to everyone in these discussions. There's no need to provide any answer to you. As well, your attitudes and limited knowledge of the application of research methodology make civil and thoughtful discourse impossible.

So my questions to you and JRusskie are formed to reveal how each of you think whenever you provide a response.

You provide enough information for me to say that I find the "best" use of my time in these discussions is to make fun of you and JRusskie.

Ariel Bitran's picture
And Chrissy

makes some sense of the whole darn thing.

Regadude's picture
Tweedle dee and tweedle dum...

The only duo that strikes here George, is you and Johnny. You both STRIKE OUT!

Site Map / Direct Links