I just found this great link that explains the different sonic signature of speakers

[Joe Schmoe]

My favorite sound is not even on the dial. It is precise and smooth at the same time (which they show as polar opposites.) My Cambridge Audio speakers have that sound. Not too sure who else does (maybe Mordaunt Short, since they are in the same "family".)

 

[gus6464]

My favorite sound is not even on the dial. It is precise and smooth at the same time (which they show as polar opposites.) My Cambridge Audio speakers have that sound. Not too sure who else does (maybe Mordaunt Short, since they are in the same "family".)

I've heard some Mordaunts and they definitely fall under that british house sound as Wharfedale, B&W, Epos, and others.

 

[cornelius]

My favorite sound is not even on the dial. It is precise and smooth at the same time (which they show as polar opposites.)

Absolutely, that is what my system sounds like. It seems like a contradiction, but it can be done...

 

[TLS Guy]

Absolutely, that is what my system sounds like. It seems like a contradiction, but it can be done...

Well of course it can be done. That is how any decent system should sound. The sound scape is just littered with those that don't.

 

[AcuDefTechGuy]

From Toole Audio Science Paper:

"The Harman International loudspeaker companies – JBL, Infinity, Harman/Kardon and Revel – have invested heavily in measurement facilities that allow them to take the fullest advantage of existing audio science."

I would think that their speakers would all sound outstanding, but that's not the case.

I have never listened them, but I would also think that the Revel speakers would sound superb, but some people said they don't even like them.

 

[TLS Guy]

From Toole Audio Science Paper:

"The Harman International loudspeaker companies – JBL, Infinity, Harman/Kardon and Revel – have invested heavily in measurement facilities that allow them to take the fullest advantage of existing audio science."

I would think that their speakers would all sound outstanding, but that's not the case.

I have never listened them, but I would also think that the Revel speakers would sound superb, but some people said they don't even like them.

I'm not surprised people don't like the Revel's. Here we go again with a design getting flat frequency response, that has to sound awful, because just the wrong set of compromises were made.

This is from the flagship Revel pdf.

Impedance 6.0 Ω (nominal)
3.7 Ω (minimum)
Indicates whether the loudspeaker presents a “difficult” or “easy” load on the
associated power amplifier. Combined with moderate phase angles, a
minimal impedance specification of 3.7 Ω allows a reasonably designed
power amplifier to drive Revel loudspeakers.
Filter Network Four-way, high-order acoustic response @
150 Hz, 575 Hz and 2.3 kHz
Indicates the acoustical characteristics of the filter network. Steep filters
indicate an optimized filter network that produces minimal acoustical
interference, low distortion, and expansive dynamic range.

Now a fourth order passive crossover at 150 Hz, is just something, that if you know what you are doing, you don't even consider for a successful loudspeaker. For one thing the caps have such huge values, they will have to be poor sounding non polarizing electrolytic types. The inductors to avoid them being gigantic will be made of far too small a gauge of wire, wound on iron cores.

But to top it off the phase angle of the crossover is 360 degrees, so at 150 Hz the mid is arriving 6.7 msec ahead of the woofers! At 575 Hz the tweeter is 1.7 msec ahead of the mid. At 2.3 KHz the tweeter is 0.43 msec ahead of the mid.

The point being, is that this speaker if you measured its transient response would be awful. I can assure time smear of this magnitude, makes for a speaker you would never want to own!

If I were shopping for speakers, those would not even make the audition list.

Speakers like these make my point. We need to ditch receivers. If these speakers were powered with digital crossovers the situation would be entirely different. We have to move away from this outdated receiver concept.

 

[mtrycrafts]

From Toole Audio Science Paper:

"The Harman International loudspeaker companies – JBL, Infinity, Harman/Kardon and Revel – have invested heavily in measurement facilities that allow them to take the fullest advantage of existing audio science."

I would think that their speakers would all sound outstanding, but that's not the case.

I have never listened them, but I would also think that the Revel speakers would sound superb, but some people said they don't even like them.

