Every roaster knows that voice. It’s the same one that makes us obsessively check our notes, reweigh samples three times, and second-guess our color readings.
I decided to answer it definitively by putting the Ikawa Pro coffee roaster sample roaster through what can only be described as roasting Groundhog Day: the same Colombian Supremo coffee, roasted 10 times in a row, with the exact same profile, back to back, measuring everything.
Just pure, beautiful repetition to see if this machine could deliver the consistency we need to do our work with confidence.
I did 10 consecutive roasts, 100g each, same bean (Colombian Supremo, washed), same profile (a standard cupping roast: preheat to 174°C, end temp 210°C, total time around 6:18). After each batch cooled, I weighed it and measured the color with an Agtron meter.
No variables. No tweaking. Just the kind of methodical testing that makes great coffee possible.
The goal was simple: see if the Ikawa coffee roaster could deliver repeatable results. Because in our profession, consistency isn’t just nice to have it’s everything. It’s what lets us confidently evaluate green samples, develop reliable profiles, and trust our quality control process.
Weight loss is one of the most honest metrics we have. Green coffee goes in, heat happens, water and volatiles leave, and your scale tells you exactly what occurred.
| Roast # | Code | Final Weight (g) | Weight Loss (g) | Weight Loss (%) | Agtron Color |
|---|---|---|---|---|---|
| 1 | F9F | 86.68 | 13.32 | 13.3% | 68.9 |
| 2 | EFF | 86.48 | 13.52 | 13.5% | 67.2 |
| 3 | E79 | 85.96 | 14.04 | 14.0% | 69.4 |
| 4 | O2A | 86.00 | 14.00 | 14.0% | 66.9 |
| 5 | 3A1 | 86.60 | 13.40 | 13.4% | 69.7 |
| 6 | A4F | 85.59 | 14.41 | 14.4% | 69.9 |
| 7 | 23C | 86.90 | 13.10 | 13.1% | 67.1 |
| 8 | 4A4 | 85.70 | 14.30 | 14.3% | 70.1 |
| 9 | 257 | 86.50 | 13.50 | 13.5% | 69.6 |
| 10 | 3A4 | 86.30 | 13.70 | 13.7% | 65.8 |
| AVERAGE | 86.27 | 13.73 | 13.7% | 68.46 | |
Key statistics:
The Ikawa coffee roaster delivered impressive consistency. Most roasts landed within half a gram of each other exactly the kind of precision we need when evaluating samples or developing profiles.
Look at that stability. No drift, no wandering. This is what lets us trust our data and focus on the coffee itself rather than wondering if our equipment is causing variation.
The Agtron meter gives us objective color data, cutting through subjective interpretation to show us exactly where our roast landed.
| Roast # | Code | Agtron Color | Notes |
|---|---|---|---|
| 1 | F9F | 68.9 | - |
| 2 | EFF | 67.2 | - |
| 3 | E79 | 69.4 | - |
| 4 | O2A | 66.9 | - |
| 5 | 3A1 | 69.7 | - |
| 6 | A4F | 69.9 | - |
| 7 | 23C | 67.1 | - |
| 8 | 4A4 | 70.1 | Lightest |
| 9 | 257 | 69.6 | - |
| 10 | 3A4 | 65.8 | Darkest |
| AVERAGE | 68.46 | - | |
Key statistics:
A 4-point spread across 10 roasts is excellent consistency, especially accounting for normal measurement variation.
The machine stayed remarkably consistent throughout all 10 roasts, giving us the reliable foundation we need for meaningful sample evaluation.
When you’re using an Ikawa coffee roaster for sample roasting, this data shows you can trust it to deliver repeatable results. Not perfection no roaster achieves that, but the kind of consistency that makes professional work possible.
| Metric | Average | Range | Std. Deviation | Coefficient of Variation |
|---|---|---|---|---|
| Weight Loss | 13.73g (13.7%) | 1.31g | ±0.48g | 3.5% |
| Color (Agtron) | 68.46 | 4.3 points | ±1.3 | 1.9% |
What this means in practice: Same bean + same profile = results within about 1 gram and 2-4 Agtron points. That’s the kind of precision that lets you confidently evaluate green coffee, knowing the differences you taste are from the coffee itself, not equipment variation.
After confirming the Ikawa’s consistency with a single coffee, curiosity took over: What happens when you use this exact same profile on completely different origins?
