Hey everyone,

This past year, I've been experimenting with obtaining the most uniform development across the coffee bean as my aim - and I have finally attained a general roadmap for that result. Assuming relatively dense beans (Ethiopia, Kenya, Colombia, Guatemala), I am hitting these markers:

Charge: 220C BT, Low airflow
Yellow: 153C @ 5:45, medium airflow
Increase Airflow to max @ 190C
FC: 200C @ 9:30
Drop: 208-210C @ 10:15-25 (45-55s post 1C Dev)

I am getting a juicy, rounded flavor profile with minimal caramelization (although there is still some there) and no hints of underdevelopment. The reward is coffee that is juicy and lively upfront with a super sweet lingering finish. (Underdevelopment would ruin the finish aspect, tasting like wet wood or soil, and be a little sharp in acidity. Slight overdevelopment would just mean less lively fruits but still delicious. Over-roasting tastes like hints of ash from even the tiniest amount of internal scorching/tipping).

When successful, I liken my coffee to be like a very fresh fruit tart - full of juicy fruit flavor with accents of delicious baked graham cracker pie crust - like 80/20 fruit to caramelized base notes.
---

The other day, I was at Black Fox in NYC, and I tasted two drip coffees (a WP Ethiopian from 49th Parallel and a washed Colombian from George Howell) that were both outstanding. I must say, however, that their roast profile was vastly different than mine. They added a ton of post 1C caramelization which contributed to a profound "coffee" flavor accented by the brighter origin notes. I found these coffees super enjoyable but at the same time totally different than my approach. This was more like a coffee/chocolate cake or pudding base with fruit on top. More like a 50/50 representation of the two elements.

I'd assume that their approach is closer to:

4-4:30 Yellow
7:30-8:00 FC
~2min development

Roast time is similar but the amount spent in post FC dev is very different. I would conjecture that the inside half of the beans were roasted to the same dev for both my coffee and these other coffees whereas the portion just outside of that were developed very differently - mine continuing in the same level of dev as the inside until the edge where as the longer dev coffees would be more caramelized. Both very delicious with vibrancy and clarity, but different.

So my question is: Do you roast fast with long development (like the 49th Parallel and George Howell coffees)? Or do you prefer to minimize caramelization and highlight acidity/sweetness?

Depends on what I'm going for. I like longer development times and shorter roasts for lighter coffees that I need high solubility for, like for washed Central American espresso.

I find there are two main roast paths for me: 1) a variety of tastes that are somewhat less distinct, transitioning on the tongue flowing and changing 2) a singular distinct flavor, full and lasting
The artisan aspect of roasting comes into play when I am thinking of how I want to shape the flavor in the cup and thus; how long or what kind of heat is needed. Generally what can I do to the coffee that will get me the desired roasts. The technical aspects come into play when I shape a profile. The way I roast for No. 1 is generally a higher heat and No.2 is a wavy type of profile that will get what I want through specific maneuvers during the roast.

Speed and duration are dependent on the bean and the flavors I am interested in.

I love your approach. I suppose that, ultimately, what we want as roasters is the ability to master all the profile pathing so that we can pull out the flavors we want, however we want. To becoming masters at our craft!

Yes, exactly

treq10 wrote: Charge: 220C BT, Low airflow
Yellow: 153C @ 5:45, medium airflow
Increase Airflow to max @ 190C
FC: 200C @ 9:30
Drop: 208-210C @ 10:15-25 (45-55s post 1C Dev)

What is your charge weight? Charge temps are relatively useless without a corresponding weight.

