Speakers that don't seem to match their measurements

[dfuller]

By which I mean, what are some speakers that sound good in spite of measuring not great, or speakers that sound pretty not good in spite of their near-perfect measurements? Let's limit to the ASR battery of tests.

I had a fascinating experience with this recently - mostly, sonics matched measurements, but there were certainly speakers that sounded better than they'd let on, and ones that sounded way, way worse.

The PMC 6-2, is without a doubt not a great speaker by Toole etc standards - their dispersion is messy and their on axis needs some help - but they were still pretty good! They did a lot more right than they did wrong (well, after you correct the usual PMC quirks by shelving the high end back to neutral); they had very good separation, really nice tight bass, and very low distortion. In contrast, the Genelec 8361 measures damn near perfectly, and yet sounded... pretty bad to me. Just kind of jumbled and muddled.

On the budget end, I'll add on the JBL 300s. I found them both unpleasantly bright and kind of tubby. The original Focal Alpha 65s were surprisingly decent in spite of mediocre at best measurements.

What about you? What speakers seem to defy measurements in your experience, either way?

 

[YSC]

Personally never experienced those, and most common of ppl claiming of sounded much better than the bad measurement is either they need boost in certain FR region or just simply get used to the lower fidelity sound which makes anything sounded more true to life sound bad to a first glance, just like those way oversaturated photos everywhere, it could be preferred to individual, but if what you're chasing is Hi-Fi, it isn't by definition

 

[ernestcarl]

I feel like this has been asked many times before.

Have all variables been truly accounted for?

It’s nice to have more stringent controls and actual verification measurements before testing (e.g. anechoic or quasi-anechoic/gated on-and off-axis) and thereafter measuring again for comparison at the listening room and MLP with a deeper analysis of the speaker units being auditioned; listening levels noted and careful volume calibration levelling (yes, the more quality assurance measurements, second verification, and third re-verification the better); LF response extension needs to be taken into account, too; then, of course, there’s room’s modal behaviour and boundary interferences… so, yes, a more detailed discussion of the state and design of the room acoustics can’t really be ignored/eschewed for us to progress beyond the more generalized anecdotal, personal preferences type only level sort of discussions here. Note, there’s no need for me to discount preference as yet just one more factor to consider, but is has to be disclosed.

Before we forget, the test tracks or recordings need to be considered and planned prior with some intelligence and analyzed as well — otherwise, we might be perpetuating more of the same old “circle of confusion” ad infinitum.

 

[ernestcarl]

Might as well mention at least one imperfectly measuring “cheap” speaker that I like: Fluid FX50

You can find a few of my own in-room measurements and impressions and compare that to Amir’s here at ASR.

 

[MarkS]

My speakers are GoldenEar Triton 7s, which I bought before discovering ASR. These get no love here, and available measurements are meh. But with Lyngdorf room correction and a single SVS Micro 3000 sub crossed at 60 Hz, they sound very good to me.

 

[sigbergaudio]

I don't think you will find speakers that "don't match their measurements" (to quote the title of the thread), but I suspect there is often a simplified understanding of the measurements.

I didn't find any measurements of the PMC 6-2, so hard to comment on that specifically. Do you have a link?

 

[RobL]

By which I mean, what are some speakers that sound good in spite of measuring not great, or speakers that sound pretty not good in spite of their near-perfect measurements? Let's limit to the ASR battery of tests.

I had a fascinating experience with this recently - mostly, sonics matched measurements, but there were certainly speakers that sounded better than they'd let on, and ones that sounded way, way worse.

The PMC 6-2, is without a doubt not a great speaker by Toole etc standards - their dispersion is messy and their on axis needs some help - but they were still pretty good! They did a lot more right than they did wrong (well, after you correct the usual PMC quirks by shelving the high end back to neutral); they had very good separation, really nice tight bass, and very low distortion. In contrast, the Genelec 8361 measures damn near perfectly, and yet sounded... pretty bad to me. Just kind of jumbled and muddled.

On the budget end, I'll add on the JBL 300s. I found them both unpleasantly bright and kind of tubby. The original Focal Alpha 65s were surprisingly decent in spite of mediocre at best measurements.

What about you? What speakers seem to defy measurements in your experience, either way?

If a “damn near perfectly measuring speaker” sounds “pretty bad” and “jumbled and muddled” to you, you need to look beyond the speaker for the issue.

 

[Digby]

What about you? What speakers seem to defy measurements in your experience, either way?