Some should, especially their speakers that are actively amplified at each driver.
But, doesn't mean all theirs would.

 

[no. 5]

From Toole Audio Science Paper:

"The Harman International loudspeaker companies – JBL, Infinity, Harman/Kardon and Revel – have invested heavily in measurement facilities that allow them to take the fullest advantage of existing audio science."

I would think that their speakers would all sound outstanding, but that's not the case.

I have never listened them, but I would also think that the Revel speakers would sound superb, but some people said they don't even like them.

That's from "Audio - Science in the Service of Art", right?

Remember that just because a maker has the measurement facilities does not mean that what is learned and known can be applied to every speaker in every model line; think of how much money it would take to make the low end models super linear, with ultra wide dispersion and zero resonance, cost prohibitive right? Add to that the need to sell a product; I'm guessing that if a typical consumer came accost a truly spectacular speaker playing pop music at Best Buy, they would say it sounds bad.

But that doesn't mean that they don't put any effort into it, ask WmAx what he has been able to do with first generation Primus drivers. As for not all of the models sounding outstanding, consider how many variables are at play; what was the room like? What was the listing material? Does the listener have any hearing loss? Was it a sighted comparison with another speaker? To name a few things.
Kevin Voecks once commented that "Looking at a good set of measurements we could say this is a good speaker, but the difference between a really good and really great speaker is still beyond our ability to absolutely predictably measure. So it’s that last level that we spend a lot of time on because it’s not as direct a path."

 

[tbewick]

I'm not surprised people don't like the Revel's. Here we go again with a design getting flat frequency response, that has to sound awful, because just the wrong set of compromises were made.

The research of Dr Toole and others show that a speaker's frequency response determines whether the speaker sounds good or not. At low frequencies, both time- and frequency-domain data need to be considered. I have seen test measurements of two Revel loudspeakers*. I do not have the expertise to analyse these data except in a superficial fashion. They appear to go along with the design-aims dictated by Harman's perceptual research.

* NRC measurements for SoundStage! of the Revel Concerta F12 and M12.
http://www.soundstagemagazine.com/measurements/revel_concerta_f12/
http://www.soundstagemagazine.com/measurements/revel_concerta_m12/

Now a fourth order passive crossover at 150 Hz, is just something, that if you know what you are doing, you don't even consider for a successful loudspeaker. For one thing the caps have such huge values, they will have to be poor sounding non polarizing electrolytic types. The inductors to avoid them being gigantic will be made of far too small a gauge of wire, wound on iron cores.

But to top it off the phase angle of the crossover is 360 degrees, so at 150 Hz the mid is arriving 6.7 msec ahead of the woofers! At 575 Hz the tweeter is 1.7 msec ahead of the mid. At 2.3 KHz the tweeter is 0.43 msec ahead of the mid.

The point being, is that this speaker if you measured its transient response would be awful. I can assure time smear of this magnitude, makes for a speaker you would never want to own!

If I were shopping for speakers, those would not even make the audition list.

Your criticism of the Revel speaker's crossover design assumes that the time-alignment errors are audible. If you can describe the perceptual research on which your criticisms are based, it would help to substantiate your argument.

 

[avaserfi]

Kevin Voecks once commented that "Looking at a good set of measurements we could say this is a good speaker, but the difference between a really good and really great speaker is still beyond our ability to absolutely predictably measure. So it’s that last level that we spend a lot of time on because it’s not as direct a path."

There no actual research that backs Mr. Voecks statement. This seems to be complete conjecture. With appropriate measurements it has been shown, through credible perceptual research, that listener preference for a loudspeaker can be determined presuming proper in room placement of both the loudspeaker and listener.

The research of Dr Toole and others show that a speaker's frequency response determines whether the speaker sounds good or not. At low frequencies, both time- and frequency-domain data need to be considered. I have seen test measurements of two Revel loudspeakers*. I do not have the expertise to analyse these data except in a superficial fashion. They appear to go along with the design-aims dictated by Harman's perceptual research.

