Would be a lot more helpful if your graphs were actually readable. The images are too small.

Thanks very much for sharing these profiles and results! Here were some comparisons I did, in which slow-start, fast-finish fared better than the alternative approaches I tried:

Roast and Learn Together - August 2014

Roast and Learn Together - August 2014

Obviously, what works best may vary based on both on the design of a particular roaster and the characteristics of a particular coffee, so I'm not sure how easy it will be to support broader generalizations based on these kinds of experiments. Rao suggests that his recommendations are based on a large pool of comparisons, but much of that data seems to come from large commercial roasters. At least in my own experience, some light roasts can taste better using a SS/FF approach, and with less development than he recommends.

What coffee were you roasting when you did the comparisons you posted above? My experiments were done with the Kenya Kirinyaga Kiangoi AA that we used for the August R&LT.

EvanOz85 wrote:Would be a lot more helpful if your graphs were actually readable. The images are too small.

Apologies. I have to manually adjust them in photoshop to get 800 px images for hb. Here's an imgur link:

[creative nickname] wrote:Thanks very much for sharing these profiles and results! Here were some comparisons I did, in which slow-start, fast-finish fared better than the alternative approaches I tried:

Roast and Learn Together - August 2014

Roast and Learn Together - August 2014

Obviously, what works best may vary based on both on the design of a particular roaster and the characteristics of a particular coffee, so I'm not sure how easy it will be to support broader generalizations based on these kinds of experiments. Rao suggests that his recommendations are based on a large pool of comparisons, but much of that data seems to come from large commercial roasters. At least in my own experience, some light roasts can taste better using a SS/FF approach, and with less development than he recommends.

What coffee were you roasting when you did the comparisons you posted above? My experiments were done with the Kenya Kirinyaga Kiangoi AA that we used for the August R&LT.

Hi Mark. Thanks for your feedback. I definitely referred to your posts (plus those of Tom C. and Rich) when figuring out roast ideas.

I think it's important that you point out the distinction between "drying" and ramp. I have always tried to have a drying/ramp ratio of >1 in line with your post's suggestion.

As far as declining vs. linear RoR, I'm not really sure what to say. To some extent, my profiles aren't all that far off from linear - they maintain a similar RoR from drying through ramp but definitely flatten a bit in 1C. So I'm not convinced that decreasing RoR from drying to ramp is necessary - even though it'll tend to happen since really fast ramp times might cause scorching. But I do tend to think that a fast, aggressive drying phase produces better internal bean development - as Rao suggested. In your first post, your drying is quite fast! So that would suggest that fast-start, slow-finish profiles can produce good results. But obviously, I'm new to this and always open to trying out different profiles. I'm just providing some data for others' thoughts rather than trying to pronounce one style right or wrong.

I'm interested in attempting something similar to your final profile. Maybe I'll try it on another Kenyan coffee - I have one more batch of the Tambaya AA left. I'd be interested in some info on your linear profile - how much heat in dry, and then your 1Cs temp and finish temp.

And, to answer your question, all three roasts were washed, high-grown, Latin American coffees from Sweet Maria's. The first roast used their Rosma Huehue Guatemala, the second their Huila Colombia, and the third their Costanza y Lopez Guatemala. When experimenting on roast details, I generally stick to high-grown, washed Latin American coffees since I want to avoid the "wilder" characteristics of African coffees coloring my judgments - even though I tend to prefer them.

By the way, I also used a very similar profile on two African coffees that day - SM's Kifle washed Sidama, and their Tambaya AA Kenya. Both were also very tasty - and reminiscent of what I've gotten from Heart. When I brewed the roasts for my coworkers, they nearly unanimously thought them to be the best roasts they've had from me. Now that is very possibly because I haven't executed a ss-ff profile properly! But I think it vouches for the results of this approach.

I had a meeting with TW last weekend on a project we are starting. Brought home a bag of TW Ethiopian espresso roast beans. The light roast was well done. We didn't talk profiles but maybe in the future I can get some thoughts from Tim to share here. Can add that some taming is done with their espresso roasts depending on how extreme the acidity is in the greens being used.

NoStream wrote:I'm interested in attempting something similar to your final profile. Maybe I'll try it on another Kenyan coffee - I have one more batch of the Tambaya AA left. I'd be interested in some info on your linear profile - how much heat in dry, and then your 1Cs temp and finish temp.

You can read the full rundown of that roast here: Roast and Learn Together - August 2014

Basically, I charged cooler than I usually would, then slowly increased the heat while keeping an eye on the BT-ROR, trying to keep things moving along at a steady pace at around 30F/minute until FC. Then I dialed things back, letting the endothermia slow down the ROR and then stopping when the harsher edges of the aromas went away and I started to smell more sweetness, at about 1:15 into first cracks. FC was at 394, and I stopped the roast at 408F.

This was on a HG/BM roaster with a single dimmer wired into the heat-gun, so air and heat were raised and lowered in proportion to each other. To execute this kind of profile on that roaster, I needed to keep the charge fairly small, so that I could keep it moving fast enough without getting ET so high that I scorched the beans.

