For the learned here, very interesting to follow audio history to maybe 1960. It shows the game of catch-up played by people trying to systematize with measurements what seemed clear to careful listeners.
No, I am definitely not saying ears are better than measurements - only in that early period. Today, the issue is more likely, as per OP, which measurements are the most meaningful.
In as much as nobody ever even remotely can be fooled into thinking a speaker sounds like a real instrument (even in a room down the hall), we really must have a long way to go, eh.
There are advocates for all kinds of speaker systems, say FR. Some ESL folks think all other kinds of speakers are woefully second-rate. I sure do. That must mean we each are focusing on one specific virtue(s) and living graciously with shortcomings in other respects. Does that mean different measurements for different folks? Would an ESL person say, "if it can't reproduce square waves it aint no good"? Would an FR person say, "if it the sound quality shifts from low freq to high (like some say is true of a Bosendorfer piano) it aint no good"?
No, I am definitely not saying ears are better than measurements - only in that early period. Today, the issue is more likely, as per OP, which measurements are the most meaningful.
In as much as nobody ever even remotely can be fooled into thinking a speaker sounds like a real instrument (even in a room down the hall), we really must have a long way to go, eh.
There are advocates for all kinds of speaker systems, say FR. Some ESL folks think all other kinds of speakers are woefully second-rate. I sure do. That must mean we each are focusing on one specific virtue(s) and living graciously with shortcomings in other respects. Does that mean different measurements for different folks? Would an ESL person say, "if it can't reproduce square waves it aint no good"? Would an FR person say, "if it the sound quality shifts from low freq to high (like some say is true of a Bosendorfer piano) it aint no good"?
On the which sounds best (Dan's polar responses) I'm going to take stab in the dark and say the rokit 6 followed by the HR624 Mk II (I'm assuming the peak in response between 10 and 20K is not too offensive).
Why did I choose those two? Mainly because the off axis frequency response varies consistently, and to boot with the rokit 6 it seems to be the least directional as well.
Tony.
Why did I choose those two? Mainly because the off axis frequency response varies consistently, and to boot with the rokit 6 it seems to be the least directional as well.
Tony.
I recently bought a set of Behringer B2031A's, just to give them a listen. Although they are very different from my Unbaffled Dipoles, they do actually sound pretty good! The only requisite is that you sit in the sweetspot, or pretty close to it. Although my dipoles do sound best in the sweetspot, the sound is still very good throughout the room. The Behringers aren't actually constant directivity - I'd rather call it constant directivity slope, or something like that.
Recently I listened to a system with a highly directional WG from about 1.5 khz, a more or less cardioid driver below it (which becomes more like a dipole at the bottom of its passband) and closed boxes below about 200 hz. The polar response of this system is very smooth. The sound was simply great! Very clean, neutral and precise. When sitting in the sweetspot, it was among the best I've heard. However, once you stood up, everything collapsed.
These days most people would probably agree that a smooth frequency response and uniform dispersion are very important when it comes to loudspeaker sound. There is still disagreement about exactly how directional the speakers should be.
For the time being, my conclusion is that highly directional speakers can give you exceptional sound quality, with great precision, accurate tonal balance and cleanness - but not very lifelike and only within a strictly confined area.
Less directional speakers generally sound bigger, more natural and involving, less precise, less tonally accurate, more quickly sounding congested when you turn up the volume - yet sounding good throughout the room.
At this point I think constant directivity dipoles are a good compromise between both, being less directional than WG speaker, but more directional than most box speakers. A downside of course, is the high cost. You practically can't do without digital or at least active filtering, which means you need many amps and you need very robust drivers or you should accept limited dynamic range.
By the way, kudos to Behringer, who made quite an astonishing product for the price!
Linkwitz comes to similar conclusions.
Spatial distortion -
Recently I listened to a system with a highly directional WG from about 1.5 khz, a more or less cardioid driver below it (which becomes more like a dipole at the bottom of its passband) and closed boxes below about 200 hz. The polar response of this system is very smooth. The sound was simply great! Very clean, neutral and precise. When sitting in the sweetspot, it was among the best I've heard. However, once you stood up, everything collapsed.
These days most people would probably agree that a smooth frequency response and uniform dispersion are very important when it comes to loudspeaker sound. There is still disagreement about exactly how directional the speakers should be.
For the time being, my conclusion is that highly directional speakers can give you exceptional sound quality, with great precision, accurate tonal balance and cleanness - but not very lifelike and only within a strictly confined area.
Less directional speakers generally sound bigger, more natural and involving, less precise, less tonally accurate, more quickly sounding congested when you turn up the volume - yet sounding good throughout the room.
At this point I think constant directivity dipoles are a good compromise between both, being less directional than WG speaker, but more directional than most box speakers. A downside of course, is the high cost. You practically can't do without digital or at least active filtering, which means you need many amps and you need very robust drivers or you should accept limited dynamic range.
By the way, kudos to Behringer, who made quite an astonishing product for the price!
Linkwitz comes to similar conclusions.
Spatial distortion -
Who says Frequency Response isn't the end all be all?
You should read some of the studies by Sean Olive of Harman. Floyd Toole's book describes one of his studies and it refered to here:
Audio Musings by Sean Olive: Part 3 - Relationship between Loudspeaker Measurements and Listener Preferences
He was able to find a mathematical weighting scheme that took the factors of; on axis response flatness, on axis response smoothness, off axis response smoothness and bass extension and then use them to rank order any group of test speakers. The objective ranking correlated well with subjective ranking in controled listening tests.
Speakers with wide, smooth and flat on-axis response and resonance free power response were prefered. (All are anechoic measurements)
Power response shape or directivity wasn't a factor. Polar curves were not a factor. Distortion was not a factor. Time response was not a factor.
David S.
