• Welcome to ASR. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

What role should listening play in speaker design

Status
Not open for further replies.
This thread is fascinating and I am surprised how polarizing it is. To me it appears to some people the concept of a "speaker designer" that makes design choices based on listening is anathema and goes against the know science of speaker design using measurements. I think the problem is they assume that the designer is mainly making EQ choices which color the sound. While that may be true in some cases the real job of a speaker designer is to manage compromises to come up with a design that best meets the target market. The design choices are things like SPL vs extension, size vs SPL and extension, making the best use of a limited budget, etc. It is not about "right or wrong" it is about the best you can do within all the constraints you are given. A perfect example are the cheap Pioneer speakers Mr. Jones designed, every single element of the design is a compromise to the theoretical ideal of what the speaker could be but many consider the choices are good ones on balance.
 
Last edited:
It seems this discussion is going again to the still unfortunately existing audio's circle of confusion where loudspeakers are wrongly seen as art and not engineering.

Sorry to quibble (hey it’s what we do here)…but…

It can be good to remember that the above is an individual value statement not some objective truth.

It’s fine for you, and any number of others, to treat loudspeaker design as purely engineering. And with specific engineering goals to converge on. That approach would obviously help, if it could be widely implemented, reduce the circle of confusion.

But it’s like saying “ the job of an amplifier is to amplify the signal with as little distortion and load variance as possible.”

Sure. If you want that. If that’s your goal.
But it’s possible to have another goal and to deliberately design amplifiers that distort the signal. (Eg some tube amps). The success of a design depends on how it meets ones subjective goal.

There’s no problem with somebody appreciating a level of “art” being involved in speaker design - if by that we mean speakers designed towards the subjective goals and preference of the speaker designer.

Somebody might prefer such a design.

(and we can add the caveat, such a person could also say; “ I like the aesthetics of the design and I like the out of the box thinking and the goal this speaker design represents.
Whether or not I would select it as the best sounding speaker under blind conditions, in the conditions I’ll be using it - in terms of my perception - I find it very pleasing, and pleasing in the ways it pleases it’s designer).
 
Last edited:
So, you are suggesting that AJ was at the recording session? Because that is what the word "nonsense" there implies. Of course, there are valid reasons to argue that not being at the recording session is irrelevant to designing a good and accurate loudspeaker. This would be a logical counterargument to AJ's point and an interesting contribution to the thread.
'Nonsense' as in it's nonsense that you need to be at the sessions or the live performance to judge if your playback is accurate to the recording.

The starting point for playback is the finished recording, not the performance, and we can be fairly accurate to that.

Industry figures in general do tend to push a 'Who knows what's right or wrong? It's all subjective' line regardless if they make accurate devices or junk. Everyone covers everyone else's back.
 
This measurement-based trend is very new and narrow.
Is that so?
I think that in professional audio engineering, a measurement-based audio environment had already been the standard long before 2016.
 
The thread is totally, TOTALLY out of reason here, debating whether listening tests are needed or what AJ uses and if he believes in measurements.

And centers at a sub-subject of his expertise as it seems convenient, cause touching the core element would leave most of us at loss.
Lots have an idea about speakers in whole, right?

Let's move now to the burning subject which as far as I'm reading all these years only couple of folks seem to know-how here:

Would anyone be able to build a driver from the ground up without measurements and end-up been decent?
Like the Kef coaxial for example which AJ set the foundation, or the Mofi one with the new architecture, etc?

Science is not about been cozy at the laurel leaves, avoiding the fundamentals just to make a point. We can be better than this.
 
Would anyone be able to build a driver from the ground up without measurements and end-up been decent?
Yes, decent - but not excellent
 
Question: If we can develop speakers with just measurements and no fine tuning by ear...then why do we have people mastering and mixing things on actual speakers, when they too could just do it all on a computer and simulate?
Huh? Making art is different than faithfully reproduce it. It is a purely subjective product with countless ingredients and preference. Not so as far as playback equipment.
 
Having watched this thread in the last day or so, it appears that the HiFi audio subjectivists are alive and well.

They all came out of hiding from the woodworks cheering on feel, gut, artistic expression and emotions as the anthem of speaker design.

We have gone back few decades of scientific progress with this thread. I bet the next thing to follow is Danny Ritchie to get promoted as the new standard bearer of speaker design.
 
Yes, decent - but not excellent
I mean my decent:
Not punishing impedance, linear, high SPL ability, low THD, etc.

I wouldn't even know where to begin on each of these attributes.

Think about it the other way:

There's thousands of speaker brands, and a lot of them are nice, using the rights tools, etc.
How many nice driver companies are around? In the tens? A hundred? A tenth of them proven through time?
And how many of them build nice coaxials?

So...
 
