Understanding TDS and extraction yield.
The two numbers that move you from guessing to knowing. What they actually mean, the SCA targets that matter, and how to read a refractometer without overthinking it.
TDS is the percentage of dissolved solids in the brewed coffee. Extraction yield is the percentage of the original dry coffee that made it into the cup. The SCA target is 1.15–1.45% TDS for filter, 8–12% TDS for espresso, and 18–22% extraction yield for both.
The two numbers, in plain language
TDS (Total Dissolved Solids) is the strength of the brewed coffee — a percentage of how much of the liquid you're drinking is dissolved solids vs. water. A 1.30% TDS pour over means 1.30% of that cup, by weight, is coffee compounds.
Extraction yield is the efficiency of the brew — the percentage of the original dry grounds that dissolved into the cup. A 20% extraction means 20% of your starting dose ended up as flavor in the beverage. The other 80% stayed in the puck.
TDS is what your refractometer measures. Extraction yield is what you calculate from TDS, beverage mass, and dose.
Why extraction matters more than ratio
Most home brewers obsess over ratio (1:15, 1:16, 1:17). Ratio is useful as a starting point, but it's not what tastes good — extraction is. Two pour overs at identical 1:16 ratio can land at 17% and 22% extraction depending on grind, temperature, agitation, and water chemistry. The 17% will taste weak and sour; the 22% might taste harsh.
Ratio sets the strength range. Extraction determines whether the cup tastes good inside that range. The SCA brewing control chart famously plots TDS on one axis and extraction on the other — the "ideal" rectangle sits at roughly 1.15–1.45% TDS × 18–22% extraction for filter.
The formula
Extraction Yield % = (TDS × Beverage Mass) ÷ Dose
Example: pour over, 15 g dose, 240 g brewed coffee in the cup, TDS reads 1.32%. (1.32 × 240) ÷ 15 = 21.1% extraction. That's solidly in the ideal range.
Example: espresso, 18 g dose, 36 g shot, TDS reads 9.5%. (9.5 × 36) ÷ 18 = 19.0% extraction. Inside the target.
For espresso, the convention is to weigh the shot in the cup (not including any milk or added water). For filter, weigh the brewed coffee in the cup (not the brew water you started with — some of it is now retained in the bed).
Target ranges by brew method
| Method | Typical TDS | Target extraction |
|---|---|---|
| Espresso (traditional) | 8–12% | 18–22% |
| Pour over (V60, Origami, Kalita) | 1.20–1.45% | 19–22% |
| AeroPress | 1.30–1.60% | 19–22% |
| Batch brew / drip | 1.15–1.35% | 18–22% |
| French press (immersion) | 1.20–1.40% | 18–21% |
| Cold brew (concentrate) | 1.50–2.50% | 18–22% |
How to read a refractometer (without overthinking it)
- Cool the sample. Most refractometers want around room temperature. Hot coffee skews the reading.
- Filter or settle. For espresso, use a syringe filter (or let it sit and skim off oils). For pour over, just let bubbles settle.
- Two or three drops on a clean, calibrated prism. Close the lid.
- Read three times and average if you care about precision past one decimal.
- Wipe the prism clean with a microfibre and a drop of distilled water between readings.
Atago PAL-COFFEE, VST LAB III, and DiFluid R2 are the consumer choices. All three give comparable readings — pick one and stick with it; calibration drift across devices makes cross-comparison messy.
Reading the result: what to actually do
The point of measuring extraction isn't proving you have a refractometer. It's a next-action signal.
- Extraction below 18%. Under-extracted. Grind finer, raise temperature, extend contact time, or increase agitation. Cup tastes sour, weak, salty.
- Extraction 18–22%. In range. Use taste to fine-tune. Slightly thin near 18%? Tighten grind one notch. Slightly bitter near 22%? Coarsen one notch.
- Extraction above 22%. Over-extracted. Grind coarser, lower temperature, shorten contact time, or reduce agitation. Cup tastes bitter, harsh, hollow.
- TDS off-target but extraction in range. Adjust ratio. Strong but balanced = lower ratio (more water). Weak but balanced = higher ratio.
Where it stops being useful
Extraction yield is a strong predictor of taste, not a perfect one. Roasters increasingly push very light roasts to extract well above 22% with exotic processes — those cups don't always taste bitter, because the soluble profile is different. Use the targets as a default and trust your palate as the override.
The other limit: TDS only tells you how much dissolved, not what. Two cups at identical TDS and extraction can taste very different depending on what came out of the grounds. Measurement narrows the search; tasting still picks the winner.
The point of tracking it
A single TDS reading is noise. A hundred readings — sorted by coffee, by method, by grind setting — is a personal extraction map. You start to learn that washed Ethiopians peak around 20.5%, that your AeroPress runs hot at 21%, that the Mythos always wants two notches finer than the Niche on the same coffee. That's the long game refractometry actually pays off in.
Track TDS and extraction on every brew, automatically.
HomeBarista's extraction calculator + brew log + reference brews. See the calculator →
