If all speakers measure flat, what is the point of auditioning?

[terryforsythe]

Historically, it wasn't very common.
Maybe it's better now?
I am somewhat skeptical.

It's never bad to be skeptical.

I'll share my experiences, though.

1. I bought a miniDSP SHD without ever first hearing it. Amir's review, in combination with researching its feature set, gave me everything I needed to know that it was the right unit for me.

2. I had my eye on the KEF LS60 Wireless speakers. After seeing Erin's measurements and subjective comments on these speakers (Erin's Audio Corner), I bought them for my office - having never heard them in person. They sound exactly how I expected them to sound based on Erin's data.

3. I bought Elac Uni-Fi Reference bookshelf speakers for my family room prior to ever hearing them. The driving factor there was that they were the best match to our decor that I could find but, again, Erin's data informed me on what to expect from these speakers. It was dead on.

4. I decided to make my Elac speakers all active to overcome some of their weaknesses (which are clearly apparent in Erin's data). To do so, I purchased a Hypex Nilai 500 Stereo amplifier and two Topping LA90 Discrete amplifiers based on Amir's reviews. They sound fantastic. I also replaced my miniDSP SHD with a miniDSP Flex HTx to get 8 output channels, having never heard it. The Flex sounds just as good as the SHD.

In sum, my systems sound fantastic, and I bought all of the equipment based on testing performed by Amir and Erin, without ever having heard any of the equipment before purchasing. In this regard, I don't know of any retailers in my area that even carries them. Perhaps I could find the LS60s or Elacs on display somewhere locally, but the rest of the equipment probably not.

 

[mhardy6647]

It's never bad to be skeptical.

I'll share my experiences, though.

1. I bought a miniDSP SHD without ever first hearing it. Amir's review, in combination with researching its feature set, gave me everything I needed to know that it was the right unit for me.

2. I had my eye on the KEF LS60 Wireless speakers. After seeing Erin's measurements and subjective comments on these speakers (Erin's Audio Corner), I bought them for my office - having never heard them in person. They sound exactly how I expected them to sound based on Erin's data.

3. I bought Elac Uni-Fi Reference bookshelf speakers for my family room prior to ever hearing them. The driving factor their was that they were the best match to our decor that I could find but, again, Erin's data informed me on what to expect from these speakers. It was dead on.

4. I decided to make my Elac speakers all active to overcome some of their weaknesses (which are clearly apparent in Erin's data). To do so, I purchased a Hypex Nilai 500 Stereo amplifier and two Topping LA90 Discrete amplifiers based on Amir's reviews. They sound fantastic. I also replaced my miniDSP SHD with a miniDSP Flex HTx to get 8 output channels, having never heard it. The Flex sounds just as good as the SHD.

In sum, my systems sound fantastic, and I bought all of the equipment based on testing performed by Amir and Erin, without ever having heard any of the equipment before purchasing. In this regard, I don't know of any retailers in my area that even carries them. Perhaps I could find the LS60s or Elacs on display somewhere locally, but the rest of the equipment probably not.

So what I read indicates you bought on and/or benefited from real measurements from real people performing (reasonably) unbiased assessments -- and not the manufacturers' specs. Which was kind of my point.

 

[Digital_Thor]

If you put your brutally analytical signal through very musical amps and speakers, or vice versa, it might live up to its billing. I had an experience like that with some Paradigms once.

"Musical amplifiers and speakers" ?? What does that even mean?

Way, way to many people still think/believe that +/-3dB is a precise speaker, and that good amplifiers somehow add something to the sound, rather than bad amplifiers actually removing or lacking something.

In my opinion, you should audition to experience how much we cheat ourselves and think we know what is going on. It's fun to find out that we sometimes can hear certain details, and other times definitely cannot - but - I hardly care to audition anymore, because it's almost the same story each time. Storytelling and bla bla. Give me some raw data and less BS instead.

First time I listened to a set of KEF R7 Meta, ignoring the salesman talking of mystic DAC's and wonderful amplifiers. I only lasted 3 minutes before I wanted to stuff a sock in those ports, add subwoofers and EQ all the linear FR errors.

Lastly to the OP.. what measurements? Unless I get to measure myself or get some trusty second hand opinion from guys like Erin, Amir or others critical reviewers, then I see measurements purely as a guideline - as a maybe.

 

[ctrl]

Here's how very similar monitors really measure in MLP in a real room, without any EQ, those are Genelec 8040 (red and green) and Neumann KH150 (blue and orange)

Such simple "truths" are always somewhat problematic.

[UPDATE: I probably misunderstood you, and you wanted to show that two on-axis very similar speakers can still differ sound wise significantly at the MLP - sorry if so.]

The microphone at the main listening position (MLP) captures the "sum" of the direct sound and hundreds of reflections. The frequency response (FR) at the MLP thus represents something like an "average," meaning even slight deviations can indicate significant differences in radiation.