Making art is different than faithfully reproduce it. It is a purely subjective product with countless ingredients and preference. Not so as far as playback equipment.
I would argue that the "art" has not yet been completely removed from speaker design. For example: there is often a trade-off between smoothness of on-axis response and smoothness of in-room response. As you wrote in your review of the Revel F208:
I was puzzled that the on-axis response was not as smooth as I expect Revel speakers to have. That disappeared when I looked at the rest of the measurements starting with predicted in-room response … Wow! With no DSP this speaker managed to get such balanced and smooth response. It appears that the off-axis response is tuned to counter the on-axis response
How would the designer decide which to optimize if not by listening and hearing which "sounds better"? (And yes, ideally this should be done blind to eliminate any possible bias.)
 
Agreed: so long as you keep in mind the caveat: “if one is looking for scientific levels of certainty.”

Each of us is a country whose borders no language fully crosses. We may dispatch ambassadors, send paintings, poems, songs, metaphors, but even when they are received with love they arrive translated.

You like to say that sometimes subjective impressions can match your own and provide you with useful information, if the conditions are right, etc*. I won't argue that, but even if it's true, it's a crapshoot. Someone else's subjective impressions are vastly more likely to be different from my own than they are to somehow align and thus are vastly more likely to obfuscate than to inform.

Let's call subjective impressions (like sighted, non-controlled listening) Door #1. It's closed. We can't see through it because we can't know all the factors that contribute to someone else's subjective impressions.

Door #2, on the other hand, leads to objective data that we can communicate to each other without translation, that we can replicate and verify, that isn't contingent on the vagaries of qualia. Door #2 is wide open for all to see.

So if I'm with my friends, I already care about how they see the world, so, well, they can say what they like--we all interface with the world subjectively, and I'm hanging out with them because I like them, not to gather information or make a purchase decision or learn something or arrive at an evaluation.

But if a person wants to communicate something to others in a way that informs them of some phenomenon in the world, if he wants to design audio equipment, evaluate it, review it professionally, sell it, or make a persuasive case about it of any kind, and he chooses Door #1, then I have to assume his intent is to obfuscate, not inform.



*imagine if the best endorsement of klippel nfs were "it sometimes works"!
 
I would argue that the "art" has not yet been completely removed from speaker design. For example: there is often a trade-off between smoothness of on-axis response and smoothness of in-room response. As you wrote in your review of the Revel F208:

How would the designer decide which to optimize if not by listening and hearing which "sounds better"? (And yes, ideally this should be done blind to eliminate any possible bias.)

Speaker design has a much smaller impact on what we hear than the content itself. Those things aren't equivalent even if what you say is technically true. It's reductive.

They can seek the flattest anechoic response with the directivity and bass extension they want by measuring and testing. How else?
 
They can seek the flattest anechoic response
Of what? On axis, or in room? Those are not the same, and a decision has to be made as to which to optimize (which is what my Revel example was supposed to illustrate). Similarly, smoothness of directivity vs smoothness of on-axis response; there is often a tradeoff there as well.

Measurements cannot tell you what to optimize for, because technology that optimizes everything does not exist (and given the nature of the problem, likely never will).
 
Since this thread has gone off the rails, I will chime in. I conceive of talented designers/engineers in this way: Magico may make some of the best sounding and measuring speakers available, but their best engineering is priced out of reach for nearly everyone on Earth. So, how do you make a good speaker that sells for $5,000, or $2,000 or $500. You have to compromise somewhere - remember how excited we all got about the Palmer Orbit 11 before they got tested? Then we learned that - while they may be excellent at what they do - you can't make perfect speakers sold at that price point. It seems to me that Andrew Jones is a talented designer who is also extremely curious. He could have kept iterating and tweaking at Elac, but he seems to like to challenge himself. Put a passive radiator in front of a woofer? Here's the Adante. A 10-inch coaxial? MoFi. And now he's messing around with field coils. They're not all gonna be home runs, and he doesn't seem to be chasing a specific idea of sonic perfection; he just wants to explore and play. The fact that so many of his speakers measure well, and are considered good deals at various price-points (again , not 100% of the time, but often) suggests to me that he knows what he's doing. There may be SOME disingenuous marketing when speaker designers lean in to subjectivity, but in the case of AJ, I think that is because he takes real risks, not because he couldn't copy Revel or Ascend or whoever. Now, if these new ones measure poorly and cost as much as a car, we can call it a failed experiment, and a knock on his reputation. But I have to assume that is unlikely.
 
