Speakers with the best measurements

[theshade]

Can we have a list of speakers with the best measurements? Additionally, can we have a discussion as to why you think which measurements are better? For example, although both speakers 1 and 2 have relatively smooth frequency response, the dip at ___ hz of speaker 1 is probably less noticeable than the peak at at ___ hz of speaker 2. And so on.


Ideally we can also have results of controlled blind listening tests. Although from my knowledge only harman does that.


For illustrative purposes:

Under $500:

1. Speaker 1 (measurements from soundstage url)
2. Speaker 2 (measurements from stereophile url)
3.
4.
5.

Under $1000:

1.
2.
3.
4.
5.

Under $1500:

1.
2.
3.
4.
5.

Under $2000:

...

Under $5000:

...

Under $10000:

...

Under $15000:

...

Cost no object:

...

 

[TLS Guy]

Can we have a list of speakers with the best measurements? Additionally, can we have a discussion as to why you think which measurements are better? For example, although both speakers 1 and 2 have relatively smooth frequency response, the dip at ___ hz of speaker 1 is probably less noticeable than the peak at at ___ hz of speaker 2. And so on.


Ideally we can also have results of controlled blind listening tests. Although from my knowledge only harman does that.


For illustrative purposes:

Under $500:

1. Speaker 1 (measurements from soundstage url)
2. Speaker 2 (measurements from stereophile url)
3.
4.
5.

Under $1000:

1.
2.
3.
4.
5.

Under $1500:

1.
2.
3.
4.
5.

Under $2000:

...

Under $5000:

...

Under $10000:

...

Under $15000:

...

Cost no object:

...

If only it were that simple. The really important thing is the waterfall plot, and you can't find that for most speakers. You still can not pick a speaker from measurements alone, but can tell which will be lousy, and that's most.

 

[Nuance AH]

I think you're right; I believe only Harman does that.

The best measuring speaker to my knowledge is the JBL Pro LSR6325. I do not know what it costs, though.

The most important measurements can be found in Floyd Toole's new book. Kal Rubinson does a "summary," though at this link. Here is a snippet of the article:

"1) On-axis Frequency Response (FR)
2) Listening Window: a spatial average of FRs in the ±10° vertical and ±30° horizontal windows
3) Early Reflections: a spatial average of the FRs of the early reflections from the four walls, ceiling, and floor
4) Sound Power: a weighted average of all of the above to assess the total acoustic energy radiating from the source
5) Directivity Index: the difference between the on-axis response and sound power"

The five measurements above are what Kal Rubinson describes in his article as "being statistically predictive of subjective satisfaction" according to Floyd Toole's research/findings.

 

[theshade]

I think you're right; I believe only Harman does that.

The best measuring speaker to my knowledge is the JBL Pro LSR6325. I do not know what it costs, though.

sean olive in avsforum said that they use jbl 6332 in all their harman reference listening rooms. he also compared the jbl 4328. I quote:

"Yes, you are correct: we chose the LRS 6332 because they have higher SPL capability. Since we are doing our own loudspeaker/ room correction we had no need for the room-mode correction (RMC) built into the 4328p. The 4328p actually has a slightly smoother frequency response and more extended LF response than the 6332. However, with the four subwoofers and our own auto-equalization, this was not an issue for us."

Thanks for the links. I also enjoyed ur journey to find the perfect speaker (although not all cause the thread is too large already). I try to read all harman studies (toole, olive, welti, etc.) available on the net although I doubt I understand all of them. Aside from that I read all sean olive posts, earl geddes posts, patrick hart posts, and wmax posts, david rich reviews, d. b. keele reviews, the audio critic, linkwitz website. So I am heavily influenced by what they say. In fact my most recent audio purchase is a sony 7506 due to what wmax said about them.

My current speakers are usher x-719 for fronts and energy c3 for surrounds with an adire audio tempest 15 122l sealed sub with a behringer feedback destroyer. I probably wouldnt have bought the usher x-719 if I were to buy now but damn they look good!

I started this thread so it would make it simpler for people to find out which speakers have the best measurements at different price levels. However, I do not have the technical ability and "know how" to confidently interpret loudspeaker measurements or even rank them at different price levels.

