Asymmetric horns?

ideally you want to have a vertical array of boxes with the same drivers etc. but with increasing vertical dispersion and splay angle towards the bottom.
so lets say two 2x12" 2x8" 2x1,4" with 5° vertical, then one with 10°  and one with 20° below giving you 40° in total but with half of the total energy in the upper 10°. you will need a different waveformer for every module though. Look at Charlie Hughes waveshaping devices for inspiration.
Im working on exaclty this type of box in the moment, but it will need some time as i dont have the money and gigs for such a big system right now.

The acosutic lense of danley is doing the same thing, but "the other way around" out of one box (because one box solutions are their thing):
the throat area of every "section" of the horn is decreasing, and therefore the amount of energy that is "going into it" , but the vertical angle this "section" covers is constant.


Edited by corell - 04 October 2017 at 3:41am

I've not heard any of the mentioned Bose products. I'm probably not missing out there though. I already have some ideas of how to transform the output of 4 compression drivers into a long curved slot. It won't be easy to make, but who doesn't like a challenge eh?

Having done some more reading, and some modelling with MAPP, it looks like I'm best using a combination of the 2 methods I mentioned, amplitude taper and angular taper. If I divide the horn into 8 sections with a splay (horn vertical angle) of 1-8 degrees from top to bottom, in addition to 6dB level taper top to bottom, I should get the coverage I need.

The slightly confusing part is how to achieve the amplitude taper. The examples I looked at assumed 8 identical boxes, whereas my sections would have progressively greater vertical horn angle. This would suggest that the output from the horns would progressively reduce due to the larger radiating area. On the other hand the mouth dimensions will be too small to really control all but the highest frequencies in such a small angle. Can I therefore ignore this issue, and treat this as a normal line array?

Am I correct in thinking that if I had 2 horns driven by a single compression driver, each horn would output 3dB less than a single horn on an identical driver would? Or should that be 6dB? In other words do I need to use 10Log or 20Log when calculating what proportion of the drivers is fed into each horn to produce the amplitude taper?

This is getting complicated, I thought it would.


Edited by doober - 07 October 2017 at 11:02pm