When we speak of global feedback we’re referring to the practice of wrapping the output signal of a device back to its input for comparison and correction. Because the input “knows” what’s right, a simple comparison circuit between the two serves to rectify any differences.
Of course, nothing in engineering is a free lunch. You cannot simply add global feedback in the hopes of perfection in the same way you cannot expect a broken leg to perfectly mend as if nothing happened. Both are improved, neither are perfect.
Along the same vein, it’s been suggested that it might be possible to wrap global feedback around the system rather than just individual components within the system. So, imagine a scenario where the speaker output is measured in real time and fed back to the source for comparison and correction—a brute force approach to lowering errors, to be sure, but an interesting notion too.
From a technical standpoint, running such a long global loop is fraught with problems: timing, latency, phase differences, amplitude changes from the level control that cannot be known to the system are certainly starting hurdles. But while it’s an innovative idea that might have a chance at implementation in the lower frequency ranges, what it suggests to me is something more interesting—just how far off the eventual output signal is from its starting point.
As we build upon the chain each component adds just a little bit of its flavor until the end result is reasonably far from its beginning.
This isn’t necessarily a bad thing. Our systems are amazing at what they do.
I just find stuff like this ever-fascinating.