Cupping and Sensory Evaluation
Cupping on Roest is used for two overlapping purposes: evaluating green coffee and diagnosing roast outcomes. This page describes how to structure a cupping session, when to taste, how to compare roasts fairly, and how to interpret common sensory signals without over-trusting any single roast metric.
What Cupping Is Best For
Cupping is strongest as a comparative tool: same coffee, same preparation, multiple roast approaches, tasted side by side. It is especially useful for green buying, sample evaluation, and learning how roast changes affect acidity, sweetness, bitterness, aroma, body, and aftertaste.
Cupping is weaker as a complete quality-control method when used alone. A roast should also be checked in the way it will actually be served: filter, espresso, batch brew, or another target method. Cupping has much longer contact time than most brews, so it can flatter or punish coffees differently than normal preparation; one contributor contrasted cupping’s 20–30 minute contact time with typical brewed coffee at around 5–6 minutes or less 2 sources. For brewing-oriented QC, pair this page with Resting and Degassing, Color Reading and Measurement, and Roast Defects Troubleshooting.
Small Roest batches have a specific role in this workflow. 50g batches are generally treated as cupping and quick evaluation batches, not as the best indicator of filter brewing behavior; larger daily-drinking batches belong in Batch Size Scaling 2 sources.
Canonical Evaluation Workflow
Use one repeatable workflow before changing roast parameters. The goal is not to make a perfect universal cupping recipe, but to keep the tasting method stable enough that changes in the cup can be attributed to the roast or green coffee rather than the evaluation setup.
1. Decide the purpose of the session
| Purpose | Best timing | Notes |
|---|---|---|
| Immediate roast sanity check | Same day or out of the roaster | Useful for obvious roast defects, aroma, and gross under/over cues, but not final judgment. |
| Early comparison | 24h to 5 days | Can reveal direction, but some contributors find this period noisy or misleading. |
| Main roast evaluation | 7–14 days | A practical default window for comparing most light Roest samples. |
| Longer-rest evaluation | 14–30+ days | Useful for very light, high-density, or air-roasted coffees that change substantially with rest. |
A practical evaluation schedule is: cup once early if feedback is needed, then re-cup at about 7 days and again around 14–18 days, because coffees can change substantially over that period source. For cleaner taste and more intense notes, one contributor recommended waiting at least 10–14 days before evaluation source.
2. Prepare the cups consistently
Use the same grinder, grind setting, water, dose, bowl, and timing across all cups. Community protocols vary, but two repeatable reference points are 12g coffee to 200ml water at 96°C, or 12.5g coffee to 200ml water at 94°C 2 sources. curated When the result matters, cup at least two bowls per roast or green sample, because a single defective bean, quaker, grind clump, or dosing error can make one cup taste misleadingly papery, dirty, hollow, or astringent.
A simple repeatable procedure:
- Grind all samples at the same setting.
- Dose 12–12.5g coffee per 200ml water.
- Pour 94–96°C water.
- Break the crust at 4 minutes if following the 12.5/200ml protocol.
- Begin tasting around 10–12 minutes.
- Revisit the cups through cooling, especially 18–30 minutes.
- If a roast is confusing, keep tasting to room temperature.
The cooling phase is critical. One common approach is to taste at 8–10 minutes, then every 5 minutes up to 30 minutes source. Another protocol focuses on 18–25–30 minutes and even up to 1 hour after pouring, because some defects and some positive aromatics only become clear as the cup cools source.
3. Compare blind where possible
Side-by-side cupping is more useful when the taster does not know which cup is which. Blind comparison is especially important when evaluating small differences in roast profile, airflow, batch size, or development time. Several community comparisons used blind cupping, and controlled cupping has sometimes contradicted informal impressions; for example, one airflow comparison had a general but not total informal consensus, while a more controlled cupping with Q graders found no detectable difference source.
Avoid smelling directly from bags when bag material may interfere. One contributor warned that bags can contribute wet-cardboard or glue aromas and distort whole-bean aroma notes source.
4. Record both sensory and roast context
Useful notes include:
- dry fragrance
- wet aroma
- first hot taste
- sweetness, acidity, bitterness balance
- body and texture
- aftertaste
- cooling behavior
- any defects
- roast age
- brew or cupping recipe
- grinder and water
- roast log link, batch size, weight loss, drop temperature, and color if measured
A simple notebook or notes app can be more useful than a rigid form. Several experienced users prefer free-text notes or pen and paper over structured cupping software, and Roest has discussed adding free-text cupping notes linked to curves and inventory source.
5. Brew the promising samples
After cupping, brew the roasts using the intended serving method. Cupping alone can mislead, because taste changes substantially with recipe, grinder, water, and contact time. One contributor stated that true QC should be done according to the coffee’s intended use, whether espresso, filter, batch brew, or another format source.
How to Read the Cup Through Temperature
Hot cups can hide differences. As cups cool, sweetness, acidity, bitterness, body, and aftertaste often separate more clearly. In one sensory training exercise, development-time differences did not make hot flavors radically different, but acidity, sweetness, and bitterness balance changed substantially; more development lowered acidity and allowed more bitterness to creep in source.
Cold cup behavior is especially diagnostic. A fast profile in that same exercise held best when cold, becoming more integrated, crisper, and clearer in flavor source. Other experienced tasters also report that the biggest differences can appear in dry fragrance and cold cup rather than the first hot sips source.