You should read some of the studies by Sean Olive of Harman. Floyd Toole's book describes one of his studies and it refered to here:
Audio Musings by Sean Olive: Part 3 - Relationship between Loudspeaker Measurements and Listener Preferences
He was able to find a mathematical weighting scheme that took the factors of; on axis response flatness, on axis response smoothness, off axis response smoothness and bass extension and then use them to rank order any group of test speakers. The objective ranking correlated well with subjective ranking in controled listening tests.
Speakers with wide, smooth and flat on-axis response and resonance free power response were prefered. (All are anechoic measurements)
Power response shape or directivity wasn't a factor. Polar curves were not a factor. Distortion was not a factor. Time response was not a factor.
David S.
But did they investigate these factors as thoroughly as they investigated the importance of on-axis response and power response?
I think over the decades Floyd Toole has done enough tests to remove those as big factors. What was interesting about the Olive study was to come up with the weighting coefficients of the known factors that created the best correlation scheme. In the end he had a high correlation between those factors as weighted, and listener preference.
This really gets at the root of the matter: If we listen to two speakers in an unbiased situation (blind test with no room factors favoring either system, for example) and prefer A to B, what measureable factors could have predicted that? The answer is that those 4 factors with the right weighting would.
That doesn't exclude all other factors having a possible effect (at some level), but it does show that the factors used are sufficient discriptors.
David S.
I thought polar response is the differentiator.
But Elias provided interesting challenge to this argument. His WTF analysis clearly show ROOM is the differentiator.
In hindsight this might be the reason why Orion and Behringer statistically sound the same when compared by a group of canadian audiophiles.
But Elias provided interesting challenge to this argument. His WTF analysis clearly show ROOM is the differentiator.
In hindsight this might be the reason why Orion and Behringer statistically sound the same when compared by a group of canadian audiophiles.
Dave,
From what I read in Floyd Toole's book, I got the impression that the speakers tested were mainly conventional box speakers, accompanied by a couple panel speakers - with obviously problematic frequency response and funky off-axis behavior. Did they also compare them with for instance a Gedlee Summa-like speaker, or well-designed constant directivity dipoles? Were their own studio or PA speakers included?
Don't get me wrong, I value their research highly, but I don't think it is definitive. There are still questions to be answered.
Exactly which post of Elias' are you referring to?
From what I read in Floyd Toole's book, I got the impression that the speakers tested were mainly conventional box speakers, accompanied by a couple panel speakers - with obviously problematic frequency response and funky off-axis behavior. Did they also compare them with for instance a Gedlee Summa-like speaker, or well-designed constant directivity dipoles? Were their own studio or PA speakers included?
Don't get me wrong, I value their research highly, but I don't think it is definitive. There are still questions to be answered.
Exactly which post of Elias' are you referring to?
Hi,
The Zu speakers are very well received by the reviewer at Stereophile, the only factor many die hard technical speakers designer will dispute is the speakers freq. response curves. Go to this web page to study the FR curves, they will floor you.
Zu Essence loudspeaker | Stereophile.com
The Zu speakers are very well received by the reviewer at Stereophile, the only factor many die hard technical speakers designer will dispute is the speakers freq. response curves. Go to this web page to study the FR curves, they will floor you.
Zu Essence loudspeaker | Stereophile.com
The highend speaker store where I used to work soled those. I listened to them before I ever saw any measurements. Let me tell you that those measurement results didn't surprise me one bit!
EDIT: btw, those guys do seem to be catching up 😛 . A couple of years ago their speakers measured even funnier! http://www.soundstagemagazine.com/measurements/zucable_druid/
Last edited:
Yes, in the end most of Floyd's studies center on conventional commercial speakers and not so much on the "alternative" system technologies the we diy fans like to experiemment with. He did look considerably at the ESL 63.
He centers on measureable performance, as he should. As a researcher the "transduction principal" isn't as important as the resultant measureable performance. All of the information on importance of off axis response or general directivity can be applied to whatever particular speaker style you are interested in.
In other words a dipole, a CD waveguide, a line array, are distinguished when you can define "the importance of power response". Here I disagree with the usual interpretation of Toole as showing the importance of power response. He waffles a bit in his own conclusions but in the end he can only show that power response smoothness is important (it is a good revealer of significant resonances) but the shape of the power response isn't a factor. Hence d.i. isn't significant.
By extension, polar response isn't important although, if axial response is crucial and listener position relative to the system is not well fixed then near off-axis response needs to be considered.
I think most of the answers are in there.
David S.
The most enjoyment I get in reading Stereophile is watching John Atkinson tap dance around disparities between the measurements he takes and the subjective impressions his reviewers give.
"The surprise for me was how much of its measured misbehavior was not too audible..."
David S.
That was funny indeed. There were so many crashing stupidities in this review that I wonder that anyone can take this stuff seriously. Of course, the main function of a reviewer is to write material that sells magazines and ad space, so JA can't actually come out and say, "This product is horribly colored; the reviewer is apparently deaf or completely insensitive to instrument tonality."
Ough, what an ugly spelling mistake 😕 . I can't edit the post anymore but I'd actually like to rephrase the entire sentence 😱 .
Zu speakers are another of any number of products that I see and listen to at shows and want to like but can't. I don't know what the attraction is.
Strangely enough, there are speakers which measure great, but sound awful. What you really want is a speaker that measures well and sounds good. More often than not, a lot of designers stop at "measures good" and refuse tweaks that may make the thing sound good, writing it off as voodoo or "audiofoolery".
John
Strangely enough, there are speakers which measure great, but sound awful. What you really want is a speaker that measures well and sounds good. More often than not, a lot of designers stop at "measures good" and refuse tweaks that may make the thing sound good, writing it off as voodoo or "audiofoolery".
John
- Status
- Not open for further replies.