The distinction of spatial information by our hearing, vertical sound radiation is perceived differently than horizontal sound radiation (see HRTF), is completely lost during recording with the microphone.
In your diagram, only the differences in timbre/sound character are shown, the spatial differences are lost in monaural recording. Even binaural recordings do not provide all spatial information, as standardized dummy head and torso, ear shells, and ear canals are used.

To get an impression of how similar the two speakers sound, short recordings of pink noise at the MLP would have been helpful instead of showing the FR at the MLP.

The differences of your measurements roughly correspond to similar differences between two speakers of the predicted in-room response (PIR). Let's take a closer look at the differences above 800Hz (where the influence of room modes hardly plays a role):


If we now normalize to one of the speakers, we get the differences in an easily understandable form:

To get an impression of how much this difference affects the sound, one could now capture this difference with an EQ and apply it to pink noise and comparing it with the unaltered pink noise:



Attached is a recording where pink noise is switched back and forth between altered and original version several times, to get a very rough impression of how much the timbre/sound character of the two speakers differ based on your MLP measurements.

With that, everyone can then judge for themselves how large (very roughly) the differences in timbre/sound character of the speakers are without the differences in spatial impression.

The differences in sound power, early reflections, and PIR that your MLP measurements suggest, despite the speakers very similarly dimensions and design, are likely largely due to the very different crossover frequencies - KH150 [email protected] and 8040B XO@3kHz.

 

[Digby]

Here measurements trump everything.

Hmm, but which measurements..the frequency response? One speaker can be "flat" (+/-3db) with a 2db peak at 1khz, on axis, and another "flat" and have a 2db peak at 4khz and a 2db trough on 1khz, on axis. They will sound different, this before a myriad of other measurements are taken into account.

Why would I need to even listen to a pair of speakers if they measure flat?

Because it is only one piece to the puzzle, not the be all and end all. You might prefer a speaker that is less flat, but performs better in other aspects. It might suit your needs/application better.

All other variables equal, would they not sound the same?

They never are equal though. Loudspeakers are the least equal part of the chain.

I really think the idea that 'if a speaker measures flat enough on and off axis' it can be bought with hearing, should be put to bed. If you look at the best rated speakers in the review index, you'd find the top speakers actually sound different from one another, some surprisingly so. How can they all be liked equally by the same listener, in the same room, given the same listening material?

There is no magic bullet imo.

 

[Purité Audio]

Really fine measuring loudspeakers don’t sound that different from one another.
Keith

 

[smallwonder]

Hmm, but which measurements..the frequency response? One speaker can be "flat" (+/-3db) with a 2db peak at 1khz, on axis, and another "flat" and have a 2db peak at 4khz and a 2db trough on 1khz, on axis. They will sound different, this before a myriad of other measurements are taken into account.


Because it is only one piece to the puzzle, not the be all and end all. You might prefer a speaker that is less flat, but performs better in other aspects. It might suit your needs/application better.


They never are equal though. Loudspeakers are the least equal part of the chain.

I really think the idea that 'if a speaker measures flat enough on and off axis' it can be bought with hearing, should be put to bed. If you look at the best rated speakers in the review index, you'd find the top speakers actually sound different from one another, some surprisingly so. How can they all be liked equally by the same listener, in the same room, given the same listening material?

There is no magic bullet imo.

From the arguments from science-minded audio enthusiasts I would say the ones that are highly rated here, and measure as fat across their entire frequency capabilities. The question I proposed was simply a thought exercise. I am attempting to get towards a consensus or perhaps find out where our opinions diverge. Either way this has been interesting!

 

[Mr. Widget]

Really fine measuring loudspeakers don’t sound that different from one another.
Keith

True we are far from the days of AR3a and JBL L100s... but as we approach perfection don't we focus in on smaller and smaller deviations?

 

[Snarfie]

Did measure quite some same size speakers from different brands. Imo they sound after correction in Charakter more the same specific lowend, more space between voices an instruments so better imaging. Only difference are more or less imaging forward or backward firing ports that result in overall more low but some have more thighter bass some less. Also bass reflex design or transmission line build are probably more responsible for those differences. But overall the bass performance is more present wich makes a hugh improvement. .

 

[ahofer]

"Musical amplifiers and speakers" ?? What does that even mean?

Well....exactly. I thought there was sufficient context to understand that I was poking fun.

 

[JeremyFife]

Wow.

I just voted and saw the vote results... 50:50. That surprised me.

I get it for digital transports and DACs. I kind of get it for amplifiers.

Not for speakers though.
Perhaps it's just me, but I don't think measurements alone tell me how speakers will perform in MY room.
I do get that if the speaker response is reasonably flat and if the directivity profile is one that will respond to EQ then it's going to sound good. If it's ability to produce SPL without distortion at the level I want then it will sound good and also loud enough.
Do any two speakers measure the same like this?
Contradicting myself;
I am pretty sure I'd be happy with Elac DBR62's or Revel M16's (I have simple needs, a small room and a limited budget) and I can think that based on measurements.
If I wanted the 'best' speakers I'm also sure that I could do better, but I'd have to audition them.

