2-way speaker narrow midrange dispersion - does it matter?

[arjscott]

I'd like to build some speakers with a large "sweet area" that would allow 2 or 3 people to have the same listening experience when seated between the speakers.

I suppose the key to doing this with conventional front-firing box speakers is using drivers with uniformly wide dispersion.

Other approaches such as omnidirectional or dipoles would achieve my goal, but 1) they are probably a lot more difficult to build and 2) the imaging is probably not as precise compared to front firing speakers.

A typical 2-way speaker might have a 6" woofer with a tweeter, crossed over at 2500 Hz. I'm not exactly sure how different woofer cone sizes affect dispersion, but I've read that as the wavelength becomes smaller than the cone surface, the dispersion narrows and "beaming" occurs.

Would the upper portion of the woofer's frequency range have significantly reduced dispersion? If so, I guess this could be heard by anyone sitting off-axis.

But, does this really matter?

With a speaker such as the Cryolite (7" Dayton RS woofer with BG Neo3 planar tweeter), would the tweeter's wide dispersion be enough for the speaker to sound pretty good for people sitting off axis?

Or, if reduced midrange dispersion really is a problem for 2-way speakers, should I look for a 3-way design? If this is the case, perhaps a 2 or 3" cone midrange is a good idea as people don't seem to like dome midrange drivers.

I haven't built any speakers before but will be as soon as I find answers to these questions. Perhaps I should stop worrying about it and start experimenting?

Apologies in advance if none of this makes sense or seems really silly.

[MBK]

The points you raise are valid enough, but there are many trade offs to consider, and not just engineering trade offs (money, chances of success for a 1st speaker etc).

Fundamentally, the easiest way to get a better sweet spot is to sit farther away from the speakers ;-) .

The problem with stereo reproduction is this: the best imaging is obtained when sitting precisely in the equilateral triangle between the speakers. As long as the sides of this triangle are just 3, 4 meters (9-12 feet) long, the width of an average couch will just be a lot more than your sweet spot. You will notice an imbalance of sound, regardless of the dispersion you touch upon, simply because one speaker will be closer and therefore louder, if you sit off-center.

If you have speakers spaced say 3 m+, or 10 ft apart, and sit at 5, 6 m or 15-20 ft away, then you lessen this problem, at the price of much more room interaction (the sound you actually hear has then been reflected by the room at least once), As a result things will sound more like "they are here" and that's great for music, but imaging and speech intelligibility (movies!) will suffer.

All in all if you take it to the extreme, achieving a couch sized sweet spot is an almost impossible task in a normal size living room if you ask me, but this has more to do with geometry and pathlength of sound radiation (time of flight and reflection problems), than with driver radiation characteristics.

[Jim Griffin]

Near field line arrays (line arrays that operate in the near field vs. the conventional speaker far field operation) achieve a very large sweet spot. I call it a sweet area. What happens is tha the NFLA contours the sound in the vertical plane which enhances the horizontal dispersion. Details in my white paper.

JimNear Field Line Array White Paper

[MBK]

Interesting paper and intriguing to try this kind of solution. I suppose a bit tough for the original poster as a first time project though...

[arjscott]

Thank you both for your responses. It's now clear that building the kind of speaker I desire requires a lot more than wide dispersion drivers.

A near field line array seems like a great solution. Although as MBK says it would be a tough project, I might just try it!

It would be great to listen to one of these. If anyone out there lives in New Zealand and has a line array they're willing to let me hear, please let me know.

[MBK]

Actually the Ueber-thing would be a dipole-nearfield line array then (I am a dipole person ). Speaking of which, by the man who published sort of a bible on dipoles on his website, Siegfried Linkwitz, here is another discussion of pros and cons of line arrays: Line array discussion

Best of luck with building one!

[el`Ol]

Here are the responses of a good 13cm woofer. Fs is not too low (38 Hz), but sensitivity of 90dB is good for this species.

http://www.expolinear.de/website/gra...h_130_k_gk.jpg