Dispersion, narrow, controlled, or bi?

At the risk of over-simplifying, imo a speaker should get two things right:  The direct sound, and the reflections. 

The direct sound is the most important, but imo the reflection field matters enough that it’s worth getting right (unless the system is set up for near-field listening). 

What constitutes "reflections done right" is an area where there are differences of opinion.  Imo the reflection field should be spectrally very similar to the direct sound; it should be neither too strong nor too weak; it should begin arriving neither too soon nor too late (the latter is virtually never an issue in a home audio setting); and the arrival directions matter, particularly for the first reflections.  

(In addition, the reflection field should decay evenly across the spectrum, and should decay neither too fast nor too slow, but this is a room acoustics issue over which the loudspeaker itself has little influence.)

My preference is generally for controlled-pattern multi-directional speakers, assuming they can be set up correctly in the room.   Dipoles and bipoles, for instance, need to be far enough out into the room.

I have been commercially involved with fullrange dipoles, fullrange bipoles, dipole/cardioid hybrids, and speakers with rear-firing drivers which are not full-range bipoles.  Imo each of these formats have worthwhile advantages, but there are always tradeoffs.

Most recently I have been involved with speakers which have relatively narrow frontal radiation patterns plus highly adjustable rear-firing drivers.  The general idea is to manipulate the reflection field independent of the direct sound, with the intention of making the reflection field more correct that it could have been with front-firing drivers only.  

When both the direct sound and the reflection field are "done right" (and obviously the room gets a vote in the latter), instrumental timbre is natural-sounding; listening fatigue is minimized or eliminated; and we are more likely to get a "you are there" spatial presentation, wherein the acoustic space on the recording is perceptually dominant over the playback room’s "small room signature". 

In my opinion.

Duke

dealer/manufacturer

I have to agree with Duke that the rear reflections matter in imaging.  We tend to focus on absorption but the rear needs a balance of absorption and diffuse reflection to really lock in that sense of presence. 

Due to my own circumstances, I'm really against the idea of adding more speakers in terms of floor space, but I've often wondered what the results would be of using a separate preamp/amp/speaker pair to reflect sound would be like.   You could completely alter the volume and even use DSP to tailor the reflections. 

@erik_squires wrote:  "... I’ve often wondered what the results would be of using a separate preamp/amp/speaker pair to reflect sound would be like.   You could completely alter the volume and even use DSP to tailor the reflections. "

If  you decide to try that some day, note that (unlike your main speakers) it is the POWER RESPONSE of the dedicated-to-reflections speakers that matters most.  You could even EQ them such that their power response "zigs" where the main speaker’s off-axis response "zags". 

@erik_squires again:  "I’m really against the idea of adding more speakers in terms of floor space"

Ime the arrival time of the additional reflection energy matters more than the arrival direction, so you might try using the vertical!  Maybe you could place small dedicated-to-reflections speakers on the floor, facing up, behind the main speakers. Often there’s some unused floor space behind the main speakers anyway.

Also, I suggest that you "shield" the dedicated-to-reflections speakers so that they cannot be seen from the listening area.  Not that seeing them is bad, but they will have some radiation to their sides, and you want to block that off because otherwise it will arrive too early and will probably have a poor spectral content.  If your main speakers are floorstanders, their width may well be adequate. 

Duke

Thanks for the suggestions, Duke, but honestly in terms of money and time the entire idea is well beyond me right now.  I was however thinking of a number of passive speakers that have implemented smallish rear firing drivers (like a tweeter or 2-3" mid) and think perhaps in the future a wave of active speakers could do the idea more justice.   Having a smallish amp and DSP settings for the rear firing may make for very interesting customization. 

I've never heard them, but the Steinway/Lyngdorf (at least some of their models) are powered dipoles with DSP.  They look quite interesting.