Do not stop at the first impression. Some cups begin flat, smoky, or astringent at 10 minutes and become sweet, clean, floral, or fruit-forward around 20 minutes; others collapse, become papery, metallic, bitter, vegetal, or hollow as they cool. This is why the workflow above includes repeated tasting through the cooling curve.
Sensory Signals and Likely Interpretation
| Sensory signal | Common interpretation | Practical next step |
|---|---|---|
| Grain, cereal, corn, peanut | Less development or underdevelopment | Check development, earlier heat application, and whether the cup improves with rest source. |
| Corn water, green plant, vegetal, hay | Underdeveloped or insufficient internal development | Review development time, Maillard/drying balance, and first-crack energy; see Development Time and Drop Decisions. |
| Papery, cardboard, weak, hollow | Often underdeveloped, stale, defective green, or poor extraction context | Re-cup after rest, brew with a controlled recipe, and inspect/sort defects. |
| Astringent, metallic, sharp chemical finish | Can be underdevelopment, water/extraction interaction, or roast-profile issue | Re-test with stable water and grind; do not rely on one cupping time. |
| Ashy, roasty, burned, smoky | Roast defect, tipping, scorching, excessive heat, or airflow/venting issue | Inspect beans and review Roast Defects Troubleshooting. |
| Dryness | May be true dryness, astringency, or bitterness | Separate tactile drying from bitter/astringent flavor; dryness can be confused with other defects source. |
| Flat, baked, low fruit, low complexity | Possible baked roast or overly long/low-energy profile | Compare against a faster or more energetic roast and track aroma loss through cooling. |
| Clean but thin, tea-like, low body | Very light or high-air profile, or insufficient body development | Decide whether the target is delicate cupping clarity or brewed-coffee body. |
| High sweetness, heavy body, low acidity, brown sugar/caramel | More developed or more traditional profile | Confirm it is desirable rather than masking acidity or florals. |
| Tomato in Kenya | Contested descriptor, not automatically a defect | Treat as context-dependent; one trusted contributor finds tomato desirable, while others see it as a roast-approach issue or defect-like note source. |
A roast can be both roasty and underdeveloped. Likewise, a roast can be very clear yet well developed. Sensory evaluation should therefore distinguish roast degree, flavor development, internal development, and extraction behavior rather than reducing the cup to one metric.
Resting, Degassing, and Evaluation Windows
Roest and other air-roasted coffees can change dramatically over rest. Some coffees taste good immediately, close down the next day, become difficult around days 5–10, and then improve again later. Others peak early and fade quickly. For this reason, a single cupping session rarely gives a complete picture; see Resting and Degassing for the dedicated treatment.
evaluation timing varies by roaster, coffee, and purpose. Some contributors cup or brew straight out of the roaster for immediate feedback, and one noted that same-day can be useful while the next day can be worse source. Others wait at least 4–5 days because early cups show distracting notes, and one contributor said under 10 hours may taste okay but not like after 5 days 2 sources. The practical compromise is to use same-day tasting only as an early diagnostic, then re-cup in the main window from the workflow above.
Pre-grinding for degassing is also contested. Some users grind and rest coffee in a closed container for later evaluation, while others dislike pre-grinding because it never tastes the same and may change more than just degassing source. If pre-grinding is used, record it as part of the protocol rather than comparing it directly with freshly ground cups.
Why Metrics Do Not Replace Tasting
Color, weight loss, density, moisture, and development time are useful only when interpreted together with the cup. A darker color does not always mean better flavor development, and similar color or weight loss does not guarantee similar taste. One comparison found a worst L100+ roast at color 115 tasting more underdeveloped than a better roast at color 119, supporting the warning that color alone is not a reliable indicator of flavor development source. Color readings also depend strongly on sample preparation, device, grind size, and calibration, so they should be used as internal references unless the preparation protocol is shared; see Color Reading and Measurement.
Weight loss is similarly contextual. Some coffees taste developed at surprisingly low loss, while others need higher loss because of process, moisture, density, or target profile. Treat weight loss as a tracking tool, not a score; see Weight Loss Targets.
The same caution applies to development time. A fixed number of seconds after first crack can produce different results depending on roast speed, momentum, batch size, process, and when first crack is marked. Development decisions belong with taste, crack behavior, drop temperature, and intended use; see First Crack Management and Development Time and Drop Decisions.
Sensory Training and Calibration
A practical way to improve sensory skill is to repeatedly smell and taste reference materials and organize them into categories. Fruit teas, fruit syrups, and fruit jams can be used to build memory for red fruit, yellow fruit, citrus, tropical fruit, and cooked-fruit families source. This is not a substitute for cupping coffee, but it helps tasters name what they perceive.
External calibration is also valuable. Coffee meetings, sending samples to other tasters, and comparing against coffees from respected roasters help prevent a roaster from calibrating only to personal habits. Several contributors describe using side-by-side commercial references or external feedback to understand what their own roasts are missing; this is especially useful when working on Green Coffee Selection or adapting profiles from Profile Sharing and Starting Points.
The strongest cupping practice is therefore simple: taste consistently, taste more than once, compare blind when possible, brew the coffee as it will be served, and write enough notes that the next roast decision is based on evidence rather than memory.