Unusual Speaker Designs

[mhardy6647]

Ugly and dubious -- they'll sell like hotcakes.
(That sounds like Wilson's motto, come to think of it)

 

[Archsam]

Holding this to the standards of an $8k speaker, I think the design can be justifiably dismissed without measurements TBH. There's no beating the laws of physics.

It's a very different concept from the Uni-Q.

Both the KEF uni-Q driver and what Sony is trying to do is to emulate a single point source by aligning the tweeter and woofers on axis, in order to match the different drivers' directivity and improve the time coherence of the HF and LF. So in that sense their design concept is similar, though the engineering solution (and the advantage / disadvantage associated with their solution) is very different. KEF cleverly uses the woofer as a wave guide for their tweeter dispersion, for example, whereas Sony has to introduce two assist tweeters plus clever software to assist with wider directivity in the HF. How that translates into sound quality I have no idea, but there seems to be a logical engineering solution to achieve what they are trying to do.

You can have a look at KEF's explanation yourself:
http://www.kef.com/uploads/files/en/series_pdf/q_tech_explained_010910_en.pdf

I don't think it's fair to dismiss a concept simply by it's suggested retail cost. I agreed $8k is beyond ridiculous and 99.999999% of us will never buy it, and Sony probably knows that. That actually makes this whole effort kind of admirable because nobody will ever expect this to be a money making endeavour. I think Sony sees this as what it is - an R&D project that may or may not bear fruit, who knows but let's try and see.

I have seen far worse snake oil salesman out there.

 

[andreasmaaan]

Both the KEF uni-Q driver and what Sony is trying to do is to emulate a single point source by aligning the tweeter and woofers on axis, in order to match the different drivers' directivity and improve the time coherence of the HF and LF. So in that sense their design concept is similar, though the engineering solution (and the advantage / disadvantage associated with their solution) is very different. KEF cleverly uses the woofer as a wave guide for their tweeter dispersion, for example, whereas Sony has to introduce two assist tweeters plus clever software to assist with wider directivity in the HF. How that translates into sound quality I have no idea, but there seems to be a logical engineering solution to achieve what they are trying to do.

You can have a look at KEF's explanation yourself:
http://www.kef.com/uploads/files/en/series_pdf/q_tech_explained_010910_en.pdf

I don't think it's fair to dismiss a concept simply by it's suggested retail cost. I agreed $8k is beyond ridiculous and 99.999999% of us will never buy it, and Sony probably knows that. That actually makes this whole effort kind of admirable because nobody will ever expect this to be a money making endeavour. I think Sony sees this as what it is - an R&D project that may or may not bear fruit, who knows but let's try and see.

I have seen are far worse snake oil salesman out there.

You're correct about the way the Uni-Q works.

What your explanation has glossed over with the respect to the Sony, however, is that the output from the "assist tweeters", having a centre-to-centre distance of approximately 3.5cm, will interfere with each other destructively even small distances vertically off-axis.

Here's a graph showing the vertical off-axis response of two tweeters separated by a C2C distance of 3.5cm:

Even 15° off-axis, there is a major null in the top octave. 30° off-axis, the null has moved down to 10kHz. 45° off-axis it is at 7kHz, etc. etc. Why you would want this, I don't know.

Ofc, the other outrageous aspect here is that we have an $8k speaker that has a 10cm bass driver. No amount of engineering can make a driver that size play at even moderate levels in the bass without extreme distortion.

There are many other aspects that also smell off about this, although it can't be said with outright certainty that these other likely deficiencies haven't been overcome to some extent by clever engineering.

Concerning the two points mentioned above, however, there is simply no hope.

 

[Frank Dernie]

Both the KEF uni-Q driver and what Sony is trying to do is to emulate a single point source by aligning the tweeter and woofers on axis, in order to match the different drivers' directivity and improve the time coherence of the HF and LF. So in that sense their design concept is similar,

That is indeed what KEF are doing but unless Sony have a delay in a DSP crossover theirs will not have "time coherency" since the acoustic centre of the tweeters is way in front of the cone driver, it will be no better, time wise than a conventional speaker.
Integration for very near field listening will be better than a conventionally spaced 2 way.

 

[andreasmaaan]

That is indeed what KEF are doing but unless Sony have a delay in a DSP crossover theirs will not have "time coherency" since the acoustic centre of the tweeters is way in front of the cone driver, it will be no better, time wise than a conventional speaker.

Although it hardly excuses the other design choices, the Sony speaker does appear to have such a delay according to the marketing.

 

[Frank Dernie]

I really like the styling of these little Sonys but with a teensy main driver they won't be any use for bass at all or volume except very near field, so only any good as a near field speaker for a home studio perhaps, and for that they have excellent competition at a fraction of the price.

