Ideal dispersion pattern of HF horns

[Tarheel TJ]

     I would love to hear some experience from people who have used horns with different dispersion patterns in their system.  I have been considering some upgrades to the HF section of my speakers and one of the main differences between the two potential upgrade paths is vertical dispersion (compression drivers in a horn vs. AMT drivers with wings).  In what ways does this affect in-room performance?

    On once hand, the very limited vertical dispersion of a AMT drivers would seem to be a benefit as it would keep excess sound off the floor and ceiling.  On the other hand, it seems that it is good to match polar coverage of your HF and LF horns at the crossover point.  Wouldn't a horn with somewhat wider vertical dispersion than an AMT (like a K-402) better match polars with a horn-loaded bass bin (like a La Scala)?  Wouldn't that produce a more even power response?  Or is this only important in the horizontal?

  

[Chris A]

The vertical coverage is an issue, trust me, especially when it's too narrow.  Also, it's a problem when it covers way too much vertically at some breakpoint frequency (loss of coverage control when descending in frequency) --like Heritage midrange horns: the K-400, K-500, K-600, etc.

 

I've heard horns having the following coverage:

 

1) 90 h x 60 v, down to 100-200 Hz (essentially full range).  These are K-402s on top of KPT-KHJ-LF bass bins (present Jubilees)

2) 90 h x 60 v, down to ~600 Hz (K-510s), then Khorn bass bins (above the same coverage vs. frequency as the above Jub bass bins), and also Belle bass bins

3) 90 h x about 15 v, down to 700 Hz (AMT-1s) then Khorn bass bins

4) all kinds of horns and other direct radiating drivers that lose vertical directivity below about 2 kHz (such as the K-400 series "pattern flip" midrange horns)

5) Full range dipole radiators (Magnepan MG-IIIa)

6) etc.

 

The AMT-1s have the same horizontal coverage as the K-402s and K-510s: 90 degrees.  Single-high horizontal coverage on top of a Cornwall bass bin, crossed at 670 Hz, below:

 

 

The problem with the AMT-1s, like you say, is the limited vertical coverage, and stacking them together actually makes this worse, except for the fact that the height of the direct arrivals from them increases (just like a large dipole radiator--like the Magnepans).  So the issues with too narrow a vertical coverage is that, once you stand up, you don't hear the highs well if standing close (within 5-10 feet) of the driver.  And their coverage decreases with increasing frequency--just like a line array.  Is this an issue? It depends on your listening style. 

 

Adding three stacked AMT-1s on top of one another, and tri-amping the three pairs of AMT-1s to created a "shaded line array" (like the last configuration "f" below, delaying the middle and top AMT-1s to create a "J" shaped line array coverage) would solve a lot of these issues for home hi-fi service:

 

 

But note that his creates less than 90 degrees of consistent horizontal coverage, however--just like "lineless" bifocal lenses for glasses with the junk on each side of the on-axis imaging).

 

The only real solution for both horizontal and vertical coverage, taken together as the measure of merit, is a point source radiator behind a full-range horn like a K-402 or a full-range MEH--like the Danley-style Unity and Synergy horns.  I find that I move around a lot and really enjoy my three-across K-402s (crossing to Jub bass bins in the corners, and dual 15" woofers in the center--on-axis), so I have no real issues with polar performance anywhere in the listening area of the room (40 feet long, 15.5 feet wide, 9 feet high), covering everything except perhaps the first 4 feet at the front wall (15.5 wide)--where I don't usually go anyway.

 

Chris

 

[Tarheel TJ]

Interesting.  Thank you for your insight.  I was hoping you would reply to this.

 

So, assuming the listener is always in the sweet spot, is limited vertical dispersion a significant drawback?  Is vertical dispersion matching between the bass horn and HF horn worth pursuing?

 

I am debating going for TAD beryllium 2" drivers for the extended HF response.  However, it seems one could go for an AMT-based HF section that would have a similar clean, extended response for much less money.  The main drawback that I can figure is the limited vertical dispersion, as compared to a CD on a horn.  I'm trying to decide if this is splitting hairs, or is it a significant issue.

[Chris A]

2 hours ago, Tarheel TJ said:

So, assuming the listener is always in the sweet spot, is limited vertical dispersion a significant drawback?

I have a somewhat less common view of this. In general, if you find yourself always seated in the sweet spot (I'm not--by design, which is a topic for another thread), high frequencies really don't bounce off the walls very well above about 4-5 kHz, so it's the direct arrivals that are important. Between 200 and about 4000 Hz, having good vertical coverage is important for perception of spaciousness in-room--the "second room" reflections (the "first room" being the one where the recording took place) that can aid the original stereo recordings, but are generally not needed for multichannel playback since you've got at least 5 surround channels for feeling of spaciousness. 

 

2 hours ago, Tarheel TJ said:

Is vertical dispersion matching between the bass horn and HF horn worth pursuing?

I've found that while it may not be highly detectable during short-term listening trails, consistent polar coverage between the bass bin and HF horn will sound better to the trained ear over time--i.e., if you put a good system in the room that has consistent and wide-enough polar coverage horizontally and vertically, 200-->10,000 Hz, you'll detect it and will welcome the increased naturalness and fullness of sound. 

 

So yes, it is worth pursuing, but describing what it sounds like is much more difficult--in the remaining 5% of stuff that Roy and Paul Klipsch used to talk about that's important.

 

2 hours ago, Tarheel TJ said:

I am debating going for TAD beryllium 2" drivers for the extended HF response.  However, it seems one could go for an AMT-based HF section that would have a similar clean, extended response for much less money.  The main drawback that I can figure is the limited vertical dispersion, as compared to a CD on a horn.  I'm trying to decide if this is splitting hairs, or is it a significant issue.

Depending on how you approach an AMT-1 (multiple stacked or single high AMT-1, wings or no wings, shaded delays while tri-amped three-high, etc.), you may be able to get everything you want--and for significantly less money than beryllium diaphragm tweeters.  The tradeoff is a little higher complexity of amplifiers and DSP crossover/wires.

 

I'd also recommend looking at the performance of dual-diaphragm drivers--like the BMS 4592 or 4594.  And lately, there is a new single-diaphragm driver on the market: the Celestion Axi2050, which is about the price of a Radian 950BePB (beryllium diaphragm). All of these alternatives are probably more cost competitive than TADs--unless you can find the TADs used and in good condition (i.e., that haven't been abused in service--like commercial service, etc.).

 

Chris