Drying and Maillard Phases
Drying and Maillard timing determine how heat moves through the bean before first crack, how evenly the interior develops, and whether the cup presents as clean, sweet, vegetal, hollow, baked, or roasty. On Roest, these phases are especially sensitive to batch size, airflow/pressure, inlet shape, and how the yellowing event is defined. This page explains how to interpret the phases, choose practical phase targets, and adjust a profile when the cup suggests the early or middle roast was off.
Phase Boundaries on Roest
The drying phase is normally treated as the period from charge to yellow or dry end. The Maillard phase is the period between yellow and first crack source. Because Roest can auto-mark yellowing at a set bean-temperature threshold, the yellow event on a graph is not always the same as visible color change. Operators should watch the roast, set the yellow threshold to match what they actually see, or correct the event later in the log source.
Roest yellowing defaults are reported differently across discussions. One report says the default yellow marker is 165°C BT, while another says Roest marks yellowing by default at 175°C BT if not overwritten 2 sources. Treat the auto marker as a configurable logging aid, not proof of actual dry end.
Yellow temperature also shifts with batch size. Reported practical ranges are roughly 170–175°C for 100g, 160–165°C for 150–165g, and 150–155°C for 180–200g batches source. This is one reason phase advice should not be transferred blindly between 100g and 180–200g work; see Batch Size Scaling.
Starting Targets and Adjustment Recipe
Use these targets as first-pass profile design references, then evaluate by cup, color, weight loss, and bean appearance. The ranges are not universal laws; they are practical starting points from Roest community practice.
| Use case | Drying / yellow target | Maillard target | Development reference | When to use |
|---|---|---|---|---|
| General light/filter Roest baseline | Keep drying above about 3:00; many profiles sit around 3:30–4:00 | At least 2:10–2:30 | 30–60s | Default starting point for clean, light filter roasting 2 sources |
| 150–185g compact profile | 3:00–3:30 to about 150°C BT | 2:10–2:30 more to first crack | 35–45s | Useful when working in the larger, more reliable BT range on Roest source |
| 150g reference target | Yellow around 155°C at 3:40–4:00; first crack after 6:00 | — | — | A 150g reference for beans that should not be stretched too long early source |
| Fuller / more body profile | 4:30–5:00 to yellow | 3:15–3:30 | 1:00+ | Use when the cup needs more body, sweetness, and traditional roast character source |
| 200g production-style reference | 3:40 to yellow | 2:30 | 45–60s | A 200g target used by a production roaster; do not assume it applies to 100g source |
| Natural-process slower reference | First crack around 8:00–8:30 | — | ~40s | Can be acceptable for some naturals when the roast remains clean and not roasty source |
A practical adjustment sequence is:
- Pick the batch-size family first, then set the yellow marker to the visible color change for that batch. Do not use 100g yellow temperatures as if they apply to 180–200g.
- If yellow arrives before about 3:00, reduce early heat, lower early inlet, add a soak, or slow the ramp. Drying shorter than 3:00 has been described as damaging or producing poor results, while Maillard shorter than about 2:10–2:30 is associated with hay-like, low-fruit cups source.
- If drying is on target but the cup is hollow, vegetal, hay-like, or weak, lengthen or energize Maillard rather than only adding post-crack development.
- If the cup is dull, heavy, muddy, or dark on the outside, reduce excessive heat pulses and avoid letting the middle phase drag without purpose.
- If more weight loss is needed, adjust drying/inlet/charge rather than reflexively extending post-crack time; weight loss is discussed separately in Weight Loss Targets.
Why Drying Quality Matters
Drying is not only the time needed for the beans to turn yellow. It determines how moisture and heat distribute through the bean before Maillard chemistry begins. Moisture conducts heat effectively relative to the coffee structure itself, so a drying approach that leaves the center wet while drying the outer shell can lead to superficial roasting, uneven color, and an underdeveloped center 2 sources.
Too-short drying can show up as weak aroma, lack of sweetness, vegetable smells, astringency, or tea-like thinness. In one geisha comparison, roasts with about 3:30 dry instead of a 4:20 benchmark developed an unpleasant vegetable smell and lacked sweetness in aroma, while the longer-dry benchmark showed fruit and much stronger sweetness source. A faster Ethiopia natural roast was described as astringent, thin, and tea-like, while the longer-dry version was sweeter and less astringent source.
