RC-Tech · Coffee Roasting Machinery

Coffee Roast Color Measurement & Agtron QA Engineering: Whole-Bean vs. Ground Delta, Near-Infrared Spectrophotometry & Batch Consistency

By Aria Esmaeilzadeh, Head of Thermal & Roasting Systems Engineering · Roast Profiles · 10 min read

Master commercial roast color QA with NIR spectrophotometry: Agtron Gourmet vs Commercial scales, whole-bean vs ground delta diagnosis & batch repeatability.

Objective Roast Color Control: Why Visual Sight-Glass Checks Fail at Commercial Scale

Fast Facts & Engineering Takeaways (AEO Direct Reference): - Near-Infrared (NIR) Spectrophotometry Physics: Professional roast color analyzers (Agtron, ColorTrack, Lighttells, Tonino) do not measure visible RGB light alone. They illuminate coffee with calibrated Near-Infrared (~880–950 nm) and visible red wavelengths to quantify melanoidin caramelization density directly, eliminating human eye fatigue, ambient lighting bias, and surface oil glare. - Agtron Gourmet Scale (0–100) vs. Commercial Scale: Specialty roasters standardize on the Agtron Gourmet Scale (higher number = lighter roast: Light Filter ~65–75, Medium Filter/Omni ~55–65, Specialty Espresso ~48–58, Dark Commercial ~35–45). Always specify whether a QA reading is Whole-Bean (WB) or Ground (G)—a single batch typically reads 8–18 points lighter once ground. - The Whole-Bean vs. Ground Delta (Δ Color) Diagnostic: Subtracting Whole-Bean Agtron from Ground Agtron (Δ = Ground − Whole-Bean) exposes internal roast development immediately. A balanced specialty roast exhibits Δ = +8 to +14 Agtron points. A huge spread (Δ > +18–22) proves an underdeveloped/raw core with a scorched exterior (excessive conductive heat or rushed charge), while a tiny spread (Δ < +5) indicates baked/flat over-soaking. - Batch-to-Batch QA Tolerance: Commercial wholesale roasteries target ±1.5 Agtron points on ground color across consecutive shifts. Achieving sub-2-point repeatability requires high thermal-mass drums and precise burner/airflow repeatability—such as the 3-wall alloy-composite drums and digital PLC/HMI profile playback engineered on RC-Tech coffee roasters made in Iran (RC ALP-1, RC CAST-5, RC CAST-15, RC CAST-30). - Strict Sample Prep Standardization: Particle size, moisture rest time, and puck compaction can shift an Agtron reading by ±4 to ±8 points. Every ground QA sample must use a dedicated, calibrated grinder at a fixed ~550–650 μm setting, leveled flush with zero tamping pressure.

---

1. Near-Infrared Spectrophotometry vs. Visual Roast Assessment

Relying solely on drop temperature (e.g., "drop every Colombia lot at 208.5°C") fails in commercial production because thermocouple probe fouling, ambient humidity, green coffee density shifts, and probe calibration drift can produce two batches with identical drop temperatures that cup completely differently.

