Filter Coffee Simulator

The sliders in detail: Coffee (dose) 8–80 g. The starting recipes of the methods are guide values to get going; in October 2026 we set them so that in the model they sit close to the classic standard (e.g. Chemex 520 instead of 600 µm, batch 560 instead of 700 µm, AeroPress 1:15.5, Clever 1:17, siphon 1:16) — recipe values, not a change to the formula. Ratio 1:10 to 1:20 — at 18 g and 1:16.7 that is around 300 g of water, and the tool works it out for you right next to it. We work in grams, not millilitres: the scale is more accurate than any measuring jug. Grind size 200–1200 µm, and those are real micrometres: the mean particle diameter. Your grinder does not show µm, which is why the method's starting value is your anchor. Water temperature 60–99 °C. Pouring technique (agitation) little, medium, a lot: how much you move the coffee bed — stirring, pouring vigorously, swirling. More movement brings more fresh water to the particles and dissolves more. Rest since roasting 0–45 days. Total hardness 17–250 ppm and alkalinity 0–120 ppm: ppm means «parts per million», i.e. milligrams of dissolved mineral per litre of water, for hardness and alkalinity usually expressed as calcium carbonate (mg/l CaCO₃). 100 ppm corresponds to around 5.6 °dH.

The brew time — a correction: In the old version, «brew time 45–360 seconds» was a slider like any other. That did not fit the way we brew. With percolation (V60, Kalita, Chemex, Origami, batch, moka) you do not set the time: it follows from grind size, dose and pouring technique. That is why the slider there is now called «flow-through time, observed, not a target» — you enter what your clock shows, and the model only needs the number as contact time. With immersion (French press, AeroPress, Clever, siphon, cold brew) the steep time is an ingredient of your recipe: you decide when to press. There too, it is not a target. That is why the tips under the result say «grind finer — the flow-through gets longer by itself», not «brew longer».

The results in detail: EstimateTDS (strength) and extraction (EY) are both given in percent, with a real unit and a real axis on the brewing control chart. Even so, they are calculated values from the model, not a measurement of your cup: they tell you roughly where your recipe lands and in which direction a slider pulls — not what a refractometer would show. That TDS is shown with two decimal places is calculation precision, not measurement precision. The beverage weight in grams is calculated: the model assumes that 2 g of water per gram of coffee stay behind in the coffee bed. The taste description is an orientation, not a tasting verdict.

A calculation correction: Extraction is defined as TDS × beverage weight ÷ coffee weight. In the old version that did not hold for immersion — the displayed extraction was 2 to 4 percentage points above what followed from TDS and beverage weight, and the dot did not sit on its own ratio line. Now the balance holds for all methods; we readjusted the immersion calibration constant so that the French press starting recipe lands at the same extraction as before (20.1 %). For immersion, strength is therefore shown somewhat higher than before — that is the geometry of the chart, not new knowledge.

The brewing control chart — what is new in 2023: Peer-reviewedThe classic Brewing Control Chart (Lockhart 1957) called the range 18–22 % extraction and 1.15–1.35 % TDS «ideal» — that is also the SCA Golden Cup standard. Guinard, Frost, Batali, Cotter, Lim & Ristenpart redrew the chart in 2023 with sensory panels and consumer data (Journal of Food Science 88, 2168–2177). Result: the same range is now only called «classic standard», because it does not hold for all preferences. Sour and citrus sit at high strength and low extraction, bitter and roasty at high strength and high extraction, «sweet» at low strength, tea-like at high extraction and low strength. 57 % of those surveyed (young adults in Northern California) liked light, low-bitterness coffees best, 43 % strong or light, acidity-driven ones. We renamed the tool accordingly: no «ideal», no «Golden Cup» trophy — a classic standard as orientation, your taste decides. The percentages come from this one group and cannot be transferred to Switzerland.

The basis: EstimateThe extraction kinetics follow Liang, Chan & Ristenpart 2021 for immersion and Moroney et al. 2015/2016 for percolation, calibrated so that the starting recipes sit close to the classic standard. The sensory mapping relies on Guinard et al. 2023, the preference groups on Cotter et al. 2021 and Guinard et al. 2023. Water chemistry deliberately acts only as a small modifier (±1 %). The earlier justification for that («SCA: hardness effect 20–250 ppm insignificant») we could not back up — it remains here as an unproven assumption and is labelled as such in the tool. The sources are documented; how we combined them into one formula and applied it to the eleven methods is our modelling: estimated, not measured.