How are we able to “isolate” sound and conversation?

157 views

[ad_1]

How are we able to focus on a particular person’s words and almost “drown out” other people around us?

In: Biology
[ad_2]

One answer to this is biology: our brains learn to do this because it’s an important skill, and over the course of evolution, people with genes that made it easier for us to pick out a stick breaking in a jungle we’re more likely to survive.

The other answer would be more mathematical and about how we write software that does this. For example you can search for PCA or noise reduction.

The human brain is extremely good at seeking patterns. Language is a pattern, so we’re naturally good at distinguishing it from background noise. We’re so good at it, in fact, that we can sometimes here language where there isn’t any and our senses are fooled.

Also, when it comes to distinguishing between people, it’s important that we focus on that person. If two people speak at the same time and you’re not focusing, you can’t follow either story, but if you focus on one (doesn’t matter which one) you can completely drown out the other. This is because we tell our brain to focus on a specific pattern (one person’s voice), which is a lot easier to distinguish than two at the same time (less signal overlap).

To put it simply: Language is a pattern which we can detect and doing so with other people talking in the background is easy when we train ourselves to focus on just one of those patterns.

Hearing scientist here.

Evolutionarily speaking, humans are ridiculously good at working out where sounds come from. We’re not totally sure why were so good at it, but we’re very good at it.

This allows us to do something called spatial release from masking. Basically, if we can work out where two different sounds are coming from, we can choose which we want to focus on.

How we work out where sounds come from is something [that comes up fairly often](https://www.reddit.com/r/explainlikeimfive/comments/ldzj17/ELI5%3A_In_8D_audio%2C_how_is_the_audio_able_to_sound_like_it%27s_behind_or_in_front_of_you_even_though_there_are_only_left_or_right_outputs%3F/gm8vmzd/?utm_medium=android_app&utm_source=share&context=3). But if you have two microphones, for example, you can do some clever maths to add the two signals together and focus on different regions of space. This is basically what our ears do all the time.