 

[Archsam]

You're correct about the way the Uni-Q works.

What your explanation has glossed over with the respect to the Sony, however, is that the output from the "assist tweeters", having a centre-to-centre distance of approximately 3.5cm, will interfere with each other destructively even small distances vertically off-axis.

Here's a graph showing the vertical off-axis response of two tweeters separated by a C2C distance of 3.5cm:

View attachment 64965

Even 15° off-axis, there is a major null in the top octave. 30° off-axis, the null has moved down to 10kHz. 45° off-axis it is at 7kHz, etc. etc. Why you would want this, I don't know.

Ofc, the other outrageous aspect here is that we have an $8k speaker that has a 10cm bass driver. No amount of engineering can make a driver that size play at even moderate levels in the bass without extreme distortion.

There are many other aspects that are also smell off about this, although it can't be said with outright certainty that these other likely deficiencies haven't been overcome to some extent by clever engineering.

Concerning the two points mentioned above, however, there is simply no hope.

Just a reminder that these speakers are designed for near field listening, they are never meant to push the SPL to even close to a room filling level. I remember reading an early reviewer's comment that the best way to describe these speakers is an experience similar to listening to headphones, but without the headphones on your head. And due to the short distance between the speakers and the listener, they can create a holographic soundstage that a pair of convention speakers, with the usually speaker / listener position, can never achieve.

The time delay associated with the 3 tweeters is addressed by a software relay, if I read Sony's explanation correctly.

Admittingly there are still some marketing BS in there, for example I don't buy for a second the need / feasibility for the extreme frequency response of the tweeters, 'natural sound all the way up to 100kHz' is just silly. What I am very curious about though, is how the engineers have come up with a different type of listening experience - if they can use these speakers to create a wrap around soundstage, with the details and clarify of a pair of high end headphones, but without putting headphones on, I think there is something very interesting going on here.

 

[Ilkless]

You're correct about the way the Uni-Q works.

What your explanation has glossed over with the respect to the Sony, however, is that the output from the "assist tweeters", having a centre-to-centre distance of approximately 3.5cm, will interfere with each other destructively even small distances vertically off-axis.

Here's a graph showing the vertical off-axis response of two tweeters separated by a C2C distance of 3.5cm:

Even 15° off-axis, there is a major null in the top octave. 30° off-axis, the null has moved down to 10kHz. 45° off-axis it is at 7kHz, etc. etc. Why you would want this, I don't know.

Ofc, the other outrageous aspect here is that we have an $8k speaker that has a 10cm bass driver. No amount of engineering can make a driver that size play at even moderate levels in the bass without extreme distortion.

There are many other aspects that also smell off about this, although it can't be said with outright certainty that these other likely deficiencies haven't been overcome to some extent by clever engineering.

Concerning the two points mentioned above, however, there is simply no hope.

In their (partial) defence, the first implementation of the triple tweeter array did this to control lobing:

The solution involved making the shared front plate slightly concave so that the central dome would be recessed a little compared with the assist domes. (Sony has applied for a patent for this.) In conjunction with the spacing of the three tweeters, making the crossover point for the assist tweeters a little higher than that of the primary tweeter, and the assist tweeters' lower sensitivity, this would eliminate lobing in the vertical plane.

https://www.stereophile.com/content/sony-ss-na2es-loudspeaker-measurements

Not the best measurements, but Atkinson notes no top-octave lobing. Nonetheless, tiny midwoofer and tiny tweeters does not inspire confidence. The array cannot possibly be crossed low enough to minimise diffraction (as in the case of Geithain) off the array frame.

 

[Ron Texas]

Some well known odd speakers which look like they were brought back from the future.

A well known vintage model for those old enough to remember.

 

[andreasmaaan]

In their (partial) defence, the first implementation of the triple tweeter array did this to control lobing:

I appreciate what they say they are attempting to do, but if this is "controlled lobing", then I'm a leprechaun

 

[Frank Dernie]

Just a reminder that these speakers are designed for near field listening, they are never meant to push the SPL to even close to a room filling level. I remember reading an early reviewer's comment that the best way to describe these speakers is an experience similar to listening to headphones, but without the headphones on your head. And due to the short distance between the speakers and the listener, they can create a holographic soundstage that a pair of convention speakers, with the usually speaker / listener position, can never achieve.

The time delay associated with the 3 tweeters is addressed by a software relay, if I read Sony's explanation correctly.

