Roaster: Gene Café CBR-301
Batch Size: 250 g
Set Temperature: 250°C
Preheat: None
Airflow: Fan 2 → 3 at ~4:00
Cooling initiated at: ~13:10 (approx.)
Output Weight: 209.7 g
Weight Loss: 16.1%
Role / Intent:
Washed Central America contrast roast. Intended for A/B comparison against Roast 19 (Honduras Catuai Washed).
Why This Roast Exists
The question behind Roast 20 was simple:
Can I reliably describe structural differences between two washed Central American coffees without relying on flavour notes?
Roast 19 (Honduras Catuai, washed) landed at ~14.3% weight loss.
Roast 20 was intended to sit in the same structural band.
It did not.
Roast Behaviour
Drying and Maillard behaved normally.
- Yellowing progressed cleanly.
- Aroma shifted from hay to sweet baked bread.
- Chaff release began ~5:00 and largely cleared by ~7:00.
- First crack was clearly audible, beginning ~11:45 and building through ~12:15.
No instability observed.
Fan progression normal.
No temperature intervention.
The machine behaved predictably.
What Actually Happened
Roast 20 landed at 16.1% weight loss.
That is materially different from Roast 19’s ~14.3%.
Even though crack timing appeared broadly similar, the endpoint was later in total applied energy.
This means:
- Roast 20 sits in a deeper structural band.
- The roast level is not matched to Roast 19.
- Any cup comparison will reflect roast depth as well as origin.
This is not a failure.
It is a clean data point.
Structural Expectation (Pre-Cupping)
At 16.1% WL, I expect:
- Increased body
- Reduced perceived acidity brightness
- More rounded sweetness
- Slight compression of clarity compared to 14–15% band
If these characteristics appear in the cup relative to Roast 19, they will likely be roast-level driven rather than origin-driven.
System Observation
More important than the roast depth:
Roast 18 and Roast 20 (same beans) landed at essentially identical weight loss under the same method.
That confirms:
- Endpoint discipline defines outcome.
- The CBR-301 remains stable under repeat conditions.
- Weight loss remains a reliable structural anchor.
The machine did what it was told.
Learning
Roast 20 reinforces something important:
Crack timing alone is not a sufficient endpoint guide.
Total applied time and weight loss define structure more clearly than development seconds alone.
If a true washed vs washed comparison is required, this coffee will need to be repeated at ~14.5% WL to match Roast 19 structurally.
Current Status
This Note remains open.
A repeat roast targeting a lower weight loss is planned to complete the comparison properly.
No conclusions drawn yet.
Only observations.