Thus I waited for posts. However to get the ball rolling I would probably just list the speakers that to me have the best measurements according to what I have read and remembered. Once again I apologize that I will not attempt to interpret measurements cause I am not technically proficient and would not like to provide erroneous comments.

1. Infinity Primus 360, 362 - numerous posts about this with measurements on avsforum and audioholics.

2. Infinity Intermezzo 4.1t - from the audio critic

3. B&W Matrix 803 series 2 - from the audio critic

4. Infinity Interlude Il 40 - from the audio critic

5. Waveform Mach 7 - from the audio critic

6. Linkwitz Orion - from the audio critic and linkwitz website

7. Gedlee speakers - from diy audio and gedlee website

8. JBL LSR 4300 series - from jbl pro website spec sheet

9. JBL LSR 6300 series - from jbl pro website spec sheet

10. Legend Acoustics speakers with deqx - from legend acoustics website (reviews with measurements are linked there).

11. NHT xd - from soundstage and stereophile

12. energy c3 - from soundstage

13. infinity beta 40 - from sensible sound

14. yg acoustics - from soundstage

15. revel concerta f12 and ultima salon 2 - from soundstage

16. axiom audio m80 v2 - from soundstage

17. paradigm reference signature s1 and s8 - from soundstage

18. psb platinum m2 and t6 - from soundstage


With regard to double blind listening tests, how I wish harman would publish preference ratings even just among the harman brand loudspeakers in controlled double-blind tests.

Sean Olive said and I quote:

"Originally Posted by theshade View Post
Is it possible to know preference ratings among the harman brand loudspeakers in controlled double-blind tests so it would be easier to choose where we set our own point of diminishing returns based on our different budgets.

Sean Olive answers:

I would have to convince our marketing department to release this information. In addition, we would have to include a specification that captures the maximum SPL output capability for each product so you know whether or not the loudspeaker meets your application requirements.

At the present time, I am trying to convince marketing to publish the comprehensive anechoic loudspeaker measurements that I've shown you, and which CEA and CEDIA are considering to adopt. JBL Pro already publishes these measurements for their products."

I really hope that CEA and CEDIA adopts this as it would be a great help to us consumers who want the most bang for the buck performance.

 

[Nuance AH]

I saw that Sean Olive quote too; don't remember which forum, though. I don't think Harman's marketing team will release which speakers measure better, but it would be coo. Who knows!?

The problem with comparing measurements between the various magazines is they each use different techniquies (and probably gear) to do it, and not many have anechoic chambers. Soundstage! has a set of measurements that I feel are very accurate, but I wouldn't trust many others.

But since we're adding to the list:

Salk SoundScape on-axis:

Salk SoundScape 30 degrees off-axis:

 

[bellevegasj]

This is a really interesting idea. I would love to compare what I hear to some objective #s.

 

[j_garcia]

My speakers measure at:

Frequency response +/-1.2 db (anechoic) 3 db down point is 55Hz

And I am quite pleased with them I have to agree though that even with a ruler flat curve, that does not guarantee a speaker that will please everyone's ear.

 

[fredk]

I have to agree though that even with a ruler flat curve, that does not guarantee a speaker that will please everyone's ear.

Nope, but it guarantees that you will get everything that is on the recording. You can then shape the response curve to your taste. On the other hand, if its not there to start with, you can't add it in.

 

[TLS Guy]

Nope, but it guarantees that you will get everything that is on the recording. You can then shape the response curve to your taste. On the other hand, if its not there to start with, you can't add it in.

Not so fast! You can have a dreadful speaker with a flat response at one recording position. There are lots of tricks to make a speaker give a trace like that. You have to have a waterfall plot. The reason: for a speaker to be pleasing and produce a believable sound stage you have to know a lot about the off axis response. A response at one mic position means virtually nothing.

You also need to know a lot about a speakers phase and time response as well as distortion.

And when all is said and done you have to listen to it, and evaluate with natural speech and acoustic instruments and the signing voice.

