bb.linuxsampler.org

Well, omni and cardioid electret elements are pretty cheap, and can be turned into very good microphones. In fact, I have a bunch of cardioid elements to turn into mics sitting right here. Combining these small microphones with "envelopes" of rockwool, following the "portable vocal booth" principle except leaving off the bottom so that the floor reflection can be captured, would likely be worth experimenting with, to see if it can work.

And that's about all I can say on this subject. Ambisonics is not something I've studied at all.

Well, the theory behind it (and stuff like gradients, guess it's from vector calculus) is rather new to me. So far the above have been sort of cargo cult calculus.

BTW, successfully approximating free-field conditions in the higher frequencies shouldn't be that hard with some rockwool. What's needed is set by the type of instrument to be samples.

I didn't read this anywhere, it just dawned on me and I started to question why nobody had spilled the beans for me during my research:

Q: Why do anechoic chambers look like they do?

A: It's all topology, baby! By using wedge-shaped treatments the total surface area of the absorption material is maximized. But if the whole wedge has the same material density while the thickness varies due to its shape the wedge (and the chambers total surface area) will not have an uniform absorption.