AV NIRVANA Speaker Evaluation Event - Tower Speakers $1200 or Less (Results)

[Matthew J Poes]

This really makes perfect sense. Dirac is able to fix some room related distortions and most speaker related distortions (linear distortions). However it can’t fix or change the speakers own peak output limitations, it’s own non-linear distortion, and most importantly its dispersion. Because it can’t fix dispersion it can’t fix how it interacts with the room above the Schroeder frequency. Yes it can fix phase and flatten the response, but it can’t change the response of the early reflections, their decay rate, etc.

I am curious to see measurements with Dirac across the speakers. I suspect that Dirac will flatten the response to a similar overall degree and this should give them a similar tonal balance. However differences in their harmonic distortion could change how that tonal balance is perceived.

I’ve also found that Dirac can’t really do a lot to tame a really harsh sounding speaker. Especially true when it’s a horn or waveguide speaker with a poor design (for instance diffraction slot designs).

It’s also good that Dirac can’t make all speakers sound the same. If it could then speaker designers would be out of a job and life would be a lot more boring.

And to honor a comment Earl Geddes made to me today about distortion, with complex musical signals there would be “in-harmonic anomalies”.

 

[AudiocRaver]

"In-harmonic anomalies." I like that.

 

[Matthew J Poes]

"In-harmonic anomalies." I like that.

Yep never heard the term before. I've since read a few distortion papers that described the complex nature of distortion with real musical signals. It appears that "things" arise for which we have no term and no real tests. Dr Geddes did develop a transfer function distortion metric which he said accounts for this, but it's not in use.

 

[Matthew J Poes]

I can't wait to see further reviews but quite frankly the change in response is shocking after Dirac. I used to see the same kinds of improvements but always questioned this.

To be sure I understand, those Dirac measurements are actually acoustic measurements and not predicted results from Dirac?

 

[bkeeler10]

Yep never heard the term before. I've since read a few distortion papers that described the complex nature of distortion with real musical signals. It appears that "things" arise for which we have no term and no real tests. Dr Geddes did develop a transfer function distortion metric which he said accounts for this, but it's not in use.

This is why I don't fully buy the notion that I've heard many times: "If you can hear it, you can measure it." And that audio can be strictly a scientific affair. Even if everything that can be heard can also be measured, that does not mean that we know how to take and/or interpret such measurements.

 

[Matthew J Poes]

This is why I don't fully buy the notion that I've heard many times: "If you can hear it, you can measure it." And that audio can be strictly a scientific affair. Even if everything that can be heard can also be measured, that does not mean that we know how to take and/or interpret such measurements.

I think they both are true. It's a mistake to say that we can measure sound quality fully. In fact we can't measure sound quality at all. We can measure parameters of a speaker which act as proxies for dimensions of sound quality, but a direct objective measure of sound quality does not exist. We can also measure a lot of things that don't impact sound. For example, no two speaker wires will measure exactly the same given sufficiently sensitive equipment. They might even change the sound of the speaker. That doesn't mean one is better than the other or that anyone would hear that difference. With amps you can (and often do) have amps with lower THD than others but which people find more offense. Some research into this phenomena found that, for example, the global feedback used to reduce THD primarily act on low order distortion, but raise high order distortion, which is known to have a lower threshold of audibility and to be more objectionable.

Speakers can do the same thing and these non-linearities are often thought to contribute to perceptions of warmth and detail. It's not uncommon for certain drivers to have more odd order distortion while others more even order, etc.

Dr Geddes paper on the Gedlee distortion metric (which uses a transfer function) had an example of a value for which the Thd was something like .2% but the Gedlee and IMD measures were much higher relative to other THD values. It turned out that this distortion was highly audible and the THD figure was meaningless.

I think there are some who would say we have better ways to scientifically evaluate sound quality but the industry is too lazy to accept it.

This shootout actually proves the point too. jBL developed a metric for sound quality based solely on linear distortion auralization technic. It explained 90% of sound quality differences in human subject trials. These speakers were all corrected with dirac such that they had virtually no linear distortion left. Yet Sonny said that they still sounded different. Then the logical explanation is that it's either un-correctable linear anomolies like directivity or it's non-linear distortion.

 

[Sonnie Parker]

To be sure I understand, those Dirac measurements are actually acoustic measurements and not predicted results from Dirac?

Yes... that is correct... actually measurements from our listening position.

This shootout actually proves the point too.

Just to clarify, this was not a shootout. We did not compare the speakers... each has it's own evaluation. I realize some readers will draw conclusions, but that was not our intention.

 

[Sonnie Parker]

The first post has been updated with new information and explanations... (I still have to add my song list).

The first evaluation is now posted: https://www.avnirvana.com/threads/a...peakers-1200-or-less-results.1640/#post-15058

We hope to get one every other day or two posted.