The description of Toole's research is largely incomplete. There are far more variables that effect perceived loudspeaker performance as shown by the research performed in rigorous standards.

As a starting point I refer you to:

Loudspeaker Measurements and Their Relationship to Listener Preferences: Part 1. JAES Volume 34 Issue 4 pp. 227-235; April 1986.

Loudspeaker Measurements and Their Relationship to Listener Preferences: Part 2. JAES Volume 34 Issue 5 pp. 323-348; May 1986.

 

[WmAx]

The research of Dr Toole and others show that a speaker's frequency response determines whether the speaker sounds good or not.

This is not true. FR is but one factor. FR has to be considered at a very wide range of angles, analyzed in small increments, and plotted to a polar map or other such graph to analyze off axis response averages across a very wide window.

Even if the speaker is superb in this fashion, if it has a highly resonant cabinet system(most speakers) or drivers themselves are resonant, then the speaker can not meet Toole's definition of a high quality speaker system.

Also, the speaker must have certain minimum abilities in regards to SPL vs. distortion. It's no where near what many DIYers would lead you to believe is required - but it's still a basic requirement. For example: a 2 way 5" speaker will never produce sufficient dynamic output for dynamic recordings with 30 or 40Hz dynamic content. Such a design would be limited to very low SPL in full range mode, exceptions being somewhat contradicted if a specific and complex and expensive horn type bass loading was used on the small driver.

-Chris

 

[no. 5]

There no actual research that backs Mr. Voecks statement. This seems to be complete conjecture. With appropriate measurements it has been shown, through credible perceptual research, that listener preference for a loudspeaker can be determined presuming proper in room placement of both the loudspeaker and listener.

You may have read that too quickly... or I may be responding in ignorance, but here is the whole quote:

Because we actually have come up with measurements based on listening tests, it’s a full circle. We don’t do measurements because we have cool equipment to do it, we do them only because we found an audible characteristic that will differentiate two speakers or one that’s more refined.

So how do we measure that? How do we quantify that so that we can really engineer the speakers to perform better in that regard. So there’s extremely good correlation between the measurements that we do, the set of measurements, and the sound quality. Looking at a set of measurements, if they’re not good, we can say absolutely for sure that the speaker will not excel in listening tests. The converse is not completely true. Looking at a good set of measurements we could say this is a good speaker, but the difference between a really good and really great speaker is still beyond our ability to absolutely predictably measure. So it’s that last level that we spend a lot of time on because it’s not as direct a path.

 

[avaserfi]

You may have read that too quickly... or I may be responding in ignorance, but here is the whole quote:

My previous statement still stands. Lets take for example the article Differences in Performance and Preference of Trained vs Untrained Listeners in Speaker Tests. Olive, Sean. JAES., Vol. 51., No 9., 2003.

In this relatively recent example Olive compares a group of trained Harman listeners (n=12) and untrained listeners of various careers (n=256). Four high end speakers with similar retail costs were compared by these groups [including certain subgroups] using a pre-designed scale of comparison that was thoroughly explained to the test subjects. While the primary concern of this paper was differences between trained and untrained listeners it was also shown that even minor differences in measurable aspects of loudspeaker performance can and will dictate relative placement with regard to listener preference [without regard to listener training]. The largest problem found is that it takes an extremely skilled individual with an intimate understanding of this perceptual research to notice these slight differences.

It is important to note that these differences in rating were consistent between groups and subgroups with the largest difference being between group/subgroup variance.

These findings are consistent with Toole's [as well as others] previous research that used hundreds of listeners and approximately 40 loudspeakers spanning over two years.

 

[no. 5]

Please excuse me for continuing this, but still I don't think were on the same page, I believe what Mr. Voecks was expressing was the idea that appropriate measurements alone are currently insufficient to predict exactly where listener preference will fall - NOT that appropriate measurements don't predict and correlate to listener preference.