Well this would be great if SM was actually contributing something. All I'm getting from this conversation is "I wouldn't do it that way.." Anyone can do that. I can do that and so can you.
Being that SM has brought nothing to the party in terms of commercial product, I have a tough time taking any of this seriously. If what he's talking about is so amazing, I find it tough to believe that no one is interested in it commercially. Personally (and I know this is harsh) I'm really tired of armchair engineers being dismissive of what's available.
A month to optimize a system for one track, really?
I call shenanigans.
"Being that SM has brought nothing to the party in terms of commercial product"
That's right, I have nothing to sell to you. Got any qualified potential investors? The last one I contacted was a large well known international company that has the resources to turn this into a product if that is possible. And it's a big if. I wrote their CEO and got a reply from their lawyer who told me they weren't interested because the patent had expired. The previous time I contacted them I got no reply at all. Same with all the others. This happend after I aborted preparation of 7 related patents having seen how it was all Apple Computer could do to defend itself from infringement by Samsung. The patent process IMO no longer can protect small inventors from large corporate predators. It boils down to a battle of money. OTOH, there is a smug satisfaction knowing I have something that may be one of a kind.
However, I have brought you something else, an idea that is not merely a new variant on an old idea but an entirely novel strategy. Every sound system you have probably ever heard or probably know about from the earliest wax cylinders to the newest high end digital whiz-bang relies on the same strategy, namely capture, store, retrieve, reproduce. This idea uses a different strategy; capture, store, retrieve, reconstruct. Why? Because the reverberant sound field scientists who've measured it tell us is the overwhelming preponderance of what we hear yet little of it gets on to a recording, what there is of it can't be entirely segregated from the first arriving sounds, and it is not in a usable form. The closest exception, binaural recording fails because it does not capture the vector nature of the reverberant field. This is critical to the model. How is it done? By reconstructing the mathematical relationship between the sound fields arriving at the microphones and the expected field arriving at the listener. This is what the mathematical model tells us and that is the real crux of the invention. The relationship is recontructed using an analog computer or its real time digital equivalent. The current prototype is a hybrid. The model is presented in the patent in graphical form. It shows relative to the first arriving sound that each arriving reflection has a relative directional angle of arrival, time of arrival, intensity, and relative spectral change.
In a laboratory using an entirely different and far more complex embodiment of the concept, in principle any sound field could be reproduced to any degree of accuracy desired. But from commercial recordings that contain their own reflections, in real rooms that impose their own acoustics, and within constraints of practicality, complexity, and cost, a specific seat in a specific place cannot be reliably duplicated. In fact given my initial expectations, what surprised me about the first prototye wasn't that it worked as well as it did but that it worked at all.
Anyway, I've given you something new to think about. I'd add this. The model strongly suggests that this approach is the only one that has any chance of working. If for no other reason, nobody has figured out a way to record the reverberant field as an array of incident vectors without recording the first arriving sound at the same time or in a way that it can later be separated.