We still can't give up on critical listening evaluation of speakers, because the science is not that good yet. The limit of measurement is to tell us which speakers will sound dreadful and to not bother with, but measurement can not yet find the jewel.

 

[Nuance AH]

Not so fast! You can have a dreadful speaker with a flat response at one recording position. There are lots of tricks to make a speaker give a trace like that. You have to have a waterfall plot. The reason: for a speaker to be pleasing and produce a believable sound stage you have to know a lot about the off axis response. A response at one mic position means virtually nothing.

You also need to know a lot about a speakers phase and time response as well as distortion.

And when all is said and done you have to listen to it, and evaluate with natural speech and acoustic instruments and the signing voice.

We still can't give up on critical listening evaluation of speakers, because the science is not that good yet. The limit of measurement is to tell us which speakers will sound dreadful and to not bother with, but measurement can not yet find the jewel.

^This. Well said!

 

[johnlewisgrant]

This is one damn good question! It goes to the core question: can one loudspeaker be said to be objectively better than another? Inotherwords, are there speakers that are clearly better than others, independent of what a particular listener or class of listener may think or perceive?

Marketing departments depend on the answer being NO. It's really subjective in the end, what's best is what you like. True, with one caviat: what's best FOR YOU is what you like.

I don't buy it. It's tautological; and it doesn't fit experience or scientific fact.

So, having laid out my position.... where is the "list" of the winners of the double-blind test!!!

JG

 

[ski2xblack]

I don't buy it. It's tautological; and it doesn't fit experience or scientific fact.

JG

You could still have a speaker with great specs and which many agree approaches the ideal. Then, inevitably, you'll come across someone with weird ears, or a large head, or other idiosyncratic anomaly of their physical anatomy, making the speaker sound less ideal to them.

I agree that the engineering should strive for the evidence supported objective ideal (and disclose relevant product information!), but it's gotta be near impossible to fully bridge the gap and say without equivocation that x speaker is objectively superior, period. Great goal, and cool thread, though.

 

[Nuance AH]

So, having laid out my position.... where is the "list" of the winners of the double-blind test!!!

JG

I do not know where the "master list is," but I do know that the Infinity Primus 360/362 constantly beat out a few pairs of speakers costing many times the price. I believe one of the other speakers was from B&W's line. Here is the link. It was finally revealed that the "speaker on the left" was in fact the Infinity 360/362. Its not the nicest measuring speaker ever, but that $500/pair of speakers beat out the pair on the far right, which cost $4000/pair. I'd say that's pretty impressive.

 

[AJinFLA]

You could still have a speaker with great specs and which many agree approaches the ideal. Then, inevitably, you'll come across someone with weird ears, or a large head, or other idiosyncratic anomaly of their physical anatomy, making the speaker sound less ideal to them.

Or that someone simply prefers/enjoys highly colored sound. There is an entire cottage industry known as the "high end" that caters to some of these folks. Hence the demand for SE amps connected to so called fullrange speakers, etc., etc.
Olives tests show that the vast majority will consistently pick/prefer the better measuring speaker every time...under blind conditions. Those measurements include far more than just the on axis FR as typically stated.
And I believe even more than those are needed.

cheers,

AJ

 

[Warpdrv]

Well this is all well and good, you find the best speaker that offers the best measurements and response in the world, then you take it home to your imperfect room and now you no longer have the best measuring speaker in the world anymore....

Just another problem to work around... you have to fix your room now...

So everything is subjective to a certain degree...

 

[lsiberian]

The challenges of measurements.

1. Room factoring.
2. Performance variance within specific speaker.
3. Personal preference still exists.

Even in double blind test people select the bottom speaker out of a sample size. This indicates preference.

4. No way to factor Psychoacoustic effects of Aesthetics and marketing.

The reality is double blind testing also relies on peoples short term memory so it isn't full-proof. Imagine if you are hungry and taste a certain pizza. The next pizza may not test as good. Especially if you finish the slice.

There is also program material familiarity in recordings. It's been shown those with intimate familiarity of a recording can detect flaws.

Finally few people have the patience and intelligence to understand the measurements.