Extraction yield calculator.
A refractometer tells you how concentrated the drink is. It does not tell you how much coffee you pulled out of the grounds. This converts one into the other.
The formula
Extraction yield is the mass of dissolved coffee divided by the mass of dry coffee you started with:
EY% = (TDS% × beverage weight) ÷ dry dose
TDS × beverage weight gives you the grams of coffee solids actually dissolved in the drink. Divide by the dose and you have the fraction of the bean you managed to extract.
The critical input is beverage weight, not brew water. A V60 holds back roughly two grams of water per gram of dry coffee in the bed and filter, so 250 g poured over 15 g of coffee gives you about 220 g in the cup. Using 250 here instead of 220 inflates your yield by around 14% and makes a well-extracted brew look over-extracted. Weigh the cup.
The bands, and what they are worth
| Brew | Typical EY | Typical TDS |
|---|---|---|
| Filter | 18 – 22% | 1.15 – 1.45% |
| Espresso | 18 – 22% | 8 – 12% |
| Immersion / French press | 18 – 22% | 1.2 – 1.6% |
The 18–22% window comes from the mid-century Coffee Brewing Control Chart and it is a useful starting frame, not a verdict. Modern light roasts are routinely pushed past 22% and taste excellent; some dark roasts fall apart before 19%. The number describes the brew. It does not grade it. If a 23% cup tastes better than your 20% cup, the 23% cup is better.
Where the reading goes wrong
- Temperature. Refractometers are calibrated at a specific temperature. Reading a scalding sample gives a low number. Cool the sample to room temperature, or let the device settle if it compensates automatically.
- Undissolved fines. Suspended particles scatter light and inflate the reading. Filter the sample through a syringe filter for espresso, and let filter coffee settle before drawing the sample.
- An unmixed sample. A pour-over stratifies in the cup as it brews — the first drops out are far stronger than the last. Stir thoroughly before you sample or you are measuring one layer.
- A dirty prism. Residue from the last reading is the most common cause of an inexplicable outlier. Wipe and re-zero between samples.
- Estimated beverage weight. Guessing the cup weight defeats the entire exercise. This is the input the whole calculation hangs on.
What to change when the number is off
Extraction yield and strength are two separate dials, and this is where the calculator earns its keep — it separates them for you.
- Low EY, tastes sour. Under-extracted. Grind finer, or raise the temperature. See the dial-in guide.
- High EY, tastes bitter or drying. Over-extracted. Grind coarser or drop the temperature.
- EY in band, TDS low. Extraction is correct, the drink is just dilute. Tighten the ratio and leave the grind alone.
- EY in band, TDS high. Same thing in reverse — widen the ratio.
- EY in band, still tastes bad. The number is not the cup. Look at water chemistry, bean freshness, or an uneven bed before you touch the grind.
Coffee Aeye reads TDS directly from a DiFluid R2 over Bluetooth, computes the yield against the session it belongs to, and charts every run against the last so drift shows up as a trend instead of a number you have to remember. It sits alongside Darcy permeability, slurry temperature, particle distribution and DTR in the Lab. Get it on Google Play — €4.29, one-time.