2,3,4,5-way speakers: how many ways is optimum?

[March Audio]

Its not BS, its an observation based on your generalisations, of which you have made a few now;

Class d all have timbre problems
Class D cant do bass
Full Active fails the "natural test" whatever that is
Many cant hear glare and timbre problems

These are the sort of comments I would expect to find in an audiophile forum, not one that looks more towards science rather than personal unsubstantiated opinion.

 

[j_j]

Irony, Ad Hominem attack and BS response noted ,


Good day

I will have to regard your claim as self-debunked, then, since you refuse to provide evidence. It is ironic that you have taken the Trumpian manoever of accusing your opponent of doing what you have already done. I would suggest that you find a way to block my comments, because you will find no tolerance whatsoever for such tactics from me.

As to your insult, do I need to send you the S&P address of the IEEE, or will you be retracting your accusation. Please answer clearly, and specifically. The answers that will work are "yes" and "no".

 

[amirm]

Let's take it down a few notch everyone. It is just a technical discussion. No need to create friction among each other.

 

[Thomas savage]

If you make sweeping generalisations/bold claims expect to be asked to back them up with evidence, if you can’t or simply can’t be arsed expect to be given short shrift here.

Please do not short circuit and resort to any kind of insult, it’s not what we want. There’s knowable people here who are to be respected, there’s a higher standard of proof if you want to challenge things here.

I welcome challenge and contrary POV but please if you can’t adear to the above keep ‘what you reckon’ in perspective and certainly don’t start lashing out in anyway.

Cheers all

 

[oivavoi]

Ok, allow me to play along here: What are the possible objective reasons that people might prefer passive over active crossovers?
The things I can imagine:
- a preference for the slight distortion that is introduced by many passive crossovers?
- do some electronic crossovers indeed muck up the signal?
- other things?

There are of course excellent passive speakers, just as there are horrible active speakers. But comparing apples to apples here, I wonder whether some people actually might show preferences for passive crossovers in a hypothetical A/B-comparison.

 

[Cosmik]

This article describes the complexity of passive crossover design, and some of the drawbacks such as performance that changes with power levels - and the short term history of power levels.
http://sound.whsites.net/lr-passive.htm

What I take away from it is the sheer effort of getting something that works, and how 'blunt' and approximate that solution is: it cannot be easily refined and modified, meaning that it is far more likely that the unfortunate designer will reach the stage of "that'll have to do" much earlier in the process than the active equivalent.

 

[Brad]

One thing that I think people forget is that some instruments like trumpet, saxophone, piccolo etc can sound unpleasant at times. Components that reduce dynamics and roll off high frequencies may make these more pleasant, and hence are preferred

 

[Cosmik]

On the general point of how many ways is optimum, would I be correct in thinking that for a given total driver surface area, the total height of the typical vertical column of drivers will be greater the more ways are introduced? i.e. it becomes less and less a 'point source'. This in itself could be a limiting factor..?

 

[A.wayne]

One thing that I think people forget is that some instruments like trumpet, saxophone, piccolo etc can sound unpleasant at times. Components that reduce dynamics and roll off high frequencies may make these more pleasant, and hence are preferred

Some of that reduced dynamics takes place in the recording Brad , dynamic compression is necessary for audio system playback due to the high crest factors ..

 

[A.wayne]

On the general point of how many ways is optimum, would I be correct in thinking that for a given total driver surface area, the total height of the typical vertical column of drivers will be greater the more ways are introduced? i.e. it becomes less and less a 'point source'. This in itself could be a limiting factor..?

Somewhat , it depends on the design goals , for an all out SOTA design best to use multiple cabinets , IMO unless you are using compression type Horn drivers , linesource works best over point source here ..

But yes you will end up with a pretty large monolith doing a
5 way setup, but also remember you lower thd and improve sensitivity when doubling driver area ..




Regards

 

[Cosmik]

Somewhat , it depends on the design goals , for an all out SOTA design best to use multiple cabinets , IMO unless you are using compression type Horn drivers , linesource works best over point source here ..

But yes you will end up with a pretty large monolith doing a
5 way setup, but also remember you lower thd and improve sensitivity when doubling driver area ..

What I'm saying is that for a three-way the distance between the centre of the tweeter and the mid-range will be 6", say, but if I increase it to an eight-way the distance between the tweeter and the centre of some of the drivers that are handling relatively high frequencies will be more like 18" or whatever. In other words, even at higher frequencies it is nothing like a point source, whereas for the three-way it pretty much is, because all the higher frequencies are handled by the one, small driver. It has a harder job to do, of course.

 

[amirm]

Ok, allow me to play along here: What are the possible objective reasons that people might prefer passive over active crossovers?

Because one design exist and the other does not. If I could buy the active version of my speakers I would. But they don't make it.

 

[j_j]

On the general point of how many ways is optimum, would I be correct in thinking that for a given total driver surface area, the total height of the typical vertical column of drivers will be greater the more ways are introduced? i.e. it becomes less and less a 'point source'. This in itself could be a limiting factor..?

It can be either good or bad, depending on how much attention you pay to driver frequency shading, array design (as well as crossover and such design), and the like. A speaker with a tight vertical dispersion pattern, ***if*** it's a broadband pattern, can help with imaging and timbre shift in many rooms.