Admittingly there are still some marketing BS in there, for example I don't buy for a second the need / feasibility for the extreme frequency response of the tweeters, 'natural sound all the way up to 100kHz' is just silly. What I am very curious about though, is how the engineers have come up with a different type of listening experience - if they can use these speakers to create a wrap around soundstage, with the details and clarify of a pair of high end headphones, but without putting headphones on, I think there is something very interesting going on here.

Yes, but I like bass because I listen to a lot of orchestral music and the first octave is entirely absent from these, swapped for a couple of inaudible octaves up top as you say.
In my study I can choose between headphones or normal monitors 3 or 4 meters away.
In terms of sound stage in the nearfield I don't see why these would do anything more than the Genelec "ones".

 

[Kal Rubinson]

Unlike the Stereophile people, I won't make excuses for the thinness (they claimed it cleared up with burn-in...

People? They? Was it more than one?

 

[Ilkless]

People? They? Was it more than one?

I misspoke, apologies. The burn-in talk was just letting the Lansche off too lightly (pun not intended), in my view. Despite the voicing, I maintain it is a stunningly good transducer and one of the few megabuck speakers (up there with Vivid and maybe the new Magico A5) that offer features, engineering and build that reflect the price level.

 

[Kal Rubinson]

I misspoke, apologies.

That is OK. I tend to reply as I did when anyone criticizes a statement made by a Stereophile writer and attributes it to many or all of us. A sensitive point for me.

 

[tuga]

View attachment 64151

Look like they came out of a comic book.

 

[Archsam]

Yes, but I like bass because I listen to a lot of orchestral music and the first octave is entirely absent from these, swapped for a couple of inaudible octaves up top as you say.
In my study I can choose between headphones or normal monitors 3 or 4 meters away.
In terms of sound stage in the nearfield I don't see why these would do anything more than the Genelec "ones".

I'm a headphone guy as well, my current daily driver is a pair of Focal Clear. I have a pair of new Audeze LCD-X coming in 2 weeks time so that I can experience their famous bass quality In headphones we all know that a (relatively) small dynamic driver can produce bass that can rival a good sub-woofer in quality and quantity, within the confines of the ear pads / within your ear canals in case of IEM, so no question there really.

I am very familiar with the quality and level of detail that a great pair of headphone can produce, and this is not something my speaker based system can match, and I don't exactly have bad gears either - have a look at my signature for what I currently have. It will cost crazy money on a hifi system to do what a great headphone can do.

But the experience of listening to music from headphones is never the same as with speakers. The holographic illusion of a singer front and centre to the listener can only be created with the complex interaction between two sound sources and our own two ears. I am no expert in physioacoustics so I will stop speaking right here

My hunch is the engineers from Sony are trying to combine the best of the headphone experience with speakers. Imagine the level of details and microdynamics we get out of good headphones, combined with the spatial soundstage of speakers?

If they can really pull it off, I will be the guy at the front of the queue, money in hand, and be the first one to buy the THIRD GENERATION of these speakers (when they become affordable)

 

[Ilkless]

I appreciate what they say they are attempting to do, but if this is "controlled lobing", then I'm a leprechaun

4.4kHz is the XO point, so not unusual, and the stuff above 10kHz you circled looks more like a relatively smooth roll-off. I'm not defending the approach. I think it accomplishes little, but it does have less lobing and raggedness than one might expect from such a configuration. IMHO the much bigger worry is diffraction. Unless the array drivers are actually small high-excursion widebanders (think Aurasound NS3) crossed very low, there's no hope of working around diffraction of the array mounting.

 

[rebbiputzmaker]

The obelisk. Met Dick Shahinian many many moons ago, interesting guy.

 

[andreasmaaan]

4.4kHz is the XO point, so not unusual, and the stuff above 10kHz you circled looks more like a relatively smooth roll-off.

Keep in mind that JA only measures out to +/- 15°. The stuff above 10kHz is not off-axis roll-off, but the null predicted in the graph I posted earlier:

Note how the 15° off-axis response goes down and then up, which is the giveaway:

We can infer that the vertical polars follow the rest of the modelled responses out beyond +/- 15° (if only JA had measured out that far).

And yes, the null at 4.4kHz is due to the crossover, but is a c. -12dB null 5° off-axis really acceptable, let alone optimal?

 

[QMuse]

Keep in mind that JA only measures out to +/- 15°. The stuff above 10kHz is not off-axis roll-off, but the null predicted in the graph I posted earlier:

View attachment 64981

Note how the 15° off-axis response goes down and then up, which is the giveaway:

View attachment 64983

We can infer that the vertical polars follow the rest of the modelled responses out beyond +/- 15° (if only JA had measured out that far).

And yes, the null at 4.4kHz is due to the crossover, but is a c. -12dB null 5° off-axis really acceptable, let alone optimal?

That particular null is relevant only to the bats and cats.