Hey guys...we need a little rallying here... - Page 216

You're exactly right to notice that the sound from a 90° horn/waveguide is only 6dB down at 45°, not 100dB. Very good point, one that is missed by a lot of DIY newbies on the constant-directivity/waveguide scene. Very astute observation on your part.

That's why I think it's important that the waveguide chosen have pattern no wider than 90°, possibly even a smidge less. Definitely better to be in the 80°-90° range than the 90°-100° range.

But then again, that doesn't even matter much because we're only talking about a decibel difference at the "edge" of the horizontal coverage pattern. Still, we definitely don't want to be too wide.

We had this discussion earlier in the H290C waveguide thread.

One thing that should help ease your mind though - No matter whether your horizontal beamwidth is 80° or 90° or even 100° - the truth is the frequencies produced by the tweeter are high enough that absorbent room treatments are extremely effective at absorbing reflections, especially at the "grazing" angle they would be coming from in a constant directivity cornerhorn. Just put wedges along the walls adjacent to the speakers. At this frequency range, even drapes will do the trick.

This applies not only to constant directivity cornerhorns, but also to DI-matched two-way speakers placed a few feet from the walls. They are still close enough that reflections from the energy radiated outside the horizontal beamwidth are fairly significant.

Where you really have the most to worry about is in the midrange band, especially the lower midrange. As frequency drops, absorbent room treatments are less able to attenuate reflections. Still, the corners can be used to your advantage if you design the speaker to take advantage of them.

• Speaker placement and wavefront launch

There's a lot more information on this and related subjects in the Pi Speakers FAQ. Most of the stuff in there relates to SEOS builds, as they use a very similar approach.

• Pi Speakers FAQ

Treatments like that are fine for HF absorption but ineffective at MF where you really need it. So absorbent material on the wall behind the speakers will help, but don't stop there. Consider taking these additional steps:

If you're running constant directivity cornerhorns, use a midhorn large enough to control the pattern at the upper end of its passband and close enough to the walls (within 1/4λ) for the walls to act as mouth extensions at the lower end of its passband.

If you're using DI-matched two-way speakers for mains, include helper woofers or flanking subs to mitigate midrange anomalies from reflections off the nearest boundaries.

• Helper Woofer Location

Panel absorbers are most effective at low frequencies and very low midrange frequencies. They damp using sympathetic vibration of a panel and are tuned by panel size and compliance. Directivity isn't really an issue at these lower frequencies where the panel absorber is designed to work.

Fibrous materials like fiberglass, foam or cloth are used to absorb higher frequencies. They damp using sympathetic vibration of the fibers. The more fibrous material the sound goes through, the more attenuation there is. So the thickness of the damping material and the direction of the source (and the angle of incidence/reflection) are relevant for this kind of treatment because they have an effect on how much fibrous material the sound travels through.

So if we're talking about treatments to deal with early reflections from the tweeter, we're talking about fibrous materials. Things like the traditional foam wedges or cloth drapes. Those are sort of easy treatments, and will easily absorb early reflections from the tweeter.

If we're talking about reflections from the midwoofer, things are much harder to deal with. You either need panel absorbers or flanking subs. You really can't get a fibrous absorber thick enough to work at those frequencies.

But it's even better if you can run constant directivity cornerhorns, because they don't suffer reflections from the midwoofer. It is tucked back into the corner, well within 1/4λ from the nearest boundaries. That's why I only talked about fibrous treatments in the earlier post you were referring to about cornerhorns.

Of course, all speakers used indoors will benefit from distributed multisubs. This is a separate issue, mitigating room modes. But self-interference notches from nearest boundaries are a similar problem, one that can be dealt with using similar approaches.

The response fluxuations caused by room modes and self-interference both can be mitigated with dense interference from multiple sound sources blended together. That's what multisubs and flanking subs do. Cornerhorns just have a unique distinction of not creating self-interference notches in the lower midrange, so there's one less set of anomalies to mitigate.