I've executed similar curves on my recently acquired USRC drum roaster, but on that design I tend to keep airflow fairly low and constant until just before FC, using varying gas settings to dial in the right ROR. The drum has so much thermal mass that it is much easier to get things moving quickly without a high ET in that design; I use higher airflow mainly when I need to put on the brakes by bleeding heat quickly out of the drum, or towards the end of endothermia during first crack, to blow out the chaff and keep things from speeding up too much.

Once again, it is hard to know how much to generalize these sorts of results. I have been too busy to do a careful side-by-side of a SS/FF versus a more traditional profile of the same coffee in the USRC. When I do, I'll let you know what I find.

farmroast wrote:I had a meeting with TW last weekend on a project we are starting. Brought home a bag of TW Ethiopian espresso roast beans. The light roast was well done. We didn't talk profiles but maybe in the future I can get some thoughts from Tim to share here. Can add that some taming is done with their espresso roasts depending on how extreme the acidity is in the greens being used.

I'd be very interested to see how Tim is profiling his coffees these days. I bought a small bag from him about a year ago, and as I recall they were among the lightest I had ever seen that (when brewed properly) avoided raw or grassy flavors.

So I went ahead and roasted the last of my Tambaya AA attempting to match your profile. I charged at 400 F MET, decent amount of heat through dry, pushed quickly through ramp - though not quite enough to match your profile, and finished quickly (though a couple degrees hotter than intended).



(12.3% weight loss - super light)

And the taste - evaluated as a 1:18, 203 F Wave 155, 24 hours post-roast:

I get stewed tomato and orange from the aroma - nice and tropical. In the mouth, it's lovely, super delicate, clean, sweet, and full of cherry, tangerine, and pineapple, with no bitterness or astringency. It's not underdeveloped. The sugars are more delicate and fruited and hit closer to the front of my palate. They're less syrupy and caramelized. I don't consider this profile better or worse. I suspect that for simpler coffees, a fs-sf profile might bring out more interesting caramelized sugar notes, but for a really top notch, complex, powerful coffee like this, the slow-start, fast-finish approach works.

I'm also wondering how much the gentle charge affects the roast. When I last tried a ss-ff approach, I charged hotter and couldn't give the roast much heat at first. With a cool drop, it was easier to finesse the roast closer to a straight-line. And I suspect there might be some issue with reactions stalling out due to insufficient heat application with an overly hot charge for a slow-start roast - but that's pure speculation.

I'll keep experimenting with this profile and let you know what happens. It's almost frustrating how so many different approaches can yield tasty results.

NoStream wrote: And the taste - evaluated as a 1:18, 203 F Wave 155, 24 hours post-roast:

I get stewed tomato and orange from the aroma - nice and tropical. In the mouth, it's lovely, super delicate, clean, sweet, and full of cherry, tangerine, and pineapple, with no bitterness or astringency. It's not underdeveloped.

I've been roasting this and a few other Kenyans very similarly with success. I've found the tomato/vegetal character drops off after about three or four days rest, revealing more fruit character.

NoStream wrote:And the taste - evaluated as a 1:18, 203 F Wave 155, 24 hours post-roast:

I get stewed tomato and orange from the aroma - nice and tropical. In the mouth, it's lovely, super delicate, clean, sweet, and full of cherry, tangerine, and pineapple, with no bitterness or astringency. It's not underdeveloped. The sugars are more delicate and fruited and hit closer to the front of my palate. They're less syrupy and caramelized. I don't consider this profile better or worse. I suspect that for simpler coffees, a fs-sf profile might bring out more interesting caramelized sugar notes, but for a really top notch, complex, powerful coffee like this, the slow-start, fast-finish approach works.

I'm so glad to hear that you got good results with this approach! To me, it comes the closest to replicating some of the best "Nordic-style" cups I've had using beans from roasters like Heart or TW. You definitely sacrifice some body and mouthfeel to get the simple-sugar sweetness and clarity in the high notes. I don't advocate this for all coffees, and sometimes I like a more traditional profile even for very good coffees, so that I can enjoy a more bodied cup with more maillard & caramel notes. But it is a nice arrow to have in the quiver when you want to emphasize acidity and sweetness over all else.

Absolutely agreed. I think your profile comes close to what I've experienced from TW, Kaffa, etc. - delicate, tea-like, fruited sweetness with zero roast character. I do think that many of the coffees I get from Heart have more caramelization of sugars - not to the point where it's a "comfort food" cup, but enough to give length to the finish and more of a semi-caramelized candied sweetness. And I attribute this to their work with Rao. (I'm sure you've read this: http://www.heartroasters.com/blogs/hear ... -r-i-p-soe.)

I think a fast-finish approach is definitely pickier about which green coffees it'll work with whereas a slow-finish will produce at least "good" results with mostly anything. And it'll please a broader range of palates while still retaining plenty of varietal character - as long as you roast (and ramp, especially) aggressively fast and give the roast plenty of air, which I think are important characteristics of both the "Rao-ultra-light" and "Nordic-ultra-light" styles.

In any case, I'm really looking forward to doing side-by-side profile roasts of your ss-ff vs. a fs-sf of the Acetenango Gesha I just ordered from Sweet Maria's.