I'm trying to wrap my head around how does "beamforming" work when there's no physical "space" between receivers.
My background knowledge:
I've always thought about beamforming as a sort of spacial FIR filter. We have a number of antennas (or in this case microphones) that are at some distance one from another. Due to that distance, the signal will arrive with different phase at each of the receivers. Afterwards, we multiply the received signals with weighting factors that basically decide in which direction we'll look at.
What I did:
I have a laptop with an integrated stereo microphone and stereo microphone input. Driver for the soundcard, among other options, gives option to enable beamforming.
I connected a LW/MW/SW/FM radio with stereo signal output to the microphone input of the computer. Signal provided by the radio is a mono signal from an AM station duplicated to two channels for stereo output on the radio. In time-domain, signals on both channels are completely same and in phase.
I was quite surprised that after enabling beamforming under soundcard options, I got a significant noise reduction.
My thoughts about this:
My understanding so far was that since I have two channels that are 100% same, that should be equivalent to having two microphones in same physical location. Because the distance is zero, beamforming should not be possible. From what I could dig up, the laptop itself uses this product as the audio DSP. There's not much information available on the manufacturer's web site and the whitepaper they show doesn't tell much above the usual beamforming stuff with space between microphones.
So my question is what could be happening here? How do I understand this application of beamforming where both receive channels are getting the same signal and yet the beamforming does work.
Update So commenters think it's not beamforming at work. Could it be that they are using noise-shaping and then low-pass filtering as well as as step in-between?