 

[bkeeler10]

Good to hear that the consensus on the Chane A5rx-c is similar to last time. A great speaker no doubt. Makes me wonder how much better the L series could really be . . .

 

[Matthew J Poes]

Good to hear that the consensus on the Chane A5rx-c is similar to last time. A great speaker no doubt. Makes me wonder how much better the L series could really be . . .

It might just be that it plays lower, louder, and maybe a bit less distortion in the low frequencies?

Given Chanes support for AV NIRVANA I assume a review will come in time.

 

[Todd Anderson]

Great section on the A5rx-c. I'd love to hear these speakers in action.

@Sonnie I liked your comment that they don't necessarily need processing to sound great. That bodes well for a huge slice of buyers!

 

[AudioThesis]

I'll be interested in your take on the S60. I was the first Polk dealer to get a shipment of the Signatures (I have connections ) and still have them in house.

 

[Sonnie Parker]

Given Chanes support for AV NIRVANA I assume a review will come in time.

Absolutely... we have first dibs on a review pair of L7's... no way we can pass those up. We'll probably review the new A5.4 as well. I can see the L7's in the front stage and the A5's for the surround in my room.

@Sonnie I liked your comment that they don't necessarily need processing to sound great. That bodes well for a huge slice of buyers!

Yep... one of the few that I've heard in my room that do pretty well without processing when out in the room.

I'll be interested in your take on the S60. I was the first Polk dealer to get a shipment of the Signatures (I have connections ) and still have them in house.

It's towards the end, but we'll eventually get it posted. Pretty interesting take from my perspective... interesting design with the down-firing port. I have not read what Wayne or Dennis wrote, so I am interested in seeing what they say about it as well.

 

[bkeeler10]

It might just be that it plays lower, louder, and maybe a bit less distortion in the low frequencies?

It ought to do that for sure, since it will be quite a bit bigger acoustically. However, it was not designed to simply be a bigger A5. But we still don't know much about it.

I'm sure the law of diminishing returns will apply when moving up from the A5 to the L7, though.

 

[Matthew J Poes]

It ought to do that for sure, since it will be quite a bit bigger acoustically. However, it was not designed to simply be a bigger A5. But we still don't know much about it.

I'm sure the law of diminishing returns will apply when moving up from the A5 to the L7, though.

I thought I read something about premium finishes and a new upgraded dual gap midbass drivers.

Any hint of upgraded crossover or tweeter? Sometimes when people get a second chance designing a speaker they use it as an excuse to further refine the crossover. Since this is a more premium speaker I could see adding a few more parts for a smoother response.

 

[bkeeler10]

What I understand is that it will have three or maybe four 6.5" midbass drivers in a sealed alignment, and the planar tweeter surrounded by two 5.25" midrange drivers up top. I imagine the crossover will be "upgraded," whatever that ends up meaning. A somewhat nicer finish than the A series to start, with perhaps more premium options later.

The woofers were originally to be Split-Gap, but so much has changed since then (it's been 3.5 years since this model was alluded to by Jon!) that I'm frankly not sure what's going on there now. It's definitely a big speaker, and it ought to be fairly sensitive and should carry a sense of ease and dynamic freedom, I expect. I'm a big believer in lots of radiating area, and the L7 will definitely have that.

 

[AudiocRaver]

I can't wait to see further reviews but quite frankly the change in response is shocking after Dirac. I used to see the same kinds of improvements but always questioned this.

To be sure I understand, those Dirac measurements are actually acoustic measurements and not predicted results from Dirac?

That is correct, the measurement plots shown are from measurements taken with Room EQ Wizard.

 

[AudiocRaver]

This is why I don't fully buy the notion that I've heard many times: "If you can hear it, you can measure it." And that audio can be strictly a scientific affair. Even if everything that can be heard can also be measured, that does not mean that we know how to take and/or interpret such measurements.

I agree, pretty much. We are getting better at indirect measurements, i.e., hearing something that can be explained by a seemingly-unrelated measurement. An example that I have been exploring recently is the clarity and openness of a soundstage at high frequencies, for which there is no direct measure, but which can be at least partially explained in many cases by multi-channel amp (or AVR) channel separation, or crosstalk, at high frequencies, which is rarely specified any more above 1 kHz for an AVR.

 

[AudiocRaver]

OK, a few final tweaks were made to the Chane A5rx-c post, mainly a little added descriptive detail.

 

[AudiocRaver]

I thought I read something about premium finishes and a new upgraded dual gap midbass drivers.

Any hint of upgraded crossover or tweeter? Sometimes when people get a second chance designing a speaker they use it as an excuse to further refine the crossover. Since this is a more premium speaker I could see adding a few more parts for a smoother response.

