I've converged on two basic shapes: linearly-declining RoR, and stepped. These curves of BT illustrate the difference:



Both roasts finish at the same temperature and reach first-crack at the same time. Despite that, they taste different as you might expect, and the people I've sampled them to are equally split on their preferences, which I think might be kinda ideal as a launch point for experimentation within each basic shape.

I finally made a profile for the Ikawa Pro100 that didn't taste like it was for sampling, roasted darker than medium, and didn't char or scorch. I've been trying all my naturals with it, which is not how you're supposed to do it. But it works really well for about half the greens I've tried (including my favorites) and doesn't suck on any of them.

I'll get around to varying this profile when I roast something washed or get up the nerve to experiment. Small batch size roasters are supposed to remove risk aversion, but I'm still afraid to experiment.

I find myself doing this too. When I changed from a Hottop to a Bullet my profiles changed a lot, even for the same beans. I might make some small tweaks here and there but overall the curves end up having a very similar shape for each roaster.

I experimented with different profiles on the Hottop and the Bullet. I didn't like some results, and other times the differences between my "typical" roast profile and the experimental one was so small to be barely noticeable. My theory is that I've learned how to work with each roaster's quirks.

However, sample roasting on my IKAWA is a whole other story. There are very noticeable differences in the roast profiles and it's fun to experiment with them. Plus if I don't like the results it's very little to toss out. I can roughly translate the profiles to the Bullet if something really stands out but I still tend to prefer my typical roast profile. It might be just familiarity, and when I try different approaches I end up making some kind of mistake so it's not worth it to try to fix something that is working just fine for me.

GDM528 wrote:I've converged on two basic shapes: linearly-declining RoR, and stepped. These curves of BT illustrate the difference:

Both roasts finish at the same temperature and reach first-crack at the same time. Despite that, they taste different as you might expect, and the people I've sampled them to are equally split on their preferences, which I think might be kinda ideal as a launch point for experimentation within each basic shape.

How do they differ in taste? If I want to roast in a low-acid Neapolitan style resembling Saka coffee, I'll purposely slow BT ROR early in the ramp from dry to first crack. Credit for this to OldmatefromOZ.

Starting with a disclaimer: I don't have a trained palate, nor did any of the other coffee drinkers I've shared the roasts with. This will be largely in terms of what the roast is reminiscent of. And it's also significant that I generally target City+ to Full City roast levels.

That said,
Linearly-declining RoR = "this is kinda like the stuff I buy", "a smooth winner", "thick crema", meaning similar to store-bought and tasted like its apparent roast level. This is why I use this roast profile when comparing different origins.

Stepped = "funky off flavor", "Norwegian fishing village", "smooth and sweet", "dark chocolate milkshake". Responses were bipolar, just like the roast profile. It seems to carry both light and dark roast notes, which each taster picks up and responds to differently. Also more sensitive to brew ratios.

Circling back to the OP's question: I may have converged on just two profiles - but I have not converged on exactly what to do with those profiles, and I move the points around pretty regularly. Perhaps what might seem like a rut is actually just a platform for more nuanced adjustments - it seems like the OP did exactly that by shifting around the finish time.

Do you notice any tipping with the stepped profile? Ramping temp when the beans are dry may cause them to tip easily? Also are those intake profiles?

One thing Ive noticed with intake profiles is that the roast level varies quite a bit between different coffees. This is to be expected since the beans will respond to heat output differently where the exhaust profile makes them stick to a "BT" profile so they all are heated to the same temperature. How do you target different roast levels with your intake profiles?

Milligan wrote:Do you notice any tipping with the stepped profile? Ramping temp when the beans are dry may cause them to tip easily? Also are those intake profiles?

One thing I've noticed with intake profiles is that the roast level varies quite a bit between different coffees. This is to be expected since the beans will respond to heat output differently where the exhaust profile makes them stick to a "BT" profile so they all are heated to the same temperature. How do you target different roast levels with your intake profiles?

The stepped profiles are actually 30sec ramps to minimize temperature overshoots which may (or may not) link to tipping. I may also have been lucky so far to not have any greens that are prone to tipping. The profiles in post #10 are BT, and are simply rounder versions of the inlet temperature curve. I also underdose the roaster to promote thorough mixing and single-mode heat transfer during the roast. The difference between inlet and BT is remarkably predictable, making it easy to target and hit any roast level.

The OP's profiles spanned 90 seconds and 10F finish time/temps. It so happens I coincidentally did the same 4-greens roast thing recently, and the inlet-controlled BT profiles literally stacked on top of each other: Ikawa Home thermal performance That's what I'd call a 'rut' The visible roast levels of the finished greens did look different, which I found reassuring that there is some point to shopping for different origins.

One thing I've observed with drum roast logs, is a half-dozen or more heat input adjustments during the roast, which changes the energy flow into - or out of - the ongoing chemical reactions. The net BT readings may look relatively steady, but there are possibly important modulations of energy flow at crucial times/temperatures that alter the sequence of the roasting chemistry in ways that won't be apparent from looking just at the BT curve.

Sort of; but this thread did get me to consider roasting darker a coffee I got from Chert to see if I liked it. Most of my coffees end up 21-22 on Roast Vision or 82-85 Agtron. I didn't love my first roast of the Guatemalan El Diamante which I only pushed to 10 degrees past FCs. I was doing a batch of roasts where I pushed them all to 22 degrees past FCs for coffees I was providing to others.

I decided to really push two I was keeping the El Diamante went to 30 degrees past FCs. I am surprised but 2 days later I am liking it way better. I now wonder what it would be if I were in the 20-22 range. This coffee measured 16 on Roast Vision which is 67 Agtron. So while my roast profile was similar in terms of when I changed gas and air, but I let it right with more heat.


I still believe that for most people, degrees past FC's is the easiest way to ascertain what they like. To see this on your final graph make sure you check the box in statistics.

I have a Guatemala Royal Gem that I've been playing with recently. I tried it at three roast levels. Roastvision 22, 18, and 16. The coffee opened up at the 16 and 18 roast level. The 22 had nice acidity, floral notes and a hint of caramel but it didn't taste as lively and flavorful as the darker roasts. The florals gave way to the tropical fruit notes that were described by Royal as banana and mango. While not on the nose, there was a textural and sweetness quality with just enough caramelization to hint at banana's foster in the 16 while the 18 was more of a ripe banana and slight chocolate and caramel.

While coffees do lose their delicate notes as they are roasted deeper, this isn't always a bad trade off if the coffee develops strong and pleasant notes further in. When the deeper roast level allows bitterness/roast to creep into a well extracted pour over, then I'm typically out though.

Like Michael, I'm typically at a 20-22 roastvision with most coffees I find myself enjoying via pour over.

Milligan wrote:
Like Michael, I'm typically at a 20-22 roastvision with most coffees I find myself enjoying via pour over.

My phone is full of of heartwarming messages from those that like it darker. Probably not as kindly put as you said it Tim.

I am not yet decided, but with more heresy I will share this. I roasted an Ethiopian YIRGACHEFFE WELDE HIRBE NATURAL GR1 to RV 13, and used it 30% in my Terry's Blend. She loved it; said it was really balanced. So who knows, I guess I need to go darker. The next week will be interesting.