This page collects the roast notes from 2026 — the main working record of Coffee | One Roast at a Time.
These are not recipes or instructions. They are learning records from real roasts on a Gene Café CBR-301, written at the time of the roast. The most useful numbers in any note are usually batch size, start temperature, total roast time (TRT), weight loss, and cup result where recorded.
A heater repair occurred mid-project. Roasts from Roast 70 onward reflect a period of post-repair recalibration — the machine behaves differently and earlier assumptions needed revisiting.
Start here within the notes
- Roast 91 and 92 — Same Plan, Different Result — the most recent summary of where things stand
- Roast 70–75 — The Repaired Gene Café Has a Different Fingerprint — post-repair behaviour explained
- Roast 36 & 37 — the cold vs warm start comparison, the most referenced result in the archive
- Roast 8 — Brazil Topazio — the upper-bound reference roast used for comparison throughout
All 2026 notes
-
Brazil Santos Natural — Gene Café CBR-301 Sequential Heat Saturation Session
-
Brazil Santos Natural | 250g → 207.4g (17.0% WL) | TRT Validation (Airflow Inflection)
-
Roast 26 – Brazil Santos (Natural) | TRT Push Upper Boundary
-
Note – Roast 25 | When Time Wasn’t Enough
-
Note – On Repeatability, Correlation, and the Urge to Overcomplicate
-
Sample Roast 3 – Regional Jess (Huila, Washed Blend)
-
Sample Roast Comparison – 52g Batches on the Gene Café CBR-301 | Machine Behaviour Study
-
Roast 24 – Brazil Santos Natural – Wanted Controlled post-crack development extension
-
Roast 23 – Wanted Santos Arabica (Natural) | Espresso Baseline 1
-
Roast 17 — Ethiopia Guji Karume (Natural)Espresso Calibration Morning