This thread is fascinating and I am surprised how polarizing it is. To me it appears to some people the concept of a "speaker designer" that makes design choices based on listening is anathema and goes against the know science of speaker design using measurements. I think the problem is they assume that the designer is mainly making EQ choices which color the sound. While that may be true in some cases the real job of a speaker designer is to manage compromises to come up with a design that best meets the target market. The design choices are things like SPL vs extension, size vs SPL and extension, making the best use of a limited budget, etc. It is not about "right or wrong" it is about the best you can do within all the constraints you are given. A perfect example are the cheap Pioneer speakers Mr. Jones designed, every single element of the design is a compromise to the theoretical ideal of what the speaker could be but many consider the choices are good ones on balance.

I concur. While do not want to downplay the value of applying some level of controlled listening, speaker design is primarily an engineering discipline. In an earlier day, certainly seems reasonable that there might have been more need to rely on listening. In an industry fraught by marketing mysticism, not surprising that some designers got or get hyped to an elevated status.

Measurements are essential to making well reasoned tradeoffs in engineering. There may be some refinement or a close call that may involve more subjective calls, but the real work today is done with CAD tools to reduce design iterations and guesswork.

If I am investing in technology-based product, I want to know solid engineering practices are being applied. There may be some wiggle room for an audio speaker, but for my car, airplane or my furnace, I hope a smart engineer (or engineering team) is using the best measurements to optimize the design and NOT someone who claims superior senses to do so!

As was writing, someone posted that this thread has gone off topic. Is late here and so will let it go for now. Tomorrow may be considering whether this and some other posts need a different home. Cheers! :)
 
Last edited:
Someone else's subjective impressions are vastly more likely to be different from my own than they are to somehow align and thus are vastly more likely to obfuscate than to inform.

I think that’s likely quite exaggerated.
But in any case, noting when somebody else is accurately describing the sound of speakers you’re familiar with could give one more confidence when they describe a speaker you haven’t heard yet.

I have a long running thread on Audiogon in which I describe many different loudspeakers that I’ve auditioned (or owned).

The thread has plenty of comments along the lines “I’m familiar with those loudspeakers and your descriptions correlate very well with my own impressions.”

This has led to people telling me the thread has been very useful in leading them to check out certain loudspeakers based on my descriptions. And they virtually always report to my descriptions were correct.

As one example, a member who owned Harbeth SuperHL+ speakers was intrigued by the Joseph Audio Perspective speakers, wondering if he should go for a secondhand pair for sale. Since he had found my description of the Harbeth’s very accurate, he asked me how those two loudspeakers would compare. Since I owned both of those loudspeakers I gave a detailed rundown on the sonic characteristics of each, the strengths and weaknesses of each with respect to one another, what he could expect to gain with one and lose with the other, etc.

From my description, he decided to buy the Joseph speakers and he was very happy with the results, saying that I absolutely nailed the way each of those speakers sounded. And his descriptions when he compared the two were just like mine.

That kind of stuff happens quite a bit in the audiophile world. As I keep mentioning a tool is only useful if you decide to use it. If somebody is actually interested in the subjective evaluations from reviewers or other audiophiles it is possible to navigate the field to some degree, finding those whose descriptions you have reasons to build some confidence in, and it does not necessitate having to fall for pseudoscience and woo woo.

Again… if you’re looking for more specific and more reliable information about how gear performs, then obviously this is where measurements and a more scientific approach comes in.

But plenty of us still find that exchanging subjective reports can be fun and also sometimes useful.

Along those lines… I really like seeing measurements for AJ’s speakers because they are informative. But I also personally can find certain subjective reports informative. For instance, I found Erin’s subjective part of his MoFi V10 review informative about “ putting together all the measurements. How does this thing sound?” He mentioned some things that I personally wouldn’t necessarily have pulled strictly from the measurements. (And so many people comment under his videos about how appreciative they are that Erin is able to not only talk about good measurements but put into words “ how those measurements sound when listening to the speaker”)

*imagine if the best endorsement of klippel nfs were "it sometimes works"!

* Imagine if somebody said the informal method of cooking used by most people in the world was useless because it was not being done with scientific rigour.
 
I would argue that the "art" has not yet been completely removed from speaker design. For example: there is often a trade-off between smoothness of on-axis response and smoothness of in-room response. As you wrote in your review of the Revel F208:
In all the times I have tried to correct the PIR, I have only succeeded once or twice. All other times, on-axis correction ruled.

On-axis neutrality is the most important characteristic of a speaker. Due to precedence effect, it arrives at the ear first and sets the stage. Getting this wrong and then relying on off-axis to compensate is a completely crapshoot. No way can you predict the reflections in anyone's room. Today, many homes have open floor plans, meaning some of the reflections don't even exist!

This is why I keeps saying flat on axis and well behaved off-axis. We want the off-axis to be similar to on-axis so that should there be reflections, have them tonally similar to on-axis.

If flat on-axis response makes off axis worse, then that is designer's job to fix. Not make on-axis worse as a cure.
 
Status
Not open for further replies.
Back
Top Bottom