Hot takes

[carlsnor]

Hey, long-time lurker, first-time poster here.

I was wondering if you all have any (educated) unpopular opinions. Here’s mine:
All non-coaxial speakers are by design a compromise(or better said flawed), therefore there is no reason to spend premium money on them!

Change my mind, or drop your own hot takes below!

 

[levimax]

Hey, long-time lurker, first-time poster here.

I was wondering if you all have any (educated) unpopular opinions. Here’s mine:
All non-coaxial speakers are by design a compromise, therefore there is no reason to spend premium money on them!

Change my mind, or drop your own hot takes below!

Over simplified, all speakers at any price are a compromise. Coaxial solves some problems i.e. it is a point source, but introduces other issues and limitations due to the physical realities of having 2 speakers on top of each other.

It is best to look at the measurements as a whole and not just focus on one design element. There is more than one way to design a excellent speaker.

 

[carlsnor]

Over simplified, all speakers at any price are a compromise. Coaxial solves some problems i.e. it is a point source, but introduces other issues and limitations due to the physical realities of having 2 speakers on top of each other.

It is best to look at the measurements as a whole and not just focus on one design element. There is more than one way to design a excellent speaker.

I agree every speaker is a compromise. Maybe I should have used the word "flawed" instead.
If you mean SPL capabilities by "limitations," I wouldn't agree. Measurements of even entry-level coaxial speakers have shown that compression in speakers like the Q150 is negligible, and this should have improved even more with the newer Q1 Meta.

By 'non-premium speakers and systems,' I mean something like the Primus 150 or 305p paired with their respective subwoofers. So basically everything under $1k, to give a conservative guess, maybe even as low as $500 for a Primus set on the used market.

 

[Redacted]

All coaxial speakers are just a different compromise.

 

[Jim Taylor]

All loudspeakers, whether they are coaxial or not, depend not simply on their native design limitations but also on the manner in which designers deal with those limitations. Coaxial speakers have long been seen as having high intermodulation distortion. By the use of new materials and adhesives, new and better analysis software and better surround and cone designs, the negative characteristics of coaxial designs have been greatly reduced.
The assessment can then be stated in a manner similar to amplifiers: are the flaws audible? If they are not audible to the listener, then is further argument not moot?

 

[wwenze]

People can't even decide what type of dispersion pattern and boundary loading is "the best", you think they know if coaxial would be better than normal 2-way in a typical scenario. And that's before even talking about the performance disadvantages of going coaxial, which even KEF doesn't completely get rid of.

You use coaxial in situations to solve problems that require coaxial to solve, that's it.

Waveguideless and uncontrolled wide treble, give me that sweet room reflected sound baby.

 

[wwenze]

All non-coaxial speakers are by design a compromise, therefore there is no reason to spend premium money on them!

Change my mind, or drop your own hot takes below!

The concept of convergent evolution, or in other words, different people reach the same solution.

The first speakers were full-range. The cheap speakers still are. And people realized they need improvement. Thus two-way speakers were born.

The ideal sound source would psychokinetically project vectors into the space in your room. But we don't even record and save this data to begin with (Where's mah object-based audio). The ideal sound source that would work with only a scalar value is a point source, but if that worked in reality a speaker would be the size of IEMs. We don't trust dome tweeters to handle anything below 2kHz, so we need... a woofer. Here we go again.

You are right, these are design compromises, just like how an amplifier with 100 transistors is a compromise because real transistors don't follow the perfect transistor equations which means an amplifier made of 1 transistor would measure and sound horrible. Engineering is finding how to fix nature's imperfections.

 

[kemmler3D]

I have LS60s at home so I'm pretty pro-coaxial but I don't regard them as universally superior either. You give up cone area and increase certain types of distortion so you can get better dispersion but it's not magic.

However, if you changed the question to "is a design without directivity control (waveguide, horn, coax, etc.) a deal-breaker, I would answer yes.

 

[illusonic]

I have LS60s at home so I'm pretty pro-coaxial but I don't regard them as universally superior either. You give up cone area and increase certain types of distortion so you can get better dispersion but it's not magic.

However, if you changed the question to "is a design without directivity control (waveguide, horn, coax, etc.) a deal-breaker, I would answer yes.

Taking the LS60 as an example, the midrange seems to be the least limited frequency range. It's the bass/tweeter that are limited at 86/96dB SPL.
IMD looks good for midrange and treble, too. Only THD is a bit high with around 1% at 96dB but still acceptable. Could be improved with a larger speaker. Coax size in LS60 is limited by enclosure width. So a visual design factor not a technical necessity.

While dispersion characteristics require some attention to detail to get right, distortion isn't an obvious limitation of coax speakers. The coax in the Orbits runs down to 250Hz including cardioid losses. In this configuration it also shows about 1% THD below 300Hz at 96dB (Stoneeh unechoic measurement). IMD looks good for this coax, too. The same speaker without cardioid losses would perform much better. And crossing over to woofers at 250Hz will not create lobing. So again, I see no general distortion issue for coax speakers. But I do see a general vertical lobing issue with non coax speakers.

 

[dfuller]

All non-coaxial speakers are by design a compromise(or better said flawed), therefore there is no reason to spend premium money on them!

This is, to be blunt, total nonsense. Everything is a tradeoff. You're trading vertical dispersion evenness for SPL handling (mostly a concern in 2-ways, to be clear) and driver complexity. Plus, designing performant coaxials is, to be blunt, quite difficult. That's why it's basically only Genelec and KEF doing it well, and everybody else isn't.