I will come right out and say I don't trust measurements, which is not to say that I don't trust them full stop, but I don't trust them to guide me to a pair of speakers I want to listen to.

You mention the JBL 300 series. Well reviewed here, good measurements, by rights I should be happy, yet they sound very V shaped and plasticky (resonances in the thin cabinets?) to me. When comparing say JBL 308, Hedd 07 MKII, Behringer B2031A and Eve SC208, these were all speakers that should be within spitting distance of each other, but even though they all had strengths and weaknesses, some I could live with, some not. That and there is quite the variety in presentation.

The most similar (sounding), and the ones I preferred, were the Hedd and the Behringer. The didn't seem to be much in it, save a price difference of about 4 times more for the Hedd.

If a “damn near perfectly measuring speaker” sounds “pretty bad” and “jumbled and muddled” to you, you need to look beyond the speaker for the issue.

My experience with the 8030C wasn't that they were jumbled or muddled, more so they have a somewhat 'diffuse' sound. I wonder if this is to do with the small cabinets for driver size. Maybe the absence of a much of a baffle creates this quality, perhaps this is similar to what dfuller heard...who knows?

I will say that I found the 8030Cs to be among the most singular (as in unusual) speakers I've heard. I really don't think they'd sound much like the KH80/KH120 they are often compared to, even though many recommend to just buy either one and 'be happy'. For me, at least, it is not as simple as it seems.

 

[ahofer]

I don't trust them to guide me to a pair of speakers I want to listen to.

I'd think of them more as a way to guide you *away* from many speakers that either don't have measurements, or measure poorly. Good measurements should be a baseline.

I still like my Harbeths. They are linear on-axis, but with poor dispersion. you need to be in position and on the closer side to really enjoy them as much as my Revels.

 

[Ron Texas]

One of the things I recall about Toole's work is the frequency bands are weighted equally. Perhaps a bit too much energy in the wrong place could have a negative effect on SQ which is not fully reflected in the preference score.

 

[Pearljam5000]

Everyone hears differently
That's my guess .

 

[RobL]

My experience with the 8030C wasn't that they were jumbled or muddled, more so they have a somewhat 'diffuse' sound. I wonder if this is to do with the small cabinets for driver size. Maybe the absence of a much of a baffle creates this quality, perhaps this is similar to what dfuller heard...who knows?

I will say that I found the 8030Cs to be among the most singular (as in unusual) speakers I've heard. I really don't think they'd sound much like the KH80/KH120 they are often compared to, even though many recommend to just buy either one and 'be happy'. For me, at least, it is not as simple as it seems.

The difficulty with making sighted subjective comparisons of anything is the usual biases that we all bring to the table with us. These biases will often influence our perceptions. If this thread wasn’t about dfuller’s opinion, I would disregard it outright. I think though, he has the experience to overcome personal bias so there is something else at play here (I hope!).
Regarding measurements, properly calibrated measurement equipment should contain no bias of any sort and will lay bare the actual performance of the DUT. If the DUT achieves what would be considered excellent measurements I believe we can be confident it will do what it was intended to accomplish. In the case of loudspeakers, an excellent measuring speaker (á la Toole) will faithfully transduce the signal it is fed.
If a speaker sounds bad to you, any cause related to the loudspeaker itself will be evident in it’s measurements. Outside of that the problems are probably source related or environmental. In regards to this particular thread, we are talking about a “damn near perfectly measuring speaker”. It should sound excellent. I live with a pair, they do.
I do have no problem with anyone saying that a speaker with less-than-stellar measurements sounds good to them. After a certain threshold, the perceptual improvements that “better measurements” provide are pretty incremental and small improvements in some areas in one speaker might be swamped by your personal preference for something that a poorer measuring speaker “does right” (for you!). I don’t agree that a “damn near perfectly measuring speaker” can sound bad. The speaker in question is one of the handfull of SOTA transducers frequently discussed (and owned) by members in this forum. Flaws inherent in the speaker of the magnitude implied here would certainly have been uncovered already.

 

[Digby]

The difficulty with making sighted subjective comparisons of anything is the usual biases that we all bring to the table with us.

True, but I feel I could pick out Genelec 8 series speakers, unsighted, compared to another pair (perhaps several) from the sound signature alone.

Regarding measurements, properly calibrated measurement equipment should contain no bias of any sort and will lay bare the actual performance of the DUT. If the DUT achieves what would be considered excellent measurements I believe we can be confident it will do what it was intended to accomplish. In the case of loudspeakers, an excellent measuring speaker (á la Toole) will faithfully transduce the signal it is fed.

