The Truth About Roast Curves, How to Tune Flavour Step by Step
https://www.rubasseroasters.com/article_d.php?lang=en&tb=7&id=1260
Roasters love a tidy curve, although the cup does not always follow the graph. This guide breaks a roast into its working parts, shows what happens when you change the length of one segment at a time, and gives you a flavour map you can actually use. Everything here is grounded in practical experiments rather than dogma.
Hero image: “Roast curve segments and taste: related, but how?”. Caption, “Taste first, then read the curve.” Rubasse NIR Digital Coffee Roasters
What is a “perfect roasting profile”
Quick take: Curves are useful, taste still has the final vote.
Rate of rise brought welcome structure to roasting, although a pretty line does not guarantee a delicious cup. Even with digital logging and automation, you can feel more puzzled than before, which is why disciplined experiments matter. Change one variable at a time, hold everything else steady, then cup with intent, this is the only way to connect the curve to flavour with confidence.
The variables inside a roast profile
Why this matters: End temperature is not the only decision that shapes taste.
A working roast has three segments. Drying to yellow. Maillard from yellow to first crack. Development from first crack to drop. In the experiments referenced here, all batches were dropped at the same bean temperature, about 212.8 °C, with first crack temperatures between roughly 200 and 203.3 °C. Only one segment length was changed per test, the coffee was a washed Costa Rica, and the cups were compared for sensory differences.
Image: Diagram of the three segments. Caption, “Drying, Maillard, Development.” Rubasse NIR Digital Coffee Roasters
Experiment 1, changing the length of the Maillard phase
What to expect: Body and overall flavour weight are the most sensitive to this change.
When Maillard time was varied, body moved the most. The “overall flavourfulness” bar shifted clearly, with the four-minute Maillard treatment cupping as the most complex and robust of the set. If you are chasing weight and mid-palate structure, Maillard is your first lever.
Images: “Length of Maillard phase” graph, then the “Maillard results” bar chart. Keep the same order. Rubasse NIR Digital Coffee Roasters
Experiment 2, changing the length of the development phase
What to expect: Acidity type and intensity travel with development time.
Stretching development generally reduces total acidity and shifts it toward a softer, more malic impression. Shortening development leans brighter and sharper. Push the extension too far and you risk baking, so use this lever to round the cup, not to flatten it.
Images: “Length of development phase”, then “Development phase results”. Rubasse NIR Digital Coffee Roasters
Experiment 3, targeting a different turning point
How to read it: Turning point is a proxy for early energy application.
There appears to be a useful window for turning point temperature. Think of turning point as the moment the system settles, beans, air and probe meet somewhere in the middle. It also reflects how quickly you applied energy at the start. Different coffees absorb heat differently, so expect washed lots to prefer a slightly higher turning point, naturals a touch lower.
Images: “Turning point differences”, then “Turning point results”. Rubasse NIR Digital Coffee Roasters
Experiment 4, changing the peak rate of rise
Use case: A fine-tuning tool for sweetness and dryness in the early stages.
Earlier energy can be tuned by setting a higher or lower peak rate of rise. In the example shared, a washed Guatemala with a lower-than-usual peak was compared at about thirty four versus thirty nine degrees Fahrenheit per minute. The higher peak tasted distinctly sweet, although drier and sharper, which is a handy mental model when you want to lift sweetness without making the cup syrupy. You can also aim for a target peak rate of rise rather than a fixed yellowing time.
Images: “Peak ROR differences”, then “Peak ROR results”. Rubasse NIR Digital Coffee Roasters
Roasting dynamics and the flavour wheel
Why this is useful: A visual guide helps you troubleshoot without guesswork.
The session closes with an annotated flavour wheel that links different roast dynamics to flavour families. Use it as a quick diagnostic. If you find a trait you do not want, glance at the map to decide which segment to adjust next time. It will not replace cupping notes, it will make them faster to act on.
Image: Annotated flavour wheel. Caption, “Dynamics to flavour families.” Rubasse NIR Digital Coffee Roasters
Mindset for better experiments
A working mantra: Roasting is momentum, and you are managing it.
Time targets do not always translate between machines. Keep your eye on momentum and on the rate of rise, then you will see more of what needs adjusting in your own setup. Also, taste everything. Do not bin a roast only because the line offended a rule of thumb. Your best lesson often hides in an imperfect curve.
Practical steps to copy the tests
Field guide: Keep it simple and rigorous.
- Pick one coffee and lock drop temperature.
- Change only one segment length per roast, keep charge, airflow and power steps constant.
- Record first crack temperature and time for each.
- Cup blind, score for overall flavour weight, acidity type, body and sweetness.
- Map your notes to the dynamics flavour wheel, then choose the next single change.
Summary, connect each segment to a flavour outcome
Maillard length moves body and flavour weight. Development length shapes acidity and roundness. Turning point and peak rate of rise tell you how much and how fast you push early energy. Change one lever at a time, cup with intent, and your curve will start to read like a set of dials rather than a mystery.