Too much early heat is also risky. Early heat can create brothy, savory, smoky, or dry-distillate flavors and may age poorly, while very high early inlet on Roest can roast the exterior faster than the interior 2 sources. For detailed control of early heat shape, see Inlet Temperature Management, Power Curve Strategies, and Soak Phase Techniques.
Managing Maillard
Maillard is the main phase for building sweetness, body, and roast-style character before first crack. Longer Maillard has been associated with more body, more sweetness, and a more traditional cup, while shorter Maillard can produce a cleaner, lighter roast with fewer caramel or traditional roast flavors 2 sources.
The risk is on both sides. If Maillard is too short, cups may become soft, vegetal, hay-like, bland, or low in fruit. If it is too long, the outside can darken excessively, body can become heavy, and cups can move toward baked, toasty, or muddy impressions. One production-roast review found many preferred roasts around 3:15 Maillard, while another Roest target treats 2:10–2:30 as a minimum practical floor rather than a universal optimum 2 sources.
A rising inlet or ET after yellow can shorten Maillard and push acidity or weak flavor when the middle phase becomes too fast. Lowering inlet after yellow is a common correction when the roast races from yellow to first crack source. Conversely, if the cup is flat, underdeveloped, or malty despite adequate post-crack time, the fix may be more inner-bean development earlier in the roast, not simply more development time after first crack source.
Process and Bean Differences
Natural, fermented, or heavily processed coffees may behave differently from lower-moisture washed coffees, but processing style alone does not reliably indicate green moisture. Moisture and heat needs should be judged from measured moisture, density, bean behavior, and cup results rather than from the process label alone. Some processed coffees have been described as needing more early power through charge and drying, while others can accelerate later in the roast, mark easily, or burn sugars on the surface if heat is not controlled 2 sources. For dedicated process guidance, see Natural Process Roasting and Washed Process Roasting.
Dense, large, or high-moisture beans may need more time or more early energy. A washed Indonesia example only became good after charging 50°C higher; adding 15–20s of later development did not fix the low-charge underdevelopment flavor source. Ethiopia can also take longer to dry, and attempts to force shorter drying with much more early power did not always give good results 2 sources.
Small beans and some Ethiopians may benefit from shorter total times than larger or wetter beans, but that does not mean drying and Maillard can be collapsed indiscriminately. Smaller Ethiopia examples include 4:00+ dry, around 3:00 Maillard, first crack after 7:00, and total roast around 7:45 as one proposed reference source. For origin-specific notes, see Kenya and Ethiopian Coffees.
Airflow, Pressure, and Moisture Retention
Airflow affects phase behavior by changing heat transfer, moisture removal, pressure, and inlet response. Lower airflow before yellow has been used to keep moisture in the drum longer and dry more evenly, while later increases help remove chaff, smoke, and gases 2 sources. Detailed fan and pressure settings belong in Airflow and Fan Settings and Pressure Management, but phase work should not ignore them.
A common Roest pattern is near-neutral or minimal negative pressure up to yellow, then more negative pressure as the roast approaches first crack. Christopher Feran described running slightly positive until about yellow or halfway, then neutral or slightly negative, and up to about -10 at crack source. Machine differences are significant, so fan percentages should be treated as local calibration values rather than portable targets.
Reading the Cup
Phase problems are easiest to identify by combining cup, color, weight loss, and bean appearance rather than relying on RoR alone. Weight loss can show whether a roast is on track or far off, while color and whole-bean appearance reveal whether development is mostly surface-level or more even; see Color Reading and Measurement and Roast Defects Troubleshooting.
Common interpretations are:
- Sharp acidity with low sweetness and body can indicate a start that is too fast or a dry/yellow point that should be pushed later.
- Hay, vegetal, bland, or weak cups often point to too little Maillard time or insufficient inner development before first crack.
- Heavy body, low acidity, flat fruit, slight bitterness, or burned-bread aftertaste can point to an overlong or poorly energized middle/end phase.
- A cup that seems underdeveloped despite enough post-crack time often needs better drying and Maillard development, not just a later drop.
First crack timing and post-crack decisions are covered in First Crack Management and Development Time and Drop Decisions.