What is a DUT? I'm not sure I agree in that a loudspeaker, more so than any other piece of audio equipment, is a collection of compromises of one sort or another. It is the most flawed part in the chain, introducing far more distortion, of one sort or another, than any other part.

This is without going into whether narrow or wide directivity is more accurate and other questions which are more a question of taste/application, than accuracy.

If a speaker sounds bad to you, any cause related to the loudspeaker itself will be evident in it’s measurements.

Doesn't that presume that all the measurements today (as done by Amir) are everything we need. Again, it is not necessarily as clear cut as it seems. I feel that certain types of flaws (which all speakers have by their nature) may be more egregious to one listener or another, listening to one type of music or another, than others.

A lot of the 'wonder box' speakers reviewed here (I use the term not to disparage them, but to suggest they do very well on measurement tests) would have trouble filling a room 6 or 7 metres in each direction with sound with 1 person in, let alone 5 or 15 people. So, it is rather more a case of which areas a speaker is excellent and whether that is relevant to the listener/application.

The speaker in question is one of the handfull of SOTA transducers frequently discussed (and owned) by members in this forum. Flaws inherent in the speaker of the magnitude implied here would certainly have been uncovered already.

Well, one flaw is it is not going to fill anything more than a moderate size room with sound. Does that matter, is it truly a flaw?....well, it all depends. The preference scoring is weighted in favour of certain qualities in a speaker, but does it actually function as a true ranking table, like some (certainly most newcomers) think it does. No, I think we've established it doesn't.

I think an individual probably needs a relatively high level of knowledge to read the charts and be able to pick which speaker they would most prefer. Some say that if you don't like a speaker that gets a high preference score, then the fault is with the individual - for me this is putting the cart before the horse.

FWIW, speakers with relatively similar performance can have rather different presentations, that may or may not suit an individual. I am most concerned, given how difficult it is to properly audition speakers in your own home, with finding out what positive aspects I need in a speaker and what negatives are less worrisome, rather than leaving it all to the charts/preference scores.

This is my somewhat fumbling approach to the matter anyway.

 

[rynberg]

Everyone hears differently
That's my guess .

Sigh...no they don't.

 

[dasdoing]

at moderate volume any smiley curve speaker will sound better than a flat one

 

[rynberg]

By which I mean, what are some speakers that sound good in spite of measuring not great, or speakers that sound pretty not good in spite of their near-perfect measurements? Let's limit to the ASR battery of tests.

I think if you really want real answers, you'll need to identify what you mean by "not measuring great."

Frankly, a lot of listeners have grown up with inaccurate sound and that's what they fall back to, especially in a short-term listening situation (and sighted, at that). I mean the v-curve or batman shape sounds great to a lot of listeners, even experienced ones, for a few moments. It's only in comparison to neutral speakers or longer listening sessions that most people prefer more neutral speakers.

 

[ahofer]

It's only in comparison to neutral speakers or longer listening sessions that most people prefer more neutral speakers.

Agree. Time in seat with familiar music is an important part of auditioning (and one of the reasons I find it time-expensive). Hopefully without a salesperson prattling in your ear.

 

[Digby]

Sigh...no they don't.

Well, he is correct in a sense, no two people have precisely the same hearing. It may also be a case of some people be more/less sensitive to certain flaws than others.

Although, in essence most people do hear the same, in that their ears function the same way.

 

[Digby]

Agree. Time in seat with familiar music is an important part of auditioning (and one of the reasons I find it time-expensive). Hopefully without a salesperson prattling in your ear.

Many here say auditioning has little to no value outside your own room, do you disagree?

 

[ahofer]

Many here say auditioning has little to no value outside your own room, do you disagree?

Kind of. The problems of how we have to search and buy audio are myriad, that's one of them. The last time I did it the old-fashioned way - auditioning in stores - was in 2018, and I did a decent job of finding a flat on-axis response by just long-term listening (not really using measurements, alas, although I did audition and like KEF). But I missed the dispersion part and now I have a clear comparison to another set of speakers I bought later.

Now I'd say narrow it down to some models that hit all the measurement hurdles at their price point and try to get a refundable purchase. My last purchase was a pair of Revel F228Be second hand. Not refundable, but I've been really happy with them, and I got a good price (with cosmetic defect - one of them says "PEPFORMA"). I did take my REW mic and laptop and take some cursory measurements before buying. Just looking for obvious stuff, like a busted driver.