I prefer a large radiation area and front firing, tall horn or horn hybrid design with a controlled, narrow and fairly uniform dispersion pattern. Smaller and lower horn-based designs tend to be too "beamy" and upfront sounding for my taste and more easily give off themselves as speakers than a more uninhibited field of sound.

Multi directional speakers can have an advantage with a more spacious soundstage, but generally I find them too diffuse sounding - indeed almost too or even annoyingly spacious (in which case they mayn’t have been well implemented).

A front firing only, tall and large radiation area horn-based design as described above to my ears compensates immersively through the sheer size of its radiation field with a more distinct, yet still relaxed and fuller presentation - a good, a rather natural sounding compromise, I find. 

@audiokinesis wrote:

[...] note that (unlike your main speakers) it is the POWER RESPONSE of the dedicated-to-reflections speakers that matters most.

Why isn’t the power response of the main speakers of a primary concern here? Introducing a reflective sound field "actively" with additional speakers is also an additional measure to get right. Isn’t that a degree of complexity that can invite more problems than what it tries to solve?

@phusis wrote:  "I prefer a large radiation area and front firing, tall horn or horn hybrid design with a controlled, narrow and fairly uniform dispersion pattern."

Me too!  But I haven’t done anything as large as your speakers (yet), so you are getting good radiation pattern control down lower than I am.  What are the radiation patterns of your big horn and tweeter, if you don’t mind?

Phusis again:  "Why isn’t the power response of the main speakers of a primary concern here?"

Well I suppose the NET in-room power response is the primary concern, and what I was suggesting to @erik_squires was a way of CORRECTING the in-room power response by adding correctively-EQ’d speakers whose response arrived late enough that it was only contributing to the in-room reflection field.  To put it another way, the main speakers’ power response is essentially unchangeable without also changing the direct sound, so if we want to leave the direct sound unchanged but improve the in-room reflection field, adding a pair of dedicated-to-reflections speakers is one way to do that.

Phusis once more:  "Introducing a reflective sound field "actively" with additional speakers is also an additional measure to get right. Isn’t that a degree of complexity that can invite more problems than what it tries to solve?"

Yes!  That’s why I was making a few suggestions to Erik in response to him having "often wondered what the results would be of using a separate preamp/amp/speaker pair to reflect sound." 

Tying back in to the topic of this thread, "dispersion":  Reflection-field-correcting rear-firing drivers are something I’ve been doing passively, as an integral part of the loudspeaker system design, for many years. 

@audiokinesis wrote:

Me too!  But I haven’t done anything as large as your speakers (yet), so you are getting good radiation pattern control down lower than I am.  What are the radiation patterns of your big horn and tweeter, if you don’t mind?

The large format Constant Directivity EV HP9040 diffraction horn (as the model number implies) has a 90 and 40 degree horizontal and vertical coverage (down to about 500Hz, it seems), and the dispersion pattern of the JBL 2405 Alnico tweeter sits at 90 and 30 degrees horizontally and vertically - in both cases according to the specs linked.  

Well I suppose the NET in-room power response is the primary concern, and what I was suggesting to @erik_squires was a way of CORRECTING the in-room power response by adding correctively-EQ’d speakers whose response arrived late enough that it was only contributing to the in-room reflection field.  To put it another way, the main speakers’ power response is essentially unchangeable without also changing the direct sound, so if we want to leave the direct sound unchanged but improve the in-room reflection field, adding a pair of dedicated-to-reflections speakers is one way to do that.

I see. So these measures being corrective are a compensation for what the main speakers are typically able to, or rather incapable of doing in and by themselves. In principle however it ultimately follows - or so it could be deduced - that the power response of that produced by the main speakers alone DOES matter as a primary factor - insofar it can be uniformly achieved in a given listening space?

From your chair: is it even possible for main speakers as standalone units to get the NET in-room power response right (front firing, di- or bipole or otherwise), or is this a matter that can be at least partially alleviated with the design approach (i.e.: from the likes of horns) and through sheer physics/size as I implied earlier? 

