Direct or Reflect ... What's your flavor

I never look back. Someone might be gaining on me. But if you ever do, remember Meatloaf: speakers in the rear view mirror may appear closer than they are.

Erik, I have been using dipole line sources exclusively since 1979 including ESLs, ribbons and planar magnetics. I have come to the conclusion that ESLs are the best. My next and probably last set of speakers is going to be Sound Labs Ultimate 845s. I have absolutely no interest in dynamic speakers other than the Sonos units I use for background music and the old Mirage speakers I use in my workshop.
I find horn speakers intriguing but I have yet to hear a set that really wow me out. It is difficult if not impossible for a dynamic loudspeaker to match the accuracy of an ESL. But, the cheapest full range ESL is $14K and they are big things that will totally dominate the environment. It is a shame Acoustat went under as they produced ESLs at a price most of us could afford. Having said that the most expensive Sound Labs are about 50K. Given that there is no loudspeaker at any price that a good number of us would prefer makes them an extreme value. Why anyone would spend $250K on Wilson's or Magico's is totally beyond me. 
millercarbon, there is a great white about to bite you in the a--. 

Eric......If I may......Bose 901s are not ambient reflectors, as 89% of their output was purposely sent to the back wall. Then you have speakers designed by Stewart Hegeman, and all of the manufacturers who use his design ( s ), ( look them up ). Tweeters on the rear of a box, yes. And talk about ambient retrieval ? This can be done electronically, or, with other speakers placed about the room. It is my contention ( however, no arguments, please ), that recordings have ambience, and I want to hear all of it, before any additional reflections of the room. This is my preference, and feel, there is no right or wrong, as long as the listener is happy. Panels, omnis, box, horns, whatever, is no different, than ice cream flavors. Same with analog vs digital, tubes vs solid state. Most folks know what I like, but, the important thing, and this should be with every one, that when you find what you like, good for you, and no one can say different. Enjoy ! MrD.

@mrdecibel 

OK, maybe the 901s were not, but a lot of future Bose speakers were. :)

Eric, don’t you hate it when some smart-@$$ comes along and argues with the facts? Well, today it’s my turn to be that guy.

"NONE of these ambient reflectors are accurate. None of these latter elements can be called a way of credibly and accurately increasing the transmission of information from the recording to our ears."

This statement makes intuitive sense. How can adding something which wasn’t on the recording to begin with (the additional in-room reflections) possibly increase accuracy? I know what you’re thinking... "Duke, right about NOW would be a good time to stop typing"...

I’m not going to argue in favor of every single technique for deliberately adding reflections; rather, I’m going to argue that DONE RIGHT can be a worthwhile net improvement, especially if we are designing for ears rather than for test instruments.

Perhaps our first question ought to be, are reflections ever benefical in a home audio setting? Since anechoic listening is not a desirable goal, the answer must be "yes".

Next question: Are all reflections equally desirable? Apparently not, else acousticians would be out of work.

What, then, makes reflections desirable? This is a complicated subject so I’m just going to focus on one aspect for now: Reflections should sound like the direct sound. That is, they should be spectrally correct. The more they sound like the direct sound, the more beneficial they are likely to be. And the less they sound like the direct sound, the more detrimental they are likely to be. (Things like timing and level and diffusiveness and arrival angle and decay rates all also matter, just for the record.)

So if reflections are going to happen, and if they can be a good thing, does it not make sense to MAKE them as beneficial as we reasonably can?

Let’s look at how we might accomplish this. Suppose we start with a good wide-pattern conventional speaker. The speaker’s pattern averages perhaps 160 degrees wide over most of the spectrum, narrowing considerably in the top octave as the tweeter starts to beam. The spectral balance of the first-arrival sound will be the on-axis response (or close to it), while the spectral balance of the reflections will be dominated by the off-axis sound, and because of the tweeter’s beaming, it will be weak in the top octave or so.

Now suppose we were to take this speaker’s 160 degree pattern and chop it in two, with 80 degrees firing forward and 80 degrees firing backwards. Each half gets its own tweeter so the highs hold up better out to the sides of each 80 degree wedge. Now let’s mentally examine the spectral balance of our reverberant energy, and which includes a backwave identical to the frontwave, and we’ll see that there is much less spectral discrepancy between the first-arrival sound and the reverberant sound. The overall loudness of the revereberant sound relative to the direct sound has not changed, but its spectral balance has been significantly improved.

This is but one example of a way that intelligent manipulation of the reverberant field can be benefical. (Note that virtually everything a professional acoustician does falls into the category of "intelligent manipulation of the reverberant field".)

"They add... perhaps an illusionary venue."

In a home audio listening room there is, in effect, a competition between the acoustic signature on the recording (whether it be real or engineered or both), and the acoustic signature of the playback room. At best, in home audio we are presented with a poverty of recording venue cues, compared with actually being there. Still with a good recording and good system setup, we are often able to perceive a plausible illusion of the recording venue. The trick is to effectively present the cues on the recording while simultaneously minimizing the inherent "small room cues" which our playback room super-imposes atop the recording. This is a complicated topic which I will go into more detail about in another poster’s thread, entitled "The "They are here" vs "You are there" sound topic".

But imo this is another instance where intelligent manipulation of the reverberant field can pay significant dividends. By way of a quick anecdotal example, consider the experience of countless Maggie and Martin Logan and Acoustat and SoundLab and so forth owners with pulling their speakers out into the room: With sufficient distance from the back wall (i.e. sufficient time delay on the backwave energy), they hear significantly more variation in acoustic signature from one recording to the next.

So in my opinion, "done right" (which includes appropriate set-up), a good polydirectional speaker gets some things more correct than a conventional speaker does, particularly from a perceptual standpoint. 

Duke
speaker manufacturer