2x15" front loaded horn F2B

Well, the theory is sound. It hasn't been experimentally verified much in low-frequency horns for PA applications partly because research is expensive, time-consuming, and you need mics and interfaces capable of measuring >165 dB linear average without distortion.

They do exist but are not cheap.

Here's the 2019 paper on this topic that I was referring to:
https://pub.dega-akustik.de/ICA2019/data/articles/000129.pdf



HF horns are another matter. The same principles should apply.

Here's Klippel with a model and a method to cancel out nonlinear throat distortions with processing, and a model by Holland for HF horns. Both from 1996:
https://www.aes.org/e-lib/browse.cfm?elib=7878
https://www.aes.org/e-lib/browse.cfm?elib=7891
Found those when I was looking at DIY 'Guinness focusing' aka Fulcrum's tQ.

Most of this work is in the field of vocal range horns. They're just used much more often, and we are more sensitive to distortion in those ranges. Voishvillo has a good one here:
https://www.aes.org/e-lib/online/browse.cfm?elib=11243

And here's one that's a bit more heavy on the theory:
https://asa.scitation.org/doi/abs/10.1121/1.1362688?casa_token=KZpJYrlKKr0AAAAA:glJBpmp-iaouCiDLWBbQR5li5DnRFb6e2IuZtLpB3NY2kX6x_kpIdvoDrySEPLBAHIc7XJPy3k7I

These all rely on one parameter models though, which don't predict the full modal response of the horn. Kolbrek published a Matlab toolbox for a semi-analytical modal propagation method s while back, but the more advanced version lives in his HornCAD software which is unreleased.

If you want to do it 'properly' like the big boys, then you just need a monster PC or server farm, a license for COMSOL with the appropriate add on models, and lots of free time and experience:
https://www.comsol.com/blogs/simulating-nonlinear-sound-propagation-in-an-acoustic-horn/

Edited by toastyghost - 23 October 2021 at 2:29pm

It’s a cut-down version of a presentation, so I think some things are missing or glossed over. I have questions about their calculation of the throat adapter’s contribution since the simple equation used isn’t accounting for the actual SPL offset.

The concept is justified. A high compression ratio is fine, if the rear chamber is suitably large. The modern drivers available play well to the parameters which horns are most sensitive to - BL, Mmd, etc.

I’d like to see a comparison of the relative distortion levels at the same throat SPL, for example. That would make answering your questions much easier.

There is one level matched in both setups; 109.8 and 109.7 dB at -1 dBFS and -3 dBFS input respectively. But it looks like they don't use those two different drive levels to determine the contribution of the throat adapter alone    

An increased level should increase the distortion anyway, as can be seen in the tables. It isn't fully intuitive, since the effect is non-linear.

You can kind of do it yourself though. Looking at the fifth-order harmonic, it goes up by 13.3 dB with the throat adapter for that 109.7 dB level. There's not enough other info to make predictions as to *why* though. No diagram or photos of how they inserted mics at the throat while changing the throat adapter and rear chamber for example. Is the horn a straight one, or folded?

I've looked but can't find the presentation recording, and they've not published follow-up papers yet.

There’s also the issue of the anechoic chamber used. There’s no documentation of the cutoff frequency, so an increased level might be increasing the reflection contribution - which would show up as distortion.

If we had more details about the construction of the horn and driver used, it could be recreated in a coupled non-linear model within COMSOL, including the effects of/on driver suspension and the like.

Again this is something I wanted to do myself, but I've spent the summer looking into other things first. Which unfortunately became a very big rabbit hole of their own

Sonic the hedge wrote: Thanks - some good resources! From what you and Snowflake have said, it does seem possible, that a basic principal behind F2B is to deliberately induce distortion in the throat, which is then acoustically filtered, by the horn, to produce a full and pleasing tone. It is a warm sounding cabinet, and I guess that well-tuned acoustic filtering just sounds natural, because it is. Interesting too, that there seems to be a bit of a gap in the underlying theory here - it does make you wonder, if MA have some secret unpublished formulae, or if 215/F2B was developed and refined by experimentation.

madboffin wrote: I can assure you that it was developed and refined by experimentation. It followed on from the bass horns used in the 215Mk3 and RS1200 cabinets. There was also the ergonomic and economic imperative of making it fit in a standard semi-trailer, which dictated one of the mouth dimensions (570mm or 22-1/2" to pack four across). You need to understand that when we were developing this in the late 1980's there were no computer simulations and the main measuring instrumentation was B&K 2010's and chart recorders. We had just acquired a TEF-10 (a piece of transformatively innovative technology, look it up) and were beginning to learn how to use it, but it played almost no part in the F2B's development.

Don’t think so, I see the links when I’m logged out (which is often because the new login system for the AES hates me).

But the really good one is also available here:
https://www.prosoundtraining.com/2021/02/04/heyser-videos/