Message to golden-eared audiophiles posting at ASR for the first time...

[sergeauckland]

Without polar pattern, the single on-axis FR measurements give an extremely incomplete picture for tonal balance.

Whilst this is true, how a loudspeaker sounds within any particular room will depend on that room. At least, the on-axis response gives a 'Best Guess' at a response without knowing about the room. A 'bright' room with few soft furnishings, highly reflective, (as beloved by loudspeaker advertising art-directors) will sound completely different to an overstuffed 'Victorian' sitting room full of furniture, bookcases, plants and other 'stuff'.

The interaction of a loudspeaker's off-axis response and the room acoustics will be equally dependent on the room and the loudspeaker's dispersion characteristics.

I like to see an accurate on-axis response, and hopefully a benign off-axis response such that it will integrate into the room much like any other sound source. That's about the best we can hope for in a normal domestic environment.

S.

 

[Sal1950]

Whilst this is true, how a loudspeaker sounds within any particular room will depend on that room. At least, the on-axis response gives a 'Best Guess' at a response without knowing about the room.

I completely agree. Though completely out of fashion today, on-axis tells me a big part of what I want to know about a speaker. Then if we have some knowledge of room acoustics and that room's design, a lot can be deduced from that information.

 

[j_j]

Without polar pattern, the single on-axis FR measurements give an extremely incomplete picture for tonal balance.

All of that, in 64 point bold italic.

 

[MattHooper]

As I’ve brought it before, when it comes to the average audiophile trying to correlate speaker measurements to understand what he/she prefers, there is the “Klippel Problem” (where I’m using that device as a stand in for any similar comprehensive suite of measurements).

If Klippel level measurements are required to truly make this correlation, the extreme paucity of those measurements can be a problem.

On one hand, you can take the route of just looking to a basic criteria: “ choose loudspeakers that measure like this!” (which would be something approximating what people call the Harmon Curve, even if ultimately that’s a misnomer).

And then you can try and find Klippel measurements for loudspeakers, which is going to be exceedingly limiting your choice.
You end up on a place like ASR where the same few brands are recommended over and over (KEF, Genelec, Revel, Neumann, JBL…)

Which is perfectly fine if somebody wants to take that route. It’s the “ this speaker is likely to sound good” route.

But that’s an entirely different scenario than an individual trying to understand how loudspeaker measurements exactly predict how a speaker will sound to him, and by that I mean not just the narrow range of speakers that happen to measure “Revel-like” but all the wonky speaker measurements that one would find from the wide array of loudspeakers one has experienced or that are available.

In other words if you’re not going to limit yourself to speakers that have Klippel measurements, then IF such measurements are required for understanding/predicting which speakers you will like, then you are mostly out of luck.

So it comes down to being able to listen to them…biases and all.

And personally, I don’t mind that approach…simply deciding on loudspeakers by listening to them. It’s served me very well over the years. The loudspeakers that I have liked listening to in stores or other venues are the ones I have enjoyed once I got them into my own home.

 

[MediumRare]

Would anyone like to suggest an additional measurement or two to fill in what a Sprinorama is missing? How can dynamics or cu ft room capability be measured, for example? Looking for the missing ingredients to an amazing speaker that also might compensate for lack of Spinorama perfection.

 

[SIY]

At least, the on-axis response gives a 'Best Guess' at a response without knowing about the room.

Not a particularly good one.

Yes, a full-on Klippel is the best way to look at things, but you can get a pretty good idea from a set of quasi-anechoic measurements at various angles. I didn't look at all the reviews Matt extracted the graphs from, but as I recall, the polar measurements appear in the reviews as well, yet they aren't mentioned in his post. Pareto's 80:20, they take a small fraction of the expense and complication of a Klippel but will give you 80% of the information you'd want for evaluating tonality.

 

[Jim Taylor]

But that’s an entirely different scenario than an individual trying to understand how loudspeaker measurements exactly predict how a speaker will sound to him, and by that I mean not just the narrow range of speakers that happen to measure “Revel-like” but all the wonky speaker measurements that one would find from the wide array of loudspeakers one has experienced or that are available.