 

[tmtomh]

I got my Genelec 8351b's without ever having heard them. Best speakers I've ever had, by a wide margin.

I think when it comes to actives with built-in DSP, it's not too difficult to buy based solely on properly done measurements. If the Klippel looks good and the directivity is good (i.e. no major errors and therefore easily EQ'd), then it's really just about low-frequency extension, SPL capability, distortion vs SPL, self-noise (less important for conventional mid/far-field hi-fi style listening), and of course price and looks.

With that said, I think where extensive listening could have helped me is that I might have felt that something like Neuman 310's with external DSP, or perhaps KEF LS60's, or maybe Sointuva AWG's with a Purifi amp and external DSP, could have given me more or less equally good performance for less money.

But I wouldn't trust my listening comparisons unless I could get all that stuff into my own listening space, all at the same time, and live with all four setups, properly calibrated, for at least a couple of weeks. Totally impractical, and probably impossible - and even then I would have confidence but not certainty in my listening impressions.

Time is money too, and if we're fortunate enough financially to be able to consider $6k-$8.5k setups like these, it's not always worth the time, physical hauling, hassle, and stress to go to extraordinary lengths to see if you can get closer to $6k than to $8.5k.

I got free overnight shipping on the 8351b's, they fit on stands I already had, GLM was dead easy to use, and they were dialed in after a single GLM measurement calibration, 1-2 slight tweaks of the target curve, and 20 minutes spent experimenting with slight differences in toe-in/toe-out angle. That time and aggravation saved/avoided is very valuable to me, and if these speakers last 10-15 years, the possibility that I paid $2k more than maybe I had to for the equivalent sound I'm getting, will work out to about 25-50 cents a day. I can live with that, to say the least.

 

[suttondesign]

the next project should be how to measure and predict the magic of dipoles. i agree that measurements of monkey coffins is a great guide, and the genelec/neumann/dutch/kii speakers are awesome; but non-box speakers are beguiling in a way that someone ought to try to quantify. i assume the polar plots would show something distinctive?

 

[Mr. Widget]

Wow.

I just voted and saw the vote results... 50:50. That surprised me.

I get it for digital transports and DACs. I kind of get it for amplifiers.

Not for speakers though.

I agree, rather surprising... I also agree with those that say you can't expect an in store demo to be the last word in whether or not you will like the performance in your room. I really need to hear them in my listening room before I will sign off on them.

In general I tend to be very particular in making most purchases and life decisions. That may be a factor in why I won't buy clothes on-line or choose speakers by measurement alone.

I got my Genelec 8351b's without ever having heard them. Best speakers I've ever had, by a wide margin.

Glad it worked out, but depending on your expectations it could have been a disappointment.

 

[holdingpants01]

Glad it worked out, but depending on your expectations it could have been a disappointment.

I also bought them without listening, no complaints, otherwise I think I would question my taste first rather than their performance lol

 

[Mr. Widget]

I also bought them without listening, no complaints, I think I would question my taste first rather than them if I wouldn't like them lol

Genelec is definitely a quality brand who produce excellent products, but they also produce quite a range. Even if you decide that Genelec is the brand for you, measurements alone wouldn't satisfy me on which product I would choose.

Again, I think this is more about how picky we are or accepting we are rather than the utility of measurements.

 

[holdingpants01]

Genelec is definitely a quality brand who produce excellent products, but they also produce quite a range. Even if you decide that Genelec is the brand for you, measurements alone wouldn't satisfy me on which product I would choose.

I can understand that, but on the other hand that was rather simple in my case, as I wanted coaxials with possible W371A upgrade path later, that meant only 3 models, one a bit too small and one too big to later not fully use it, so basically I had no choice. There's not much overlap between the series if any, only a lot of the sizes in specific model ranges

 

[napfkuchen]

I recently ordered a pair of speakers after reading Erin's and a SoundOnSound review. Let's hope the in-room-response improves on my current living room setup. Otherwise my "faith" in measurements/spinorama would be shattered.

 

[tmtomh]

Glad it worked out, but depending on your expectations it could have been a disappointment.

Well, that's the point I and others have been trying to make - we tend to overestimate the chances for satisfaction (or for having bought the best speaker for our situation) when we select through uncontrolled listening auditions, and we tend to underestimate the chances for success from buying "blind" using measurements (that is, properly done measurements combined with and understanding of what measurements we need to see and what they mean).

So of course, it could have been a disappointment. But I didn't get "lucky" - I made an informed decision and the chances were very good that I would not be disappointed.

What I was talking about above is the converse: I might also have ended up feeling satisfied with one of the other setups I named, and those would have cost slightly less.

 

[solderdude]

I have spent the better part of this week trying to understand shopping for speakers. I come from a home theater background and there accuracy is less important...I think. Here measurements trump everything. Why would I need to even listen to a pair of speakers if they measure flat? All other variables equal, would they not sound the same?

all variables equal... yes of course because in that case they are the same.

But are all variable equal when shopping for speakers ?