How Do You Buy High-End Speakers?

[AcuDefTechGuy]

If a speaker measures well it will sound good. That has always been true.

I disagree. It may be usually true, but not always true.

I've auditioned 2 speakers that measure very well (Paradigm Studio 100v3 & PSB Imagine T2) that I personally did not like very much.

PSB Imagine T2 Tower loudspeaker Measurements | Stereophile.com
Paradigm Reference Studio/100 v.3 loudspeaker Measurements | Stereophile.com

It doesn't mean they suck. It just means I didn't think they sound good to my ears. But others may think they are the best sounding speakers they have ever heard.

 

[Irvrobinson]

I guess this is the choice one has to make. Do I want a speaker that measures closest to the reference, or one that sounds closest to the reference to me.

I mean, if my hearing is weak above 10kHz, any reference above 10kHz sounds weak to me. But that sound is not what the reference is actually producing or how the reference was intended to sound. So should my speaker perpetuate that weakness in my ear? If the speaker is a little "bright" above 10kHz, the thing I actually hear is closer to what the instrument is producing and was designed to produce.

Maybe understanding this issue, and deciding which I want, will make selecting my speakers a bit easier.
More accurate or Sounds better?

Am I understanding this debate correctly?

Edit: I know there are no instruments, (short of a dog whistle), above 10kHz. It's just an example.

The debate I'm having with Irv and Andy is whether or not you should choose speakers by comparing them to a reference recording and referring to measurements, or simply buying what sounds good. It isn't about more accurate or sounds better, just how many and which selection criteria are used.

As for getting speakers that compensate for a lack of hearing sensitivity in certain octaves, I think that's folly. Since everyone's ears are nonlinear the only reasonable objective IMO is to have speakers produce sound as similar to the original source as possible.

 

[Irvrobinson]

I disagree. It may be usually true, but not always true.

I've auditioned 2 speakers that measure very well (Paradigm Studio 100v3 & PSB Imagine T2) that I personally did not like very much.

PSB Imagine T2 Tower loudspeaker Measurements | Stereophile.com
Paradigm Reference Studio/100 v.3 loudspeaker Measurements | Stereophile.com

It doesn't mean they suck. It just means I didn't think they sound good to my ears. But others may think they are the best sounding speakers they have ever heard.

That means the measurements taken were really insufficient to completely characterize the sound of the speakers. As a scientist you know that.

 

[AcuDefTechGuy]

The debate I'm having with Irv and Andy is whether or not you should choose speakers by comparing them to a reference recording and referring to measurements, or simply buying what sounds good.

No, our criteria is that the speakers must 1) measure great objectively and 2) sound great subjectively.

If the speaker ostensibly measures great, but does not sound great subjectively, then do not buy the speakers.

 

[AcuDefTechGuy]

That means the measurements taken were really insufficient to completely characterize the sound of the speakers. As a scientist you know that.

And how are we to know that the measurements were "insufficient"?

It's not like they underwent 4 years of clinical trials and then were approved by the FDA.

That is the reason we tell people to buy speakers that both measure great and sound great. You do not just go by the measurements alone because some measurements may not be as complete as they claim.

I do believe in evidence-based practice. Evidence is not just some numbers in the lab. The numbers must match the actual outcomes.

 

[Irvrobinson]

And how are we to know that the measurements were "insufficient" by your standards?

Because if it measures great and sounds bad then the measurements must be insufficient (or improperly taken or presented). For example, some presentations of measurements can lie, in a way. Like 1/3rd octave smoothing in frequency response graphs. Even 1/6th smoothing can make some in-room measurements look a bit too good to be true.

 

[AcuDefTechGuy]

Because if it measures great and sounds bad then the measurements must be insufficient..

Yes, but we only know that AFTER we listen to the speakers. Thus, the speakers need to measure great and be confirmed by our ears to sound great.

And even then, there are different grades of "great", which can only be determined by listening to the speakers.

And that is why (after the weeding out process) we buy the speakers that sound the greatest to us, and each of us has our own idea of great.

Is anyone claiming to have the most accurate ears? Not I.

I mean are any of us actually disagreeing here or are we just saying the same thing differently?

The speakers must measure great. The speakers must sound great.

 

[Rickster71]

.... Do I want a speaker that measures closest to the reference, or one that sounds closest to the reference to me.

I mean, if my hearing is weak above 10kHz, any reference above 10kHz sounds weak to me. But that sound is not what the reference is actually producing or how the reference was intended to sound. So should my speaker perpetuate that weakness in my ear? If the speaker is a little "bright" above 10kHz, the thing I actually hear is closer to what the instrument is producing and was designed to produce.

Bingo.
That's been my contention all along.

 

[BoredSysAdmin]

I personally see no reason to spend that much, I think one of the best speakers I ever heard was the sound scape 8's {something that stands alone with no sub, seems to be alive and actually felt like the sound followed you around, they are under 10K unless you get crazy with the finish, but to my ears for 2.0 listening you will not find a better speaker, I have heard speakers well over 50K and the sound scapes are all over where you think you should be...

Herbu, like many folks here - I don't see myself spending that kind of money on speakers just yet.
That said I wholeheartedly agree with many others - you will have to listed them for yourself, with your favorite tracks and ideally in your room to make an educated decision.