Tying back in to the topic of this thread, "dispersion":  Reflection-field-correcting rear-firing drivers are something I’ve been doing passively, as an integral part of the loudspeaker system design, for many years. 

Why isn’t this a more widely addressed field of concern from other/more speaker manufacturers?

What I meant by "actively" (i.e.: not how the speakers are configured crossover-wise) was to point to the corrective measures being done additionally with extra speakers to aid the reflective field as an actual, actively addressed area by its designer.  

@phusis wrote:  "The large format Constant Directivity EV HP9040 diffraction horn (as the model number implies) has a 90 and 40 degree horizontal and vertical coverage (down to about 500Hz, it seems), and the dispersion pattern of the JBL 2405 Alnico tweeter sits at 90 and 30 degrees horizontally and vertically - in both cases according to the specs linked."

Sweet! 

I really like your radiation pattern choice, it’s wide enough and consistent enough to give very good coverage of the listening area, but not so wide that you have an excess of early-onset reflections. And you will have a good directivity match between your woofers and the horn in the crossover region. Those big woofers in their bigfaced box will have useful pattern control probably down to the Schroeder frequency just from their sheer size.

@phusis:  "So these measures being corrective are a compensation for what the main speakers are typically able to, or rather incapable of doing in and by themselves."

Yes!

@phusis:  "In principle however it ultimately follows - or so it could be deduced - that the power response of that produced by the main speakers alone DOES matter as a primary factor - insofar it can be uniformly achieved in a given listening space?"

Yes. There are definitely limits on how much correction one can do to the in-room power response via secondary rear-firing drivers.

@phusis:  "From your chair: is it even possible for main speakers as standalone units to get the NET in-room power response right (front firing, di- or bipole or otherwise), or is this a matter that can be at least partially alleviated with the design approach (i.e.: from the likes of horns) and through sheer physics/size as I implied earlier?"

Absolutely!   There definitely are other approaches than the one I use which “get the NET in-room power response right”, and imo YOUR approach (big horns and big woofers) is one of them!

Imo there can also be a spatial quality benefit arising from the contribution of rear-firing drivers, even if "correction" of the reflection field’s spectral balance isn’t really needed.

@phusis:  "Why isn’t this a more widely addressed field of concern from other/more speaker manufacturers?"

Manufacturers tend to embrace paradigms which they have found to work well, and presumably for most of them rear-firing drivers don’t offer something they place a high priority on.

For instance, most manufacturers make cone-n-dome speakers that have wide radiation patterns, so they are already getting a lot of reflection energy out into the room. They may not see much potential benefit to adding more reflection energy even if it is corrective.  (The idea that rear-firing energy CAN be corrective may not be very widespread.)

And some manufacturers deliberately have rather narrow radiation patterns, which results in a high direct-to-reverberant sound ratio. For these manufacturers, reducing the direct-to-reverberant ratio via the addition of rear-firing drivers probably seems counter-productive.

Then there’s the question of cost/benefit: Even if there is some benefit from adding rear-firing drivers, could the same amount of money spent elsewhere make a bigger improvement?  And, perhaps the most practical question:  Will having rear-firing drivers make a worthwhile difference in sales?

My own paradigm goes something like this: “A speaker should get two things right: The direct sound, and the reflections”. Getting the reflections “right” is the hard part, and imo it’s worth a fair amount of effort, but clearly I’m in the minority.

@phusis:  "What I meant by "actively" (i.e.: not how the speakers are configured crossover-wise) was to point to the corrective measures being done additionally with extra speakers to aid the reflective field as an actual, actively addressed area by its designer.  

Got it, thanks.

Are your speakers "active", crossover-wise? 

Duke

@audiokinesis wrote:

Sweet! 

I really like your radiation pattern choice, it’s wide enough and consistent enough to give very good coverage of the listening area, but not so wide that you have an excess of early-onset reflections. And you will have a good directivity match between your woofers and the horn in the crossover region. Those big woofers in their bigfaced box will have useful pattern control probably down to the Schroeder frequency just from their sheer size.

