What is the difference between absolute vs relative band powers


What would be more useful in a given context or scenario?

I know the [B]absolute [/B]band powers are just the sum of the powers for the frequencies in a particular band of the spectrum. The [B]relative [/B]band powers are the sum of the powers in a certain frequency band divided by the sum of the powers in all the other frequency bands. So the relative band powers represent how much of the signal (i.e. a percentage) is made up of oscillations in a particular band.

I still don’t know what that means though, they both look the same and are moving the same.


The absolute band powers are useful mostly for people building their own algorithms/metrics who want to combine/analyze band power information in some way that doesn’t involve relative powers. The relative band powers are useful because, as you say, they express a percentage. So if you want to know how much of the overall signal is accounted for by a given band, you can do that.

The two measures will be correlated of course, but they won’t vary in exactly the same way. You can check that by graphing them at the same time. I’ve attached a screenshot of MuseLab graphing alpha_absolute and alpha_relative for the left ear electrode. Not how there are some similarities, but they’re quite different.


Hi Tom, that still doesn’t make much sense to me though… the language is too technical?

What is an algorithm metric?

I just want to know which one to choose for viewing a standard rising or falling of any given band. If gamma goes up then down, i want to see it go up and down just like it did in my brain. Should I click absolute or should I click relative?


Ah, fair enough! :slight_smile: I’d recommend trying the relative band powers, then.


Hi, could you send me the screenshot, please? I have looked up for it and haven’t found it.