Using [exhaustive] measurements to predict personal preference takes time ... and time ... and more time ... as well as a great deal of insight, work and dedication. That work and dedication is the same with any complicated hobby, such as golf, painting, billiards, poker or target shooting, among others.

So it comes down to being able to listen to them…biases and all.

Thirty years ago, I would agree wholeheartedly with that. However, nowadays there are so many choices that a buyer can't personally locate that it's not funny. It's just my personal opinion, but I think that there are more and more young(er) people today searching for a decent audio system in situations that are totally different from 30 years ago ... and they know it. They know and understand full well that their apartment isn't like Mom and Dad's house. They know and understand full well that their access to digital music via streaming ... especially over phones ... is a total game changer. Not only that, but the majority of the ones that I've met want to have equivalent sound on their speakers as on their IEMs or 'phones.

I'm not saying that they WANT to depend on measurements, I'm saying that they are FORCED to depend on measurements. Many of them have no choice.

If that were not true, how do you explain the people scanning this site as guests 24 hours a day, 365 days a year, to the tune of thousands per hour?

p.s. - I apologize if this is technically a bit off-topic.

 

[Mr. Widget]

Without polar pattern, the single on-axis FR measurements give an extremely incomplete picture for tonal balance.

100%

However while the current gold standard, the CEA-2034 or CTA-2034 as discussed by Floyd Toole, is certainly a very practical way to compare loudspeakers. Dr. Toole's research has shown the majority of listeners prefer speakers that have linear on axis and smoothly rising power response when using this standard. That said, I propose that it is not universally relevant. Near universal perhaps, but many audiophiles like speakers that sound really colored, or even downright peculiar.

I will define audiophile as someone who derives an unusual or exceptional level of joy when listening to music played back on a sound system. My theory is that members in this group have an unusually high number of members who do not respond to music playback in the same way as the general public. Many of these audiophiles do prefer objectively neutral well behaved speakers, but I think an unusually large percentage of this group find joy from a wide range of nonlinear systems.

This doesn't explain people spending great sums on power conditioners and over sized cables, but it may explain how someone in this group prefers a high distortion tube amp played through an objectively terrible speaker.

 

[Sal1950]

This doesn't explain people spending great sums on power conditioners and over sized cables, but it may explain how someone in this group prefers a high distortion tube amp played through an objectively terrible speaker.

IMHO, they all belong somewhere in a group labeled "audiophools".
They may have a love for music, but are too lazy or uncaring to do any homework on the subject.
They belong in that "sounds good to me" group of untrained market purchasers who build their systems composed of components that
attracted them to lay down their cash with little to no interest in whether it's actually High Fidelity. They may find the cosmetic design screams
"expensive", enjoy the warm romantic glow of tubes, or the distortion/FR deviations from accurate "sounds good" to him.
It also lumps in the ones spending $30k + on silver speaker cables again without doing any homework but are buying into a great marketing campaign
to the point that he does actually hear the improvements being sold by the snake-oil man. At least till one day he wakes up and does some blind
listening tests. But that requires too much work and those sparkly silver cables DO look good and are great conversation starters over their cost.
See my signature.

 

[SIY]

That said, I propose that it is not universally relevant.

I have said the same thing, but for different reasons. IMO, Floyd's research is absolutely solid and reliably predictive for a particular speaker paradigm (which, to be fair, covers the majority of use cases). But there are other approaches...

 

[Jim Taylor]

I will define audiophile as someone who derives an unusual or exceptional level of joy when listening to music played back on a sound system. My theory is that members in this group have an unusually high number of members who do not respond to music playback in the same way as the general public. Many of these audiophiles do prefer objectively neutral well behaved speakers, but I think an unusually large percentage of this group find joy from a wide range of nonlinear systems.

There are people who fall into the category you describe, but I don't believe that there are as many of them as what you imply.