I also urge you to get and use excellent software from Harman - How to listen - Harman How to Listen
It will help you to train your ear and know what's good sound and what is not

p.s: Salk's SoundScape 8 is by far the best speaker I ever heard - over B&W 800d and Wilson's Maxx (100k+) - but don't take my or Imloud word on it - try to get their audition before purchase if possible

 

[Swerd]

Buying speakers that measure perfectly flat or play the music just as it was recorded, only works for the few people that have perfect hearing.
The rest of us have to contend with the fact that our hearing isn't the same as when we were 18 years old.
We all need recent hearing tests to determine at what frequencies we can still hear and where on the spectrum we have a loss.
That means the solution is, what sounds good to 'You'.

I see two problems with that idea. One is philosophical.

If your hearing is less than "normal for an 18 year old", everything you hear will be affected by that, including live human voices and live musical instruments. Your brain adapts to that and perceives those sounds as the normal reference. If you get a hearing aid to correct defective hearing, at first things will not sound normal, not unlike people's reaction with new glasses.

Getting speakers with built-in correction would make all recordings sound altered, but would not affect how you hear live sounds. It would be like driving a car with vision correction in the windshield instead of wearing glasses.

The other problem is more technical. To apply correction to a speaker will require active crossover controls. Passive crossover networks can lower responses by filtering, but cannot boost anything. In contrast, eyeglasses can magnify or demagnify things we see to achieve focus.

 

[Rickster71]

I see two problems with that idea. One is philosophical.

If your hearing is less than "normal for an 18 year old", everything you hear will be affected by that, including live human voices and live musical instruments. Your brain adapts to that and perceives those sounds as the normal reference. If you get a hearing aid to correct defective hearing, at first things will not sound normal, not unlike people's reaction with new glasses.

Getting speakers with built-in correction would make all recordings sound altered, but would not affect how you hear live sounds. It would be like driving a car with vision correction in the windshield instead of wearing glasses.

The other problem is more technical. To apply correction to a speaker will require active crossover controls. Passive crossover networks can lower responses by filtering, but cannot boost anything. In contrast, eyeglasses can magnify or demagnify things we see to achieve focus.

My idea was one I've thought about for a while now and by no means do I have it hashed out completely.

It would work better with headphones rather than speakers, obviously. That would make it more analogous to wearing glasses.

We are all in a similar situation now.
Either we have perfect hearing or we don't, that, combined with the fact none of us have speakers that play perfectly flat.

 

[Dennis Murphy]

I see two problems with that idea. One is philosophical.

If your hearing is less than "normal for an 18 year old", everything you hear will be affected by that, including live human voices and live musical instruments. Your brain adapts to that and perceives those sounds as the normal reference. If you get a hearing aid to correct defective hearing, at first things will not sound normal, not unlike people's reaction with new glasses.

Getting speakers with built-in correction would make all recordings sound altered, but would not affect how you hear live sounds. It would be like driving a car with vision correction in the windshield instead of wearing glasses.

The other problem is more technical. To apply correction to a speaker will require active crossover controls. Passive crossover networks can lower responses by filtering, but cannot boost anything. In contrast, eyeglasses can magnify or demagnify things we see to achieve focus.

You're certainly correct in stating that passive crossovers can't by themselves increase output. But many tweeters--particularly ribbons--will have a rising output in the 12k - 20k region. That's exactly where we old farts are deficient. So the designer has to decide whether to flatten out the tweeter response at the top, or to leave it somewhat elevated. I think the correct solution is open to debate. Fortunately, I can ignore the issue with the RAAL tweeters I use, since the foam pads that are supplied can be adjusted to do whatever you want in shaping the highest octave. As for the argument that we should choose speakers that both measure well and sound great--I dunno. If it sounds great, why does it have to measure great? My personal opinion is that speakers that measure well in all important dimensions (including distortion, off-axis response, and decay) will sound great to people who are used to what live music sounds like. But I'm not going to tell someone they shouldn't be enjoying what they are enjoying simply because it doesn't meet my measurement standards. I might suggest that they listen to a couple of speakers that do meet those standards just to make sure that they aren't overlooking something they might enjoy even more.

 

[AcuDefTechGuy]

As for the argument that we should choose speakers that both measure well and sound great--I dunno. If it sounds great, why does it have to measure great? My personal opinion is that speakers that measure well in all important dimensions (including distortion, off-axis response, and decay) will sound great to people who are used to what live music sounds like. But I'm not going to tell someone they shouldn't be enjoying what they are enjoying simply because it doesn't meet my measurement standards.

I think the salient point is enjoying the sound.

The measurements just help us get there.

 

[zieglj01]

As for the argument that we should choose speakers that both measure well and sound great--I dunno. If it sounds great, why does it have to measure great?

But I'm not going to tell someone they shouldn't be enjoying what they are enjoying simply because it doesn't meet my measurement standards.

Bingo - It is about preference, and about one enjoying their adventure.

 

[AcuDefTechGuy]

I guess this is the choice one has to make...

Among many.

So can we rule out bookshelf speakers at least?

Then you have to decide passive vs. fully active speakers vs. partially active speakers.

Is there a weight or size limit on your towers?

Are there speakers that clash with your aesthetic taste?

 

[GranteedEV]

So can we rule out bookshelf speakers at least?

Negative.

Genelec 8260A

 

[AcuDefTechGuy]

Negative.

Genelec 8260A

Yeah............ I don't think the OP is going to spend $10-15K on those.

 

[GranteedEV]

At 12k / pair he can't get them much cheaper. Sorry they're not red

 

[AcuDefTechGuy]

Sorry they're not red

Well.....then......they're just no good.

My younger brother was over my house last weekend. He had the nerve to say, "So why are your speakers red again?"

I said, "Why are some pianos red and some violins red? Because I like the damn color!"

 

[GranteedEV]

How bout I just buy you some rose-tinted shades?