Two mornings in a row I brewed the same coffee at the same dose, ground on the same setting, with water out of the same pitcher. The only thing I changed was what the water ran through — a paper cone on day one, the stainless mesh basket that came in the box with my dripper on day two. The paper cup was clear and a little lean. The metal cup was heavier, rounder, and left a fine silt in the last half inch. Two noticeably different coffees out of one bag of beans, and the variable was the thing I had always thought of as packaging.
What a filter lets through and what it holds back
Three things are trying to get out of the coffee bed and into your carafe. The dissolved flavor compounds — the acids, sugars, and aromatics that make coffee taste like coffee — are in solution, and every filter passes them freely. The other two are physical. There are the lipids, the oils that carry a lot of the aroma and most of the mouthfeel, and there are fines, the dust-sized particles that any grinder produces alongside the pieces you actually wanted. Filter material decides how much of those last two ends up in the cup, and that is the whole story in one sentence.
The National Coffee Association's own brewing guidance puts the hierarchy plainly. Its pour-over guide notes that paper filters absorb more of the oils in coffee, which contributes to a cleaner, lighter taste, while metal filters allow more oil and sediment through for a more intense taste and texture. Cloth, it says, lands in the middle — holding back more than metal does and less than paper. That ranking has matched everything I have poured at home.
Paper, and the rinse it asks for
Paper gives the cleanest, most transparent cup available in a kitchen. If you are drinking a light-roasted single origin that somebody roasted to show off a specific fruit note, paper is the filter that lets you find it. It catches nearly all the fines, so there is no grit and no lingering bitterness in the bottom of the mug, and it goes in the bin afterward with no cleaning at all.
Two habits make paper behave. The first is rinsing: set the filter in the brewer, pour hot water through it, and dump that water out before the coffee goes in. That washes out the faint papery flavor and, just as usefully, warms the cone and the vessel underneath so your brew water is not losing heat into cold ceramic. The second is watching your grind. Paper is the filter most easily blinded by fines, and a bed that stalls halfway through the drawdown is almost always telling you to go coarser rather than to switch brewers.
Bleached versus unbleached is the question I get asked most and it matters least. Brown paper is simply not bleached; it is not cleaner or more natural in any way that reaches the cup, and in my experience it carries a slightly stronger papery note that makes the rinse more important, not less. Shape and fit matter far more — a cone cut for one brewer sitting crooked in another will channel water down the side and under-extract the bed.
Metal baskets, body, and the silt at the bottom
A permanent metal filter is a sheet of etched or woven stainless with holes big enough to pass oil and small enough to hold back most of the grounds. Most of them. Fines slip through by design, which is why a metal-filtered cup has visible sediment and why the last mouthful is the one to leave behind.
What you get in exchange is body. The oils come through, the mouthfeel gets thicker and rounder, and a dark or chocolatey roast that tasted thin through paper suddenly has some weight to it. A French press is nothing but a metal filter with a long steep attached, which is exactly why press coffee tastes the way it does. Nothing gets thrown away either, which is a real argument if you brew every day.
The catch is maintenance, and it is not optional. Coffee oil builds up in the mesh, goes rancid, and starts flavoring everything you brew through it. Hot water and a rinse do not cut oil. A metal basket wants actual dish soap and a soft brush regularly, and a soak in a proper coffee-equipment cleaner every month or so if you use it daily. A neglected metal filter is the single most common reason someone decides they do not like metal filters.
Cloth, and the reason it stays a minority taste
Cotton flannel is the oldest of the three and, brewed well, the most flattering. It holds back the fines the way paper does while letting more of the oils through, so you get clarity and a silky texture at the same time — no grit, no thinness. There is a reason the Japanese nel drip tradition has stayed with cloth for a century.
Then there is the upkeep, which is why most people try it for a month and go back to paper. A cloth filter must never be allowed to dry out with coffee oil in it, because dried oil turns rancid and is not coming back out. The standard routine is to rinse it thoroughly after every brew, keep it submerged in clean water in a sealed container in the fridge between uses, change that water regularly, and boil it periodically. It is a piece of damp, food-soiled fabric that you are storing on purpose, and it needs to be treated like one. When it starts smelling of anything other than coffee, it is finished.
Your grind has to move when your filter does
This is the step people skip, and skipping it is why filter comparisons so often end in a shrug. Each material wants a different particle size. Metal passes fines, so grinding finer just puts more silt in the cup without gaining much extraction; nudge coarser and the sediment drops off sharply. Paper traps fines, so it tolerates a finer grind and rewards it with a faster, more complete extraction — right up until the bed clogs.
So when you swap filters, change one thing deliberately and let the dust settle. Go a step coarser moving to metal, a step finer moving to paper, brew that way three or four times before deciding anything, and keep everything else fixed. If your grinder cannot hold a setting well enough to make that comparison meaningful, that is worth knowing too, and the piece on blade versus burr grinders covers why.
Matching the filter to the roast in the bag
After a couple of years of switching back and forth, my rule of thumb is simple. Light roasts, where the whole point is acidity and a specific origin character, go through paper — the clarity is what you paid for. Medium and dark roasts, where you want body and sweetness and chocolate, go through metal, which flatters exactly the things those roasts are trying to do. Blends usually land wherever the darkest component wants them.
What lives on my counter, and when each one comes out
A stack of rinsed paper cones for weekday mornings with whatever light roast is resting on the shelf, a stainless basket that comes out on weekends for the darker house blend, and a cloth filter in the drawer that I used faithfully for two months and now reach for about twice a year, because I am not honestly disciplined enough to keep it right. If you have never done the comparison, do it tomorrow: weigh out two identical doses of the same coffee, brew one through paper and one through metal back to back, and drink them side by side while they are both hot. It costs you one extra scoop of beans and it will tell you more about your own taste than another month of reading will.