Sonic benefits of conventional vs coaxial design?

[Mort]

Conventional speakers (not coaxial) have engineering and cost benefits.

Is there any real audible benefit to coaxial? Where can that be seen in the measurements?

This is not about corner cases, just assume both are designed competently.

 

[voodooless]

Is there any real audible benefit to coaxial? Where can that be seen in the measurements?

The main advantage is in the off axis behavior. It’s generally the same in all directions, mainly dictated by the enclosure, and much less by driver location. So no lobing and interference artifacts. If done right one can output a phase coherent signal in a wide range of directions.

Just generate a 100 Hz square wave, play it on your system (not too loud!), and see the recoding of it via REW or similar tool (via microphone). On a normal speaker, the recording will look nothing like a square wave. And if it does, move your microphone, and look again

In a good coaxial, the recording will be very square wave like over a wide range of locations.

The real question is: how audible is this? Hard to say, but in general I would say: not always easy.

There are excellent normal speakers, and very bad coaxials.

 

[Beave]

In terms of measurements, the biggest difference is typically seen in the vertical off-axis plots.

Speakers with tweeters above or below midranges (or woofers) will have angles and frequency ranges at which the two drivers are out of phase and thus canceling out. You will see holes in the vertical off-axis plots, usually around crossover frequencies.

Coax speakers generally don't have this issue.

It's not clear how audible they are provided you're staying on axis. I would suspect it depends on the reflective behavior of your floor and ceiling.

 

[Mr. Swordfish]

The theory is that by co-locating the drivers you don't introduce phase anomalies around the crossover point due to path length differences.

In practice, those anomalies are not all that audible, and the extra cost of co-locating the drivers isn't worth the benefit. At least according to the designers of the hundreds of non coaxial speakers. YMMV.

 

[staticV3]

Another benefit is that midwoofer and tweeter will sum earlier.

Whereas traditional bookshelves need 1m listening distance or more, Coax designs work well below 1m as well.

 

[pablolie]

Another benefit is that midwoofer and tweeter will sum earlier.

Whereas traditional bookshelves need 1m listening distance or more, Coax designs work well below 1m as well.

I tend to prefer coaxial designs (from KEF). Seems to be an overall balance I am tuned into since the LS50 came out.

That said, 1m is not a lot of distance, even for desktop speakers - and positioning can overcome quite a few of those (mostly mathematical) limitations, in my listening experience.

 

[Mort]

Can we say anything about soundstage one vs the other?

 

[Flaesh]

Conventional speakers

IMO it's WG+15" 2-ways. Using a coaxial tweeter allows you to lower the crossover and avoid CD-comp equalization.
As far as I remember, you were recently interested in horn-speakers-under-7500-pair.

about soundstage one vs the other?

For me, soundstage is not the main value*. And I haven't used non-coaxial tweeters for a long time. I don't think there will be noticeable differences in the soundstage, all other things (horn\eq\cross\room) being equal.

* At the same time, I absolutely hate sound from a speaker; sounds should come from the air. The speakers just stand in the corners like weird, unnecessary furniture. I can move the speakers, sit on a chair, and listen to Chesky's samba walk around me, but it doesn't captivate me.

 

[Mort]

IMO it's WG+15" 2-ways. Using a coaxial tweeter allows you to lower the crossover and avoid CD-comp equalization.
As far as I remember, you were recently interested in horn-speakers-under-7500-pair.


For me, soundstage is not the main value. And I haven't used non-coaxial tweeters for a long time. I don't think there will be noticeable differences in the soundstage, all other things (horn\eq\cross\room) being equal.

Good memory, appreciate it. Yeah I am interested in a good pair of (probably used) horns at some point.
You prefer the WG+15"? Like the Dutch and Dutch monster? I guess I'd like to hear that too.

 

[bmc0]

In a good coaxial, the recording will be very square wave like over a wide range of locations.

Only if constant delay (linear phase) crossovers are employed. Most passive coaxial speakers will not produce something that looks much like a square wave at any location.

Can we say anything about soundstage one vs the other?

Probably not much. While vertical directivity matters to an extent, the effect seems to be mostly timbral rather than spatial. Horizontal directivity is the more important one for spatial quality.
I've heard several anecdotes claiming better rendering of spatial depth with virtual point source loudspeakers (coaxial, multiple entry horn, etc.), but I'm unaware of any substantive evidence.

 

[2020]

Another benefit is that midwoofer and tweeter will sum earlier.

Whereas traditional bookshelves need 1m listening distance or more, Coax designs work well below 1m as well.

I tried the Adam A8H, a conventional speaker that is marketed as both midfield and nearfield, with a different reasoning behind it.

The company considers it midfield because of it's power/projection potential (116db at 1m) and "conventional wisdom" says it's also midfield due to the 8" woofer that specs at -3db at 34hz. However, Adam also says it's nearfield because it's minimum listening distance is only 0.6m, which is quite small. So it is possible to go below the standard 1m.