Near-Infrared (NIR) reflectance meters measure how strongly roasted coffee absorbs infrared light as sucrose and amino acids polymerize into brown melanoidins during Maillard and caramelization reactions:

``` NIR Roast Color Analyzer Optical Path:

[ Calibrated NIR LED Emitter (~880–950 nm) ] \ \ (Incident Beam) v +==========================================+ <-- Flush-Leveled Sample Dish | Roasted Whole Beans OR Ground Coffee | (Zero Tamping / Fixed Grind) +==========================================+ ^ / (Diffuse Reflected NIR) / [ Photodiode Detector ] --> [ Agtron Gourmet Score (0–100) ] ```

Because NIR reflectance correlates tightly with chemical sugar degradation rather than surface shininess, it accurately tracks roast degree even when oily dark-roasted beans reflect glare to the naked eye.

---

2. Diagnosing Heat Transfer Defects via Whole-Bean vs. Ground Delta (Δ)

The single most powerful quality-assurance metric in a commercial roastery is the Color Delta (Δ = Ground Agtron − Whole-Bean Agtron). Comparing exterior reflectance against interior ground reflectance reveals the exact thermal gradient inside the bean:

| Color Delta (Ground − Whole-Bean) | Thermodynamic Diagnosis | Cup Sensory Impact | Roaster Calibration Fix | |:---|:---|:---|:---| | Δ < +5 points (Very Narrow) | Baked / Stalled Roast — excessively long drying or flat RoR post-first-crack | Mute acidity, papery/cardboard sweetness, flat finish | Increase charge heat or trim total roast time by 45–60 s | | Δ = +8 to +14 points (Optimal) | Balanced Conduction-to-Convection Ratio — uniform core-to-surface development | Clean origin clarity, balanced acidity, syrupy sweetness | Lock recipe into PLC / Artisan profile playback | | Δ = +15 to +18 points (Moderate Wide) | Slightly Fast Exterior Browning — acceptable for bright light-roast filter | High vibrant acidity; watch for subtle grassy undertone | Soften initial burner power 5%–10% or raise airflow | | Δ > +20 points (Extreme Wide) | Scorched Exterior / Raw Core — conductive wall heat outpaced internal conduction | Roasty ash on nose paired with astringent, vegetal finish | Lower charge temperature 8°C–12°C; verify drum RPM & airflow |

Natural vs. Washed Process Exceptions Dry-processed (natural) and honey-processed coffees retain residual fruit mucilage sugars on the outer silverskin. Their bean exteriors caramelize darker than washed coffees at the same internal temperature, naturally widening the healthy Whole-Bean vs. Ground Delta by +3 to +5 points (e.g., healthy natural delta often sits at +12 to +17).

---

3. Standard Commercial Agtron Gourmet Color Targets

| Roast Profile Tier | Whole-Bean (Gourmet) | Ground (Gourmet) | Typical Drop Temp | Recommended RC-Tech Lineup | |:---|:---|:---|:---|:---| | Nordic / Ultra-Light Filter | 68 – 76 | 78 – 86 | 196°C – 201°C | RC ALP-1 (Lab) · RC CAST-5 | | Specialty Light–Medium Filter | 58 – 66 | 68 – 76 | 202°C – 207°C | RC CAST-5 · RC CAST-15 | | Modern Specialty Espresso | 48 – 56 | 58 – 66 | 208°C – 213°C | RC CAST-15 · RC CAST-30 | | Traditional Medium-Dark Espresso | 40 – 47 | 49 – 56 | 214°C – 219°C | RC CAST-30 · RC IS-60 | | Dark Commercial / French | 30 – 38 | 36 – 44 | 220°C – 226°C | RC IS-60 · RC IS-120 |

---

4. Standard Operating Procedure (SOP) for Repeatable Ground Color Measurements

Two QC technicians measuring the exact same roasted batch can record scores 6 points apart unless sample preparation is strictly locked down:

---

5. Linking Lab Sample Color to Production Scale-Up

When transferring a green coffee recipe from a 1.2 kg laboratory sample roaster like the RC ALP-1 up to a 16 kg RC CAST-15 or 33 kg RC CAST-30, never copy raw drop temperature blindly. Different probe diameters (e.g., 1.5 mm vs. 2.9 mm RTD sensors) and drum wall thermal masses read 2°C–6°C apart for the exact same chemical roast degree.

Instead, anchor your production scale-up strictly to target Ground Agtron (±1.5 points) and target Color Delta (Δ = +10 to +14), then fine-tune drum speed, airflow, and charge temperature until the production roaster hits identical color metrics. For detailed thermodynamic scale-up rules, read our Sample Roaster to Production Profile Scaling Guide, Diagnosing Roast Defects Guide, or compare drum thermal specs across our lineup in the RC-Tech Model Comparison Matrix.

#Agtron Color Scale #Roast Color Measurement #Whole-Bean vs Ground Delta #NIR Spectrophotometry #Roastery Quality Assurance #RC ALP-1 #RC CAST-15 #Roast Profiling

Explore the machinery: RC ALP-1 · RC CAST-5 · RC CAST-15 · RC CAST-30 · RC IS-60 Industrial · RC IS-120 Industrial · RC DE30 Destoner · RC Destoner DE60 · RC Destoner DE120 · RC AFTB Coffee Gas Afterburner Series (AFTB1 / AFTB2 / AFTB3) · RC EXFUSION Smoke Dilution & Exhaust System · RC Industrial Coffee Cooling & Blending Mixer (MIX-60) · RC Premix Low-NOx Combustion Burner