This happens a lot. One other speaker we have heard several times had tweeter design changes every time we heard it. We thought something was different, but not until we asked directly did we receive the admission. that changes had been made.

 

[Matthew J Poes]

I agree, pretty much. We are getting better at indirect measurements, i.e., hearing something that can be explained by a seemingly-unrelated measurement. An example that I have been exploring recently is the clarity and openness of a soundstage at high frequencies, for which there is no direct measure, but which can be at least partially explained in many cases by multi-channel amp (or AVR) channel separation, or crosstalk, at high frequencies, which is rarely specified any more above 1 kHz for an AVR.

That wouldn’t be a terribly difficult experiment to do. Measure a setups channel separation. Ensure it’s as high as possible. You could even use a dual mono setup end to end if need be to ensure maximum channel separation. Then artificially reduce crosstalk through cross feed on the computer. You could probably custom mix a set of test material with the crossfeed added so you can use. A group of three to four listeners and maybe ten trials each would be sufficient to characterize the effects.

There is a measure of spatial clarity used for performance spaces which I feel might be adaptable to a home environment. It requires a two channel recording of the impulse and that is a part of the problem, REW can’t be used for it. It uses two microphones in a dummy head or at a head spaced distance recording an impulse from the stage. My idea for adaptation is to have the L, C, and then R speaker each produce an impulse that you capture with the two mics. That’s 6 separate impulse measurements. You measure the ratio of early to late energy for each microphone. You then calculate the ratio of the ratio, which gives you the ratio of early to late energy heard by the left ear vs right ear for each channel. Higher amounts of direct sound in the correct air then characterize the room/system as having greater clarity in the soundstage. Better image delineation.

No idea of that would work or make sense, it was just a thought. My main concern in a small room is the short decay times.

 

[bkeeler10]

I agree, pretty much. We are getting better at indirect measurements, i.e., hearing something that can be explained by a seemingly-unrelated measurement. An example that I have been exploring recently is the clarity and openness of a soundstage at high frequencies, for which there is no direct measure, but which can be at least partially explained in many cases by multi-channel amp (or AVR) channel separation, or crosstalk, at high frequencies, which is rarely specified any more above 1 kHz for an AVR.

That's interesting. I suppose it is a given that less crosstalk between channels correlates to a clearer, more open soundstage. Perhaps this partially explains why people can hear differences between amplifiers that have ruler-flat frequency responses.

 

[Matthew J Poes]

What I understand is that it will have three or maybe four 6.5" midbass drivers in a sealed alignment, and the planar tweeter surrounded by two 5.25" midrange drivers up top. I imagine the crossover will be "upgraded," whatever that ends up meaning. A somewhat nicer finish than the A series to start, with perhaps more premium options later.

The woofers were originally to be Split-Gap, but so much has changed since then (it's been 3.5 years since this model was alluded to by Jon!) that I'm frankly not sure what's going on there now. It's definitely a big speaker, and it ought to be fairly sensitive and should carry a sense of ease and dynamic freedom, I expect. I'm a big believer in lots of radiating area, and the L7 will definitely have that.

Yes I agree but then I also come at this from a different ideal speaker design ideology (And my sense of normal is thus skewed). My speakers have 531 sq.cm of midrange per speaker. My bass section would be basicaly the same per channel.

It sounds like the L7's would have a tad under 500 sq.cm. for bass and 180-200 SQ.cm. for the midrange or so. Certainly not bad. More drivers is always nice because it brings down thermal compression, lowers distortion, and improves dispersion.

 

[bkeeler10]

Yes I agree but then I also come at this from a different ideal speaker design ideology (And my sense of normal is thus skewed). My speakers have 531 sq.cm of midrange per speaker. My bass section would be basicaly the same per channel.

It sounds like the L7's would have a tad under 500 sq.cm. for bass and 180-200 SQ.cm. for the midrange or so. Certainly not bad. More drivers is always nice because it brings down thermal compression, lowers distortion, and improves dispersion.

Might I ask what speakers you are running?

My calculations suggest about 214 sq cm per 6.5" driver and 140 sq cm per 5.25" driver. Did I miss something?

I hope we're not veering too far off-topic . . .

 

[Matthew J Poes]

Might I ask what speakers you are running?

My calculations suggest about 214 sq cm per 6.5" driver and 140 sq cm per 5.25" driver. Did I miss something?

I hope we're not veering too far off-topic . . .

I use Gedlee Abbeys which use 12" drivers in a two way design.

The SD of a driver never matches it's nominal size, which often includes the surround. The actual cone area of a 6.5" driver is typically more like 180-190 SQ cm at best. High Xmax versions with larger surrounds will have a smaller cone area. I went to parts express and found a similar driver and uses its come area for that calculation.