I asked Gemini AI about this a while ago and it said in a perfect scenario, where there is no problems with room acoustics or placement, traditional design would win because it would be be perceived as larger and more powerful (dynamics). It also said that it was more "pure" because each driver does less and the benefits are maximized in the process (unlike coxial, where the midwoofer acts as a waveguide for the tweeter and then the tweeter is technically being modulated and this is distortion). It said coaxial's real benefit was imaging and ease of integration. Because it's naturally in phase and has point source properties, it's less likely to interact with the room in asymmetric or "ugly" ways. Also, it will likely have a larger sweet spot so it will be more forgiving or placement or the listener moving. It went on to say that traditional design's weakness are just real spaces, but with that taken out of the equation in our hypothetical, it doesn't really have any weaknesses. Coaxial then loses by default because it's meant to address something that's no longer a real issue (unless you are obsessed with phase and prioritize that first).

 

[Flaesh]

Dutch and Dutch monster

BTW I even don't know what's installed in 15C WG, coax or not?. It would be interesting to listen to, but I definitely won’t be able to buy it))
I used to be more interested in various coaxials. But KEF, Tannoy (I haven't heard the VQ series, I believe it's suitable)), Vienna, and some others coax' didn't make much of an impression.

P. Hales sells coaxial speakers\speakers with coax:

 

[voodooless]

Only if constant delay (linear phase) crossovers are employed. Most passive coaxial speakers will not produce something that looks much like a square wave at any location.

As I recall, the OG LS50 did a pretty good job.

 

[Mort]

I tried the Adam A8H, a conventional speaker that is marketed as both midfield and nearfield

That's always been weird to me since that's a wide directivity monitor.

 

[geol]

In hi-fi, coaxial drivers don't offer any particular advantage; in fact, I know very few good ones, except perhaps from PHL (expensive and sometimes requiring you to bring your own tweeter and crossover). Or Celestion, which makes excellent dual-cone drivers (I used a pair of K12H-200TCs to repair PA speakers, reinforced with two Monacor piezo tweeters after 10 kHz; it was very good, even for "hi-fi"). B&C, DAS, 18 Sound, and Faital also make good drivers, but they're even more expensive.
You have to choose carefully. Regarding listening parameters, such as directivity, you have to see if all of that is actually noticeable.

 

[Mort]

Vertical directivity issues can be huge if you are standing and moving around the room. it's audible with my ribbon tweeter with weak vertical.

That's not a common issue I would imagine.

 

[No_hair_left]

Can we say anything about soundstage one vs the other?

I'm a firm believer that coaxial speakers offer a better soundstage when used as dolby atmos ceiling speakers, probably because they have a much better/more even off axis response but there is something about using a point source speaker for that particular application that just seems to work over all the 2 way speakers I have listened to, sounds more cohesive, bigger.

 

[HalSF]

II asked Gemini AI about this a while ago and it said in a perfect scenario, where there is no problems with room acoustics or placement, traditional design would win because it would be be perceived as larger and more powerful (dynamics). It also said that it was more "pure" because each driver does less and the benefits are maximized in the process (unlike coxial, where the midwoofer acts as a waveguide for the tweeter and then the tweeter is technically being modulated and this is distortion). It said coaxial's real benefit was imaging and ease of integration. Because it's naturally in phase and has point source properties, it's less likely to interact with the room in asymmetric or "ugly" ways. Also, it will likely have a larger sweet spot so it will be more forgiving or placement or the listener moving. It went on to say that traditional design's weakness are just real spaces, but with that taken out of the equation in our hypothetical, it doesn't really have any weaknesses. Coaxial then loses by default because it's meant to address something that's no longer a real issue (unless you are obsessed with phase and prioritize that first).

I just can’t wrap my mind around treating AI regurgitation of scrapped and plagiarized bits and fragments of scattered human statements, mindlessly processed and reconstituted into unsubstantiated confident expressions of insight and synthesizing knowledge and judgment, as authoritative and sensible.

 

[bmc0]

As I recall, the OG LS50 did a pretty good job.

I guess it depends on what one wishes to call "pretty good." J.A.'s measurements for Stereophile indicate 180° excess phase from the crossover, typical of in-phase 2nd-order slopes. I don't know of any public 100Hz square wave measurements, so I made a rough model of the crossover and bass alignment (ref: ASR measurements). Plots of the model's magnitude and phase, step response (showing a reasonable match to J.A.'s measurement), and 100Hz square wave response (6 cycles):


Note: For clarity, I'm not suggesting that the excess phase is an audible problem in this case.

 

[jonfitch]

Imaging depth seems to be better in my experience with coaxials. Even with coaxials that don't seem to have very wide dispersion, it feels like there's better row separation between instruments. I think the uniformity of the dispersion probably creates some interesting effects with the reflections that you don't get with traditional dome and cone speakers.