A key thing about sound waves is how they combine. If you take two identical sounds and play them together, the peaks of the waves align and amplify to make a very loud sound. This is called constructive interference.
But you can also do a sort of reverse. If you have a wave with peaks and troughs at certain points, you can invert the wave by taking every peak into a trough and every trough into a peak. If you play these two sounds together, the peaks of the original sound line up with the troughs of the inverted wave. These cancel out giving you destructive interference.
The neat thing is that you can take a combined signal of multiple sound waves, and further combine it with just the inverted wave of the original sound coming from the microphone and that will perfectly cancel out the microhppne sound without cancelling any of the normal audio you do want to keep.
This is also how noise cancelling headphones work. They sample the noise coming from outside the headset, invert it and add it to the headset sound. This played sound combines with the outside sounds and causes the same destructive interference, cancelling out the outside sounds so you then only hear the sound from the headset. It’s not an *exact* match to what waves actually make it through into your ear so it doesn’t completely cancel all background sound, but it’s close enough to significantly dampen the external sounds. Combined with regular sound blocking and sound proofing you get a high quality noise reduction.
This is called acoustic echo cancellation (AEC). The basic idea is that your computer knows what audio is incoming and being output to your speakers, so it can use that as a reference in order to filter out the aforementioned audio from your microphone signal.
It’s not perfect. If the filter is too aggressive it will cut out intended speech, too lenient and it will let echos through. But on the whole it works pretty well.
We’d need more info in order to give a specific answer. For instance: What microphone or devices are you using? Are there (sound) speakers? What apps are involved?
Microphones themselves have a pickup pattern and can be made such that sounds emanating from other directions are picked up at a lower volume or not at all.
The audio can be processed in order to eliminate the call audio (or any audio output). This is since you can utilize destructive interference. This can be done for phones to avoid feedback when used in loudspeaker mode.
The audio itself may not be passed on as part of the input. Without external speakers, the sound output of a computer would not be sent as part of the input to the call software as this could cause a feedback loop. Involving speakers would motivate the prior 2 paragraphs.
Latest Answers