Where signal meets metal
An electrode is the doorway between your brain and the recording equipment. Think of it like a doorbell button: if your finger barely touches it, the signal is faint and flickery; press firmly and it rings clearly. The quality of that contact is measured by impedance — roughly, how much the connection resists the tiny electrical signal trying to pass through. Low impedance means a clean, open doorway; high impedance means a clogged one, where noise can sneak in and the real signal struggles to get out.
On the scalp there are two common styles. Wet electrodes use a dab of conductive gel that bridges the gap between metal and skin, soaking past dead skin cells and oils to lower impedance — great signal, but messy and slow to set up. Dry electrodes skip the gel and press metal pins or pads straight onto the scalp; they are fast and comfortable for everyday use, but the contact is harder to keep clean and steady, so they often start out noisier.
Voltage is always relative
Here is a fact that surprises many people: you can never measure a voltage on its own. Voltage is always a difference between two points, like measuring someone's height — "tall" only means something compared to a floor. So every brain signal you record is really the difference between an electrode over the brain and a second reference electrode somewhere else, often on the earlobe, mastoid bone behind the ear, or another scalp site.
There is also a separate ground electrode, which gives the amplifier a stable baseline to compare everything against — think of it as the system's "sea level." Because every channel is measured against the same reference, where you put that reference quietly shapes what you see: a reference sitting near active brain tissue will subtract away some of the very signal you wanted. That is why labs think carefully about referencing, and sometimes mathematically re-reference the data afterward to a common average.
Montages and the 10–20 system
If two labs both put an electrode "on the front of the head," they might mean slightly different spots — and then their results can't be compared. To fix this, scalp EEG uses the 10–20 system, a standard map that places electrodes at fixed fractions (10% and 20%) of the distance between bony landmarks on your skull. Each spot gets a short name: letters for the brain region (F for frontal, C for central, P for parietal, O for occipital) and numbers or letters for the side (odd on the left, even on the right, z down the middle).
A montage is simply your chosen recipe of which sites to compare against which. Using the same reference for every channel is a *referential* montage; subtracting each electrode from its neighbor down a chain is a *bipolar* montage, which sharpens local differences. Same electrodes, same brain — but the montage decides how the activity is displayed, and so it changes what jumps out at you.
Implanted grids and arrays
Once you go inside the skull, the skull and scalp no longer blur and weaken the signal, so everything gets sharper. Resting flat on the brain's surface, electrocorticography (ECoG) grids are thin sheets of electrodes that pick up activity from the cortex right beneath them. Pushing deeper, penetrating microelectrode arrays are tiny beds of needle-like tips that nestle among neurons and can hear individual cells fire. The trade-off is the usual one: more detail and proximity, but more invasiveness.
No matter the electrode, the raw signal arriving in the wire is astonishingly small — often just millionths of a volt. So the very first thing it meets is an amplifier, an electronic circuit whose whole job is to boost that whisper into a voltage big enough to digitize without drowning it in new noise. Electrode, reference, and amplifier together hand a clean stream of numbers to the computer. In the next guide, we follow that stream into how the signal is filtered, sampled, and made ready for decoding.