Yellow @ 5:45 and FC @ 9:30 giving you the fruit tart result further reinforces my hypothesis that the drying:maillard time ratio has a very large effect on exactly what sugars get developed during the development phase. I too generally try to end up with high drying:maillaird ratios when going for fruity, clean, bright SO's. I've found the only foolproof way to do this consistently is the "soak" method, which it looks like you're imitating as well. It would be very interesting to nail down exactly what element(s) of these sorts of profiles gives this result, as it seems to give very similar results even across roaster manufacturers and roaster types.

treq10 wrote: I'd assume that their approach is closer to:

4-4:30 Yellow
7:30-8:00 FC
~2min development

Do you have a basis for that assumption? Not trying to be argumentative, it's a serious question. Many have tried to duplicate Howell's results and few have succeeded... I'd imagine if it was that easy we'd see many more Howell copy cats.

treq10 wrote: So my question is: Do you roast fast with long development (like the 49th Parallel and George Howell coffees)? Or do you prefer to minimize caramelization and highlight acidity/sweetness?

The double edged sword of hobbyist roasting is that very few of us will every fully realize a single coffee's full flavor range, and thus unlock a coffee's true potential. In fact very few roasters do this in general, however most professionals (and some hobbyists) have some inkling of where to take a coffee after a couple test roasts. The problem is that these general heuristics are shaped and molded by the prevailing trends and tastes that people have already been exposed to. Some of my most interesting and delicious roasts came from profiles that I initially thought were failures. But i've also had profiles that looked bad and tasted even worse. Humans in general are predisposed to limiting downside instead of increasing upside, so more often than not you see roasters fall back on well known and proven profiles/techniques when trying to nail down an unruly or exotic bean or to get a specific flavor. It makes sense... nobody wants to waste beans on crazy profiles that have a low probability of succeeding.

I guess this is just a really wordy way of saying that sounds like a false dichotomy to me. I think we've had it beaten into us that those are our two choices, and we fall back onto those two modes of roasting because they're safe and known. But why can't we have a SO cup with caramel and nuts as the base/finishing note along with a bright and sweet top note? I generally refuse to believe this isn't possible, partly because I believe I've had coffees like this before and partly because believing we are doomed to forever choose between third wave brightness vs more rounded comfort flavors would make me just give up and get a Tim Wendelboe monthly subscription. And if anyone will figure it out I don't think it's outlandish to think it will be a hobbyist roaster who isn't constrained by the tyranny of customer desires.

I firmly believe that as hobbyists we should roast for a specific brew method (pourover/espresso/drip/etc) and try to take a coffee to it's ultimate expression for that brew method, desires and predisposed notions be damned. Far too often I find myself trying to duplicate past results or forcing a specific outcome and it rarely leaves me fully satisfied (Oh this is a high quality Central American Geisha, it should taste floral and bright. Let me just slam it with heat and roast it fast with little development -> Why does this cup not taste like how I imagined it would?!?!). It sounds like typical pretentious coffee snob bullsh**, but letting the coffee take you where it, not you, wants to go never leaves me disappointed and always teaches me something.

As a reliable generalization, I find fast-start, fast-finish and slow-start, slow-finish works well for me. Post 1C development and pre-dry times need to be proportional.

In other words I am rarely happy with a 6:00 dry, stretched Maillard and 1:00 post 1C drop. Likewise, I've never had a good result from a 3:30 dry, 2:00 Maillard and 4 minute post 1C development.

I did a 6:30 total time (3:30/2:00/1:00, 295DE, 370*1C, 390*F) Ethiopia Idido Nat roast the other day that was just bangin'. I say was, because I brewed all of it at the bar in one day. It was bright and juicy, acidic without a hint of citrus, fully developed and sweet as could be. 30s longer would have killed it. Heck, 10s might have as well.

Can we please give FC start temp and drop temp when discussing anything roasting?

Time by itself during development is irrelevant to me.

One single temp by itself during development is irrelevant to me.

I may finish in 1 minute and only cover 5 degrees of development. I could also finish in 1 minute and cover 15 degrees of development - massive difference.

Previous post edited for DE and 1C temp.

Very interesting - towards the tail end of my time with my Huky, I was trying to do really fast roasts like that with big time energy in dev and had similarly good results. I distinctly remember a roast that filled the air with fruit, it was amazing. I almost started theorizing that all that really mattered was good energy during dev for those light roasts.

Oddly, I never got back to that style. Maybe cause I'm overly influenced by looking at royal's crown jewel analysis where their roasts completely tank at dev, with little increase in temp.