One done wrong can be profoundly creepy.

 

[j_j]

Oh, and as to the 'n' way issue, this depends in great part on the drivers available, and what you want for a low-frequency limit, and at what level.

Any of a number of fantastic ribbon tweeters exist that will do 5k to dogwhistle at insanely high levels. A few that cost more will do 2K to beyond dog-whistle for more money, and ridiculous efficiency, on top of ridiculously high power handling.

Midranges are sometimes the reason for a 4 way, depending on the choice, their available maximum power, and their xmax.

But huge subs are the big issue, because they will cut off before you would want to use any sane-sized midrange (or pair thereof), and hence you need a "mid-woofer". But this really is only an issue if you want ridiculous levels t 15Hz.

 

[j_j]

Ok, allow me to play along here: What are the possible objective reasons that people might prefer passive over active crossovers?
The things I can imagine:
- a preference for the slight distortion that is introduced by many passive crossovers?

also:
Changes in frequency response with level due to driver and crossover nonlinearity.
Doppler issues for slow crossover slopes (12dB/octave is an absolute minimum for any mid driver, if you do the physics, and you're better off with more.
Problems in transient handling in inductors,
Impedance approximations gone wild
Zoebel requirements that vary with frequency
...

- do some electronic crossovers indeed muck up the signal?

Some do.
Many digital crossovers use poorly designed filters based on analog filters, denying something like 70 years of understanding in basic filter theory.
Some omit time delay between drivers.
Some do mistaken EQ for drivers (a lot, actually, do that!)
In short, digital crossovers are due an overhaul. I can't say more at the minute.

- other things?

There are of course excellent passive speakers, just as there are horrible active speakers. But comparing apples to apples here, I wonder whether some people actually might show preferences for passive crossovers in a hypothetical A/B-comparison.

Proving once and for all that no matter what method one uses, SOMEBODY will get it wrong. If there is a constant in life ...

 

[oivavoi]

(12dB/octave is an absolute minimum for any mid driver, if you do the physics, and you're better off with more.

My monitors from British AVI have 48 db/octave crossovers. Sounds super smooth, and it's impossible to detect that there's two drivers playing even sitting very close to the speaker. Not possible with passive crossovers, of course.

Many digital crossovers use poorly designed filters based on analog filters, denying something like 70 years of understanding in basic filter theory.
Some omit time delay between drivers.
Some do mistaken EQ for drivers (a lot, actually, do that!)
In short, digital crossovers are due an overhaul. I can't say more at the minute.

Hey! What's the deal with starting up on interesting explanations, and then just saying that you quit?
Especially curious about eq'ing drivers. Very common among active speaker manufacturers, but are there pitfalls here?

 

[j_j]

Especially curious about eq'ing drivers. Very common among active speaker manufacturers, but are there pitfalls here?

Well, yes, but what do you measure, and how do you correct what you measured, with what time window, etc? I've encountered a few that led to corrections bright enough to cause sunburn, so to speak. This is a problem intimately tied to the radiation pattern and direct response of a speaker, so I can't be terribly generic.

I guess I can say that our mid/high crossover, for instance, for one set of drivers, is 100dB down at 1500, for a 3000Hz -up passband. The mid is down 100 dB by 4500 Hz, too. Not so much opportunity for interference there.

 

[fas42]

My monitors from British AVI have 48 db/octave crossovers. Sounds super smooth, and it's impossible to detect that there's two drivers playing even sitting very close to the speaker. Not possible with passive crossovers, of course.

This is a description of the subjective experience of "good enough quality" - just enhance that capability so it occurs at any volume, and for any distance of of seating from the speaker.

The "of course" doesn't apply for passive crossovers - but it does require that the electronics, especially of the power amplifier, be of a higher order of competence for a passive crossover speaker to emulate that behaviour.

 

[Sergei]

Passive speaker designers traditionally find a three-way speaker troublesome - apparently - and many stick to two-way. For example:

In other words, the passive approach cannot make a good 3, 4, 5 or 6-way speaker regardless of cost.

I would say the passive approach can lead to a very good 3-way speaker, e.g. https://www.pioneerelectronics.com/pio/pe/images/portal/cit_11221/420272384TADR1OwnersManual0228.pdf.
But it would cost you.
https://www.technicalaudiodevices.com/reference-one/

And the problems with a two-way (active or passive) would appear to be many, including:

• drivers being stretched over wide frequency range leading to various types of distortion
• drivers driven too high begin to beam, in contrast to the next driver up, leading to a lumpy dispersion characteristic against frequency

Active systems, particularly those based on DSP would appear to give a more transparent crossover, potentially without phase rotation through the crossover. They can also provide time alignment with the appropriate delays.

Agreed. I use two-way monitors only for surround channels.

If a three-way active is better than a two-way, is a four-way therefore better than a three-way? Why not five-way or more? (putting aside the practical or cost aspects, but even then we might assume that multi-way DACs and amplifiers could be made reasonably cheaply, and that the drivers would have a much easier job to do so maybe their specs could be relaxed).

How about 18-way? https://www.bang-olufsen.com/en/speakers/beolab-90.