Initially I bought the EV system with the smaller HP940 horn (it uses the same compression driver, the DH1A, regardless), and though I was offered the larger HP9040 horn at the time I feared it would "overload" the listening space and thus opted for the smaller horn. It became clear however that the HP940 horn didn’t load the very powerful DH1A driver (2.1 Tesla flux density) properly, sounding too hot in the upper presence region, whereas the HP9040 horn made for a much smoother, more fully fleshed out, relaxed and physical presentation.

It goes to show that while the larger horns might at first be deemed crazy, too up front sounding, overloading the room etc. (a classic case of "listening" with the eyes or one’s preconceptions), they’re typically the more well-balanced, relaxed and smoother sounding variants over the smaller ones. A bit like large panel speakers on steroids.

It was apparent as well that the larger horn made for a more coherent and of-a-piece sound, which confirms a good directivity match at the crossover to the woofers, as you point to. Indeed size matters here :)

For instance, most manufacturers make cone-n-dome speakers that have wide radiation patterns, so they are already getting a lot of reflection energy out into the room. They may not see much potential benefit to adding more reflection energy even if it is corrective.  (The idea that rear-firing energy CAN be corrective may not be very widespread.)

I’m wondering if the typical "cone-n-dome" speakers are so popular with the manufacturers from the perspective of a deliberate design choice or more a reflection of a smaller size factor being a better business proposition? Perhaps I’m assuming more than wondering here, but I’m guessing power response and reflection field isn’t at the top of their agenda, or at least something they’re less likely to be successful with the typical designs choices of front firing only, direct radiating smaller woofers/mids and dome tweeters. 

Then there’s the question of cost/benefit: Even if there is some benefit from adding rear-firing drivers, could the same amount of money spent elsewhere make a bigger improvement?  And, perhaps the most practical question:  Will having rear-firing drivers make a worthwhile difference in sales?

I’m not sure most manufacturers even considers the alternative of adding rear firing drivers (with the exception of tweeters, perhaps, which seems to be about something else than getting the reflection field right per se).

Are your speakers "active", crossover-wise? 

Yes they are, with the exception of the JBL Alnico tweeters that use a simple 1st order high-pass filter with Mundorf Silver/oil caps coming in at ~11.5kHz. The Alnico’s are 105dB sensitive, but match wonderfully with the ~110dB large EV horn with a rather natural sloping here - indeed a lucky match. I don’t like the top end to come across as "lively" but rather prefers it blends in smoothly as a natural extension of the mids. I’m more of a "based around the midrange"-sound kinda guy, even though I’m not british :)

Obviously in my case the configuration is outboard actively with a Xilica DSP and the great british MC² Audio amps (S800’s for the EV’s, vertically bi-amped, and a T3500 for the subs). 

What are your thoughts on active configuration, and is it something you have considered for your own designs?

I think that for the best, "pinpoint" imaging you want a narrow dispersion and little to no reflection to "smear" the locations of the various "instruments" in the room.

This would provide the best "They are here" experience.

Experimentation with the Heil AMT shows the best imaging is attained when the rear signal is blocked off.

@toddalin - My experience is that speakers like you've described give you amazing imagine in exactly one place in the room.  

But yes, it's really pinpoint. 

 Narrow but controlled perhaps in a very lively room otherwise controlled wider dispersion so high frequencies disperse similar to lower is generally considered best. Ohm Walsh do it uniquely well resulting in the natural balanced  sound they are known for.  KEF Blade 2  meta seem to be the reference standard these days in regards to most measured performance including dispersion making them suitable for optimal performance in most any room but for a price. 

There is no perfect answer, but I think it's clear that the end goal is best served by the right room with the right speaker.  

I will say that I often dream of building my own line sources, as they have to major advantages:  Controlled vertical dispersion and you can basically sit or stand anywhere and here the same thing.