I have said before that I have seen many people who buy audio gear on the basis of uncontrolled, emotional evaluation. In another year or two, they are dissatisfied, so they sell it or trade it in and buy another. One or a couple of couple years pass, and this happens again ... and so on an so forth.
I believe the answer is simple: they are using the wrong criteria, but are too stubborn to change.

My neighbors are a good example. He's a practical man, and buys his vehicle based on practical criteria. He has good luck with his vehicles.
She, OTOH, judges based on appearance. She buys a vehicle based on how well she likes the way it looks. Every couple years, she trades or sells her car because, as she says, "It just didn't work out."
She uses the same principle to pass judgement on everything in her life; her garden plants, her clothes and appearance, and the way the house is decorated. Her success in these other areas of her life is absolutely fabulous ... so she feels very confident in applying the same standard to her car.

Big mistake. A car isn't a flower, or a carpet, or clothing, or a hairstyle. But her husband can't convince her of that. She's stubborn. So her periodic automotive dissatisfaction is unending.

Many people in audio are the same.

 

[rdenney]

Would anyone like to suggest an additional measurement or two to fill in what a Sprinorama is missing? How can dynamics or cu ft room capability be measured, for example? Looking for the missing ingredients to an amazing speaker that also might compensate for lack of Spinorama perfection.

I like the linearity at various SPL that the Canadian NRC would provide with their spinorama measurements, measured at 2 meters (instead of 1) in an anechoic chamber. This more than anything told me how a speaker would perform when really pushed by a big amp in a big room, and that's an awfully important part of hifi speaker performance, it seems to me. But those are still limited to specific SPL outputs, instead of identifying the maximum SPL that can be produced before Range X or Range Y of frequencies starts to exhibit audible distortion. Yes, I realize that's a very fluid boundary, particularly when determining the audibility of distortion in narrow bands versus wider bands. Many dimensions challenges data representation.

We use "Klippel" as a stand-in for anechoic measurement, of course, because the latter is only rarely available.

I think it's important that a speaker that can't provide a realistic frequency response on axis isn't going to get any better by preserving that same response off-axis, though maybe that can be corrected using EQ. A speaker that has good on-axis anechoic performance can make itself sound poor with off-axis performance that significantly changes the coloration in unrealistic ways, which colors the reflected sound differently than the axial sound, making the result uncorrectable by EQ. That second speaker might look great on-axis (and thus that's the only graph the manufacturer will provide) but still simply not work in a normally reflective room. Making it sound good in that room will demand extensive room treatment to kill those off-color reflections, at the very least.

But all of that comes to no avail if the speaker loses its frequency response when pushed by a big amp to high listening levels, and those high levels are one of the use cases in a particular application.

I found that it was really difficult to determine, using available measurements, how loudly a speaker could be played such that it preserved its good frequency response and low distortion.

Rick "off-color" Denney

 

[Mr. Widget]

There are people who fall into the category you describe, but I don't believe that there are as many of them as what you imply.

I didn't mean to imply this was a large group. It is large in the context of people attending specialty audio shows, but compared to the general public they are less than a rounding error.

 

[j_j]

Maybe someday I will be able to say what I want about speaker testing. I hope.

 

[Madlop26]

In other words if you’re not going to limit yourself to speakers that have Klippel measurements, then IF such measurements are required for understanding/predicting which speakers you will like, then you are mostly out of luck.

So it comes down to being able to listen to them…biases and all.

sighs.... Klippel does not limit yourself, actually increases your ability to test speaker capabilities beyond our quite limited human ear abilities to measure sound. if you want to still trust your ears more, is ok, but to me sound like a person claiming to able to see better than a telescope, a delusion.

 

[Chris A]

Maybe someday I will be able to say what I want about speaker testing. I hope.

I'm sure that at least Greg Berchin and I would be interested in your insights on this subject.

If it helps, I think that I have some inking of what that subject might include. I'm actually ready for a little more in-depth discussion of that topic, i.e., not like those discussions that are typically bantered about on this site and others.

