Grind size, water temperature, the ratio of coffee to water, and how long the two stay in contact. Everything else is detail. This page covers all four, method by method.
Written for people who already own a kettle and want the cup to stop being a coin flip.
Each one pulls flavour out of the grounds a different way — by gravity, by soaking, by pressure, by steam. Pick one and stay with it for a fortnight before you judge it.
Water passes through the bed once and drains away. Gives the clearest separation of flavours, and punishes an uneven grind more than any other method.
The grounds sit in the water the whole time, so the timing matters far less. The metal filter lets oils through, which is why the cup feels thicker in the mouth.
A short steep followed by a gentle push through a paper filter. Cooler water than you would expect works best — it keeps the bitterness out of a concentrated brew.
Steam pressure forces water up through the basket. Start with pre-heated water and take it off the heat the moment it gurgles — that gurgle is the sound of it going bitter.
Finer grounds expose more surface, so flavour comes out faster. Coarser grounds slow everything down. When a cup is wrong, this is almost always the setting to change — before temperature, before ratio.
A scoop of coffee can vary by forty per cent depending on how the beans settle. Grams do not. This single change makes a cup repeatable more than any piece of equipment.
| Method | Ratio | Coffee | Water |
|---|---|---|---|
| Pour-over, one cup | 1:16 | 15 g | 250 g |
| Pour-over, two cups | 1:16 | 30 g | 500 g |
| French press, small | 1:15 | 23 g | 350 g |
| French press, full | 1:15 | 55 g | 825 g |
| AeroPress, standard | 1:14 | 16 g | 225 g |
| Moka pot, 3 cup | 1:10 | 16 g | 160 g |
| Cold brew concentrate | 1:8 | 100 g | 800 g |
Fifteen grams of coffee, two hundred and fifty grams of water just off the boil. If you learn one routine by heart, make it this one.
Hot water through the empty filter warms the cone and washes the paper taste out. Tip the rinse water away.
0:00Add twice the coffee's weight in water — 30 g — and wait. Trapped carbon dioxide escapes and the bed swells.
0:00–0:45Spiral outward to 150 g. Keep the stream slow and low; a hard pour digs channels in the bed.
0:45–1:15Top up to 250 g in the centre. A gentle swirl of the whole cone levels the bed as it drains.
1:15–1:45Let it run dry. A flat, even bed means you got it right. Craters or a wall of grounds mean the pour was uneven.
→ 2:45Altitude, climate and how the fruit is removed from the seed matter more than the country name on the bag — but the country is a decent shorthand for what to expect.
Nearly every disappointing cup is one of these. Learn to tell them apart by where you feel it — sourness hits the sides of the tongue immediately, bitterness sits at the back and lingers.
The water left before it had taken enough out of the grounds. The cup tastes empty and slightly savoury, and the finish disappears fast.
The water stayed too long and pulled out the harsh compounds behind the pleasant ones. The cup dries out your mouth and the aftertaste is ashy.
A brewed cup is about ninety-eight per cent water, so what comes out of your tap has a real say in the result. Water with almost no minerals in it — distilled, or heavily softened — makes coffee taste flat and hollow, because there is nothing there to carry the flavour compounds.
Very hard water goes the other way: the minerals scale up your kettle and mute the acidity that makes a good bean interesting. A simple carbon filter jug sits most tap water in a reasonable range, and it costs less than a new grinder.
Freshness matters on a shorter clock than people expect. Beans need a few days after roasting to settle, stay good for about a month, and lose most of their aroma within half an hour of being ground. Grinding to order is the single cheapest upgrade available.
Brew Atlas is an independent reference. There is nothing to buy here, no equipment to rank and no newsletter to join — just the small set of things that reliably change how coffee tastes, written down plainly.
Every number on this page comes from repeated brewing rather than from a manufacturer's leaflet, and the ranges are deliberately wide. Coffee is a variable ingredient; anyone quoting a single correct temperature to one decimal place is selling something.
The guide is revised as we retest. If a recommendation here contradicts your own results with your own beans and your own water, trust your results.