For a bookshelf speaker, the rear wave corresponds to the circumferential resonance that sets the baffle step corner frequency. I don't know of a Q that needs modified at 500-1KHz, except that of the circumferential resonance which is good to reduce and will be worse for larger baffles. Are you talking about room modes? I can see that being useful for the non-reverberant field but I would not want a lot of midband rear radiation to emphasize the reverberation of the room.
The first speaker I built was a big 3x6' open baffle, too large to use now and the ragged treble of the fullrangers was grating in the end. It was capable of some neat tricks, but to be comfortable to listen to I need smoother response. As it was basically a wide plank on it's side leaned up against a wall, I don't think there was much going on with the rear radiation which I actively tried to reduce. I heard open baffles at one hi-fi show, but to me the rear radiation just emphasized the reverberation of the room which was very hotel-like and fought with the reverberations of the music.
IIRC there is a PDF out there that measures room reverb time and response with and without open baffle to test this, but I can't find it.
The first speaker I built was a big 3x6' open baffle, too large to use now and the ragged treble of the fullrangers was grating in the end. It was capable of some neat tricks, but to be comfortable to listen to I need smoother response. As it was basically a wide plank on it's side leaned up against a wall, I don't think there was much going on with the rear radiation which I actively tried to reduce. I heard open baffles at one hi-fi show, but to me the rear radiation just emphasized the reverberation of the room which was very hotel-like and fought with the reverberations of the music.
IIRC there is a PDF out there that measures room reverb time and response with and without open baffle to test this, but I can't find it.
I was talking about the rear of the cone which goes inside the box. I was meaning that not all dipole diffraction should be eliminated.
Directivity and critical damping to mimic the performance of a ~reactance annulled horn same as the pioneers. 😉
Guys, this is entirely too cryptic for me. I know I talk a lot, but I am trying to work this out.
In an IB speaker, when the wavefront falls off the baffle, it generates a negative pressure wave that radiates backwards. There is then a wandering null region between the positive and negative wavefronts. Is this what you mean by dipole diffraction?
I know that you can aim the nulls toward reflections if you're lucky. Excessive diffraction helps corral sound to the front of the speaker. However diffraction occurs on the surface of the baffle as well as on sharp features and a negative pressure wave is still sound energy directed backwards and will still be heard if it becomes part of the reflected or reverberant field.
In an IB speaker, when the wavefront falls off the baffle, it generates a negative pressure wave that radiates backwards. There is then a wandering null region between the positive and negative wavefronts. Is this what you mean by dipole diffraction?
I know that you can aim the nulls toward reflections if you're lucky. Excessive diffraction helps corral sound to the front of the speaker. However diffraction occurs on the surface of the baffle as well as on sharp features and a negative pressure wave is still sound energy directed backwards and will still be heard if it becomes part of the reflected or reverberant field.
If these are not required then why wouldn't you build an omnidirectional speaker instead of a dipole? No horizontally contributed diffraction and better envelopment/diversity.
I see your point. But where does that put constant directivity? What if we build a constant directivity speaker without any diffraction hot spots like baffle edges, and that seems to sound very good to most people without having the "null beams"?
Omnidirectional and ideal dipole are both constant directivity designs only they are intended to engage the room. Variety becomes one of the factors. Maybe it's like when you choose an amp and you look at the distortion spectrum, you don't want missing parts, just smoothly falling. Having nothing and then edges coming in can be distinct.
In any case, all types have different diffraction concerns during design.
In any case, all types have different diffraction concerns during design.
Last edited:
I've already got enough to worry about with my B652-Air when the driver FR keeps changing shape. I need a JBL M2 that I can build out of garbage!
Nulls to certain direction (interference) happen only on certain frequency, or small bandwidth, making reflection sound different than direct sound! Constant directivity is the logical way to go as you suspect, to make (early) reflections sound similar to direct sound. If there are nulls off-axis, avoiding pointing them towards first reflections would make sense.
Ah yes, thanks pointing it out 🙂 Was thinking edge diffraction. Boxed speakers could also be crossed over like Linkwitz does, below the diffraction, to get similar sound to all directions, omni. Vertical there would be interference still as long as multiple non coincident drivers are used, but that is out of scope.