I find that psychoacoustics and small room acoustics are driving factors in this subject area, but I rarely (if ever) see discussions on those factors.

Chris

 

[MattHooper]

sighs.... Klippel does not limit yourself, actually increases your ability to test speaker capabilities beyond our quite limited human ear abilities to measure sound. if you want to still trust your ears more, is ok, but to me sound like a person claiming to able to see better than a telescope, a delusion.

You seem to have completely misunderstood what I wrote.

I didn’t say the Klippel measurements limit anyone. They of course give more of the type of information we would want.

I’m saying that there is a very limited implementation of Klippel measurements at this point. There are very very few speakers on the marketplace that have published measurements equivalent to Klippel scan.

That’s the limitation.

Now somebody may want to limit themselves to speakers that have Klippel level data available. Which as I said is perfectly reasonable.

But some of us don’t want to limit ourselves to those speakers, and when we don’t have the measurements we can try to hear the loudspeaker for ourselves.

Your telescope analogy has nothing to do with what I wrote.

 

[MattHooper]

Using [exhaustive] measurements to predict personal preference takes time ... and time ... and more time ... as well as a great deal of insight, work and dedication. That work and dedication is the same with any complicated hobby, such as golf, painting, billiards, poker or target shooting, among others.

Yup.

Some people are into that type of thing some people not so much.

As I’ve said, I’m interested enough and speaker measurements to have been reading them in Stereophile all these years, and also paying attention to Amir and Erin’s reviews and others. For me any review is always made better and more and informative when accompanied by good measurements.

I think one of the most potent ways of gaining personal experience correlating speaker measurements to the sonics, is getting into DIY. But like many other audiophiles I’m just not into that.

It’s like I would love to own a really nice sporty car. But even if I did, I wouldn’t get into it the way many car fanatics do, being totally into all the mechanics and understanding every part of the engine “ working on my car engine on the weekends” or whatever true car enthusiasts do. I’d mostly just enjoy the ride.

I’m not sure how being way more in to measurements would have actually elevated my own audio journey. Would it have given me more bliss over the years? I can’t imagine that since I’ve had tons of bliss.

Would I have enjoyed the hobby more? Would it have “ saved me” from buying numerous speakers are going out to audition lots of speakers? I don’t see how, because I don’t view going out to additional loud speakers as a chore… listening to different systems is one of my favourite things to do.
And there isn’t a single loudspeaker that I have owned or tried to my room that I regret.
I got something interesting out of all of them.

While some people may use speaker measurements to grade speakers from “good to bad” and so they would only be interested in the ones that measure the way they want, other audiophiles like myself are simply interested in how different gear performs. It’s all interesting “ good or bad.”

And I vastly prefer being able to listen to a loudspeaker myself versus only having measurements. For instance, the measurements of the Borresen x6 speakers told me they are not a speaker that I would want to buy. But given the choice between just staring at a chart or being able to actually listen to that speaker for myself… like if I could’ve hopped over to my buddies place while he had them… I’d much prefer listening experience. It’s just a richer experience than staring at a graph and more fun. And I’d get to shoot the shit with my buddy about what we are hearing. And maybe that chunk out of the mid range wouldn’t be as obvious or show up on as much music as I thought it might from the measurements, or maybe some other qualities could’ve made the listening experience very interesting despite the deficiencies. Etc.

So I’m certainly not positing my hearing as being more reliable than measurements.
But the experience of listening to a loudspeaker is much more enjoyable for me (and can sometimes tell me things that I wouldn’t have grasped from the measurements).

 

[Sal1950]

Maybe someday I will be able to say what I want about speaker testing. I hope.

Why hold back J_J ? Take my lead and say what you really feel.
Gets me in hot water now and then but whats the worst that can happen?
They can't eat you.
Or as we used to say FTITCTAJ

 

[j_j]

Why hold back J_J ? Take my lead and say what you really feel.
Gets me in hot water now and then but whats the worst that can happen?
They can't eat you.
Or as we used to say FTITCTAJ

Ownership issues.