Further - narrow vertical dispersion isn't necessarily a bad thing.

drop your own hot takes below!

Passive speakers are antiquated and pointless and only serve to enhance audiophilia nervosa about amp "synergy".

 

[Redacted]

Passive speakers are antiquated and pointless and only serve to enhance audiophilia nervosa about amp "synergy".

Last time I looked, all speakers needed an amp either in the cabinet, or an amp a couple of metres away via a cable, and may have an active or passive crossover - I'm not sure of the subtlety of the point about "antiquated and pointless" - what is not antiquated right now?

 

[kemmler3D]

distortion isn't an obvious limitation of coax speakers.

It is though, you'll always have higher IMD / Doppler than an equivalent non-coax, it's inherent in the concept, and THD will tend to be higher, all else held equal, due to the loss in cone area.

I really like my LS60s but I'm not going to pretend distortion at -40 is amazing for the money. You can get a SPK16 kit at -60 for less. What it is, it's an acceptable trade-off.

 

[dfuller]

Last time I looked, all speakers needed an amp either in the cabinet, or an amp a couple of metres away via a cable, and may have an active or passive crossover - I'm not sure of the subtlety of the point about "antiquated and pointless" - what is not antiquated right now?

All speakers need an amp. But we don't need to be wasting amp power through a passive crossover - insertion loss is usually on the order of 1-2dB, with really bad ones being as high as 3dB. Filters are more precise and more accurate done at line level or in digital in almost every circumstance, plus they allow much more flexibility to meet a desired response.

 

[Cbdb2]

You are right, these are design compromises, just like how an amplifier with 100 transistors is a compromise because real transistors don't follow the perfect transistor equations which means an amplifier made of 1 transistor would measure and sound horrible. Engineering is finding how to fix nature's imperfections.

Heres a single transistor amplifier with a thd of .2%. Shows the miracle of feedback.

Simplest possible 1-transistor power amplifier with THD <1%

Simplest possible 1-transistor power amplifier with THD

www.diyaudio.com

 

[illusonic]

It is though, you'll always have higher IMD / Doppler than an equivalent non-coax, it's inherent in the concept, and THD will tend to be higher, all else held equal, due to the loss in cone area.

Let's look at the measurements. Where is the IMD in the LS60? It's far below 10%. See IMD for a 2-way speaker. It's approaching 10%. The membrane area lost for the tweeter is comparably small and there is no reason not to increase external diameter to gain back any small loss in area.

I really like my LS60s but I'm not going to pretend distortion at -40 is amazing for the money. You can get a SPK16 kit at -60 for less. What it is, it's an acceptable trade-off.

If Purifi designs the coax, we will get -60dB THD, too. It's a matter of motor optimization and not inherent to coaxial designs.

 

[dfuller]

Let's look at the measurements. Where is the IMD in the LS60? It's far below 10%. See IMD for a 2-way speaker. It's approaching 10%. The membrane area lost for the tweeter is comparably small and there is no reason not to increase external diameter to gain back any small loss in area.

View attachment 529655

If Purifi designs the coax, we will get -60dB THD, too. It's a matter of motor optimization and not inherent to coaxial designs.

Let's not compare a 2-way to 3-way. If we compare a 2-way coax, it is monumentally worse than a non-coax design in that department.

And if we compare to a high performance, low distortion non-coax 3-way...

 

[illusonic]

And if we compare to a high performance, low distortion non-coax 3-way...

... we obviously also have to take a high performance 3-way with coax and get simiar performance.

 

[thewas]

My feeling is that in the last years with advanced engineered 3 or more way designs coaxials like from Genelec and KEF often offer one less compromise than typical non-coincidential designs which especially is important to nearfield and desktop or console listening (due to the close desktop or console reflections) since distortion isn't really an issue on them. In larger listening distances this advantage becomes less important though in my experience. One should add though that nowadays due to low crossable tweeters and advanced pseudocoaxial or line-array designs it is possible to design loudspeakers with smooth wide horizontal and narrow vertical radiation which can be advantageous at high listening distances and/or reflective living rooms, something which I currently have not seen yet with a coaxial driver.

 

[Ellebob]

It spomewhat depends on the room as which will be better. As a generalization most coaxials don't have good dispersion much past 60 degrees, usually in the 50 degree range is typical. You do however get it both horizontally and vertically. Non-coaxial speakers can have good off axis dispersion past 60 or even 70 degrees in some design, dep[ends on the design. Of course their vertical is usually abput 20 degrees. If you have carpet on the floor and some type of aborption on the ceiling you might prefer some non-coaxial speakers for their wider soundstage presentation. Lots of factors and each design has its tradeoffs.

 

[illusonic]

One should add though that nowadays due to low crossable tweeters and advanced pseudocoaxial or line-array designs it is possible to design loudspeakers with smooth wide horizontal and narrow vertical radiation which can be advantageous at high listening distances and/or reflective living rooms, something which I currently have not seen yet with a coaxial driver.

Definitely a fascinating speaker although I'm not sure I would like such narrow vertical dispersion. Might change my mind once I listen to it. I guess it's basically a tradeoff between large distant soundstage and a more detailed but closer and probably smaller sound image. After all, floor reflections help to determine distance.
Adjustable directivity would be a nice feature. Besides omni radiation, the SL1 also had at least two different beamwidths and directional beam steering in the horizontal plane, all controlled from an app. In a comparison of conventional and coax speakers, beamforming concepts are a separate category imho.