So I roasted three more samples same preheat, same end temperature, same timeline with completely different coffees
| Origin | Code | Agtron Color | Final Weight (g) | Weight Loss (g) | Weight Loss (%) |
|---|---|---|---|---|---|
| Colombian Supremo (baseline) | avg | 68.46 | 86.27 | 13.73 | 13.7% |
| Colombia (different lot) | D98 | 66.1 | 84.8 | 15.2 | 15.2% |
| Brazil Natural | F35 | 69.6 | 87.6 | 12.4 | 12.4% |
| Gesha Village Natural | C44 | 71.0 | 86.3 | 13.7 | 13.7% |
Look at the Gesha Village Natural: it produced the lightest color (71.0 Agtron) while having nearly identical weight loss to the Colombian Supremo baseline (13.7%).
This is fascinating. The Gesha lost the same amount of mass as the Colombian but stayed significantly lighter in color about 2.5 Agtron points lighter.
Meanwhile, the Brazil Natural showed yet another pattern: it retained the most mass (only 12.4% loss) and landed at 69.6 Agtron lighter than the Colombian baseline but darker than the Gesha.
The different Colombian lot (D98) behaved differently from the Supremo batch: higher weight loss (15.2%) paired with darker color (66.1), following the more traditional pattern we expect.
What’s happening here?
Bean chemistry and cellular structure create different roasting behaviors. The Gesha Village coffee, despite losing the same percentage of mass as the Colombian Supremo, underwent less browning. This suggests the Gesha has different sugar compositions, amino acid profiles, or cellular structure that affects Maillard reactions and caramelization.
It’s not about density alone it’s about how the specific chemistry of each coffee interacts with heat. The Gesha’s natural processing style, varietal characteristics, and terroir all contribute to how it develops color relative to weight loss.
The Brazil held onto more of its mass but still developed moderate color, while the different Colombian lot showed that even within the same origin, different harvest lots can behave distinctly.
The same roast profile produces distinctly different results across different coffees. This isn’t a limitation—it’s the beautiful complexity of what we do.
The corrected Gesha data actually tells a clearer story: same energy input (the profile), same mass loss, but different color development due to the coffee’s inherent chemistry. This is why we can’t just copy-paste profiles across origins.
The key insight: Weight loss and color are independent variables influenced by different aspects of bean chemistry. The Gesha Village Surma lot proves you can have identical weight loss to another coffee while developing completely different color. Understanding this relationship and knowing it varies by coffee is what separates mechanical roasting from skilled roasting.
This is why we measure. This is why we test. This is why roasting remains endlessly fascinating every coffee reveals something new about how its unique characteristics interact with heat.
In profile replicating mode the Ikawa coffee roaster will roast each sample consistently, so the differences you taste come from the coffee itself. But expect different origins and processing methods to develop slightly differently even with identical profiles. I would still use the same profile on multiple different origins, but there is a chance that sometimes you might have to adjust the profile and re-roast a sample which is totally OK.
The Ikawa Pro coffee roaster roaster delivers exceptional consistency: ±0.48g weight loss variation and ±1.3 Agtron points across 10 identical roasts. This is the reliability that makes professional sample roasting possible.
But the cross-origin test revealed something equally valuable: machine consistency and coffee variability are two different dimensions of our work. The Ikawa sample roaster executes profiles with precision. Different coffees respond to those profiles in wonderfully complex ways.
This is why we love this profession. The machine provides the reliable foundation. We bring the knowledge, curiosity, and skill to understand how each coffee wants to develop.
The Gesha Village that lost the same 13.7% as the Colombian baseline but stayed 2.5 Agtron points lighter? The Brazil that retained more mass but still developed moderate color? These aren’t problems to solve—they’re the fascinating complexity that makes roasting endlessly engaging.
So yes, you can trust your Ikawa’s consistency. And you can trust that every coffee will teach you something new about how heat, structure, and chemistry interact.
That combination of reliable tools and infinite variety? That’s what makes this work worth doing.
PS: In my experience the reliability suffers on the IkawaPRO 100x coffee roaster in auto first crack detection and auto development mode please check my previous article: IKAWA PRO 100x AI First Crack Detection – Brilliant or Busted?
Author :
Session expired
Please log in again. The login page will open in a new tab. After logging in you can close it and return to this page.
