narrow and wide baffles and imaging

When people start talking imaging I sometimes wonder what they're talking about. I know some people are just talking about some sort of left-to-right point specificity. If that's all you're getting, I don't consider that very good imaging. I expect a field of sound that exceeds the spacing of the speakers. It should convey at least 50 or 60 feet of convincing depth of field to have any hope of being "good", and likewise reach out towards me. A single guitar amp sitting on the floor should sound low in the field and overhead cymbals should sound way up there with clear specificity of location. All that jazz takes the right speakers in the right room. If those attributes are high priorities for you, you pretty much need narrow speakers and space to place them for proper boundary interaction. The physics of wide baffles and sharp corners can and have been modeled and the results do coorelate to poor imaging. Sound waves ride along the surface until they reach a corner and then project from it like a second source of sound with the corresponding delay and distortion. Siegfried Linkwitz did a great lecture on imaging and baffle interactions at BA that is full of great information that strongly supports small drivers and the minimal/no baffle approach to speakers for best imaging. That's not too far off from what Wilson tries to do with their narrow cabs covered with heavy felt and small drivers crossed over at the lowest possible frequencies. It seems to work. 

Looking down this thread I kind of wonder if everybody is defining "imaging" the same way. 

I'm definitely calling that a baffle too. Without a baffle dipole speakers suffer from all kinds of destructive interference as the wave fronts wrap around the drivers. 

Typically when someone describes a baffle the presumption is that it's a flat slab. Nobody presumes that they're wildly sculpted things like the IRS V or the LS50. The problem with giant flat baffles is that the do have an edge somewhere, and that edge is going to do bad things. Curving the baffle is an interesting and effective way of mitigating that edge diffraction effect. Those who really think giant baffles with big, sharp corners image well make me curious as to what they're hearing. The behavior of waves rolling off of a surface isn't a matter of speculation or opinion. It's something you can measure, and it has been measured, and the measurements make it completely clear that large baffles that terminate in anything resembling a corner diffract waves just like independent sources of sound. So while a huge baffle may solve some issues with making the drivers behave well in some aspect, it's pointless because you're creating another very audible time delayed artifact pouring off the corners of the box. No matter how you cut the cake, big baffles aren't good if they're anything resembling flat with corners. 

Kalali,

I don't know what kind of drivers you listen to, but you'd need a 7" tweeter if you don't want it's passband effected by the baffle. The mid-range would have to absolutely massive as well.