9SB, did I miss cupping notes?

Obviously the added airflow has slowed the whole roast down slightly, and it made the decline of ROR during the beginning of first crack more moderate ...

Mark, what software are you using? I can't track your ROR. You mentioned about one "cratering" but I don't follow.

Sorry, I use the Roastmaster software, and although it is easy to see the ROR curve on the iPad, it is somewhat faint on the screenshots that I can upload here. It is the fainter red line towards the bottom of the graphs.

By "crater" I meant decline precipitously, at a much faster rate than was true before first cracks began. The crater itself doesn't seem to be a huge problem, as the 25% roast had the steepest decline but still tasted pretty good. It was the fast decline followed by the upward flick in the second roast that I blame for the enhanced astringency I noted in the aftertaste.

Thanks for sharing the profiles, 9Sbeans! I look forward to hearing how they taste.

[creative nickname] wrote:Sorry, I use the Roastmaster software, and although it is easy to see the ROR curve on the iPad, it is somewhat faint on the screenshots that I can upload here. It is the fainter red line towards the bottom of the graphs.

...

Now I see the ROR with your hint.

Boldjava wrote:9SB, did I miss cupping notes?

Dave, I only cupped once 36-hrs post roast, and they were virtually identical. As I noted previously, my batch #1 (with increasing airflow during development phase) stands out at the side-by-side test due to its more pronounced pleasant acidity. I couldn't perceive any difference when it's cool. I will follow up the aging every other day, but doubt myself in the non-blinded conditions to tell any significant difference. That say, if I can only drink one cup per day, I may not notice any difference. Actually I may even prefer my batch #2 (with decreased-then-increasing airflow during development phase, but minor flick at the end) because it's a more balanced cup. My batch #1 stands out just because its sharper acidity in the side-by-side comparison.

The differences of the two batches are merely 7-seconds apart and the drop temperature diverge only 2.5F if subtracting the different 1C-start temperature. Obviously the difference in airflow has direct impact on the measured BT readouts (i.e. different measurement errors), and so far I can only say it's a good roasting exercise. I couldn't attribute the minor taste difference to a higher drop temperature or to the dip/flick, yet.

I'm aware that you (and Joe M.) take the post-1C dip & flick as natural responses of bean as long as the flick rate doesn't exceed the extension of the general declining RoR trend (as in my batch #2). But Scott Rao would point to the flick saying a flick is a flick and a rising RoR. What I'm practicing is to combine gas/airflow manipulations to minimize the dip & flick. Before I could reliably achieve that goal (minimizing dip & flick) and taste the difference, I won't dismiss any of your idea.

12/1 updated:

Evaluated on day 3 & 5 post-roast, the results were basically consistent with my previous observation. On day 3, using fingertip randomly crushed several beans from each batch, picked more fruity notes in batch #1, and the fragrance lasted longer. However, this difference in dry fragrance diminished on day 5 when bulk beans were grinded and evaluated again prior to brew. I suspect the volatile molecule had escaped during the degassing period.

I chose three beans in similar size from each batch and cut them open. Interestingly, the batch #1 (with high airflow) has very even roast level in both inside and outside. OTOH, the batch #2 (reduced then increased airflow) showed a slightly darker ring in two of the three beans. I started wondering the "baked" taste caused by the flick as Marshall & others brought up months ago.

Well, here we go.

I tried two roasts of this Kenya tonight, in the spirit of playing with air flow.

Goal of roast 1: modulate maillard and first crack phases using mostly the flame

Goal of roast 2: modulate maillard and first crack phases using mostly the fan

Overarching goals: identical drying phases, identical drop temps, identical development ratios >20%, drawn out first crack as in Joe's roast. However, holding all those things constant means something has to give, and that something is time. I let time in maillard and first crack phases flex - given how my Huky works, it would have been very very difficult to achieve similar timing with drastically different heat and fan settings. I don't have that level of technical prowess.

BE FOREWARNED: At least for me, this is how the Huky's perforated drum works - you don't get the super high peak of RoR in drying, but rather a broad flat mesa. Not sure if anyone else experiences this.

Also, the Huky with perforated drum has a shutter that, when opened, exposes the beans directly to the flame, and when closed, breaks the line-of-sight between the flame and beans. As a rule, I run my drying phase with the shutter open (for greater radiant heat), and close the shutter after yellowing (for greater convection).

Roast 1 - Modulate Development using Heat:
Huky 500, perforated drum
Charge: 227g
Drop: 196g (13.7% weight loss)



Anyway pretty good execution I thought. Fan constant after yellowing, at about 50% of max capacity (or 45V in Huky language). Modulated the flame to level out the ET at target and create a long crack.

Roast 2 - Modulate Development using Air Flow:
Huky 500, perforated drum
Charge: 227g
Drop: 196g (13.7% weight loss)



This one was tougher. In order to support higher fan settings (100%, or 60V), I had to leave the flame at a higher gas pressure through the start of Maillard. But at this high gas pressure, when the fan is first turned on, it draws more hot air into the chamber and actually drives the ET up. So in order to compensate later on, I dropped the gas pressure nearly all the way down, and tried to drive down ET and surf it down to my target. I may have baked this one, we'll see. Definitely more of a stumble on my part here.

But, it looks like my drying phase was near identical, I achieved similar weight loss, similar drop temps, and even similar development length.

Cupping notes @ 11 hours post roast:

Roast 1 - beans looked pretty even, not much going on in the dry grounds, hot cup had pronounced sharp acidity and vegetal taste, cooled cup softened considerably, lacked complexity but had a bit of stone fruit and perhaps a slight roasty note (from the dip and flick?)

Roast 2 - more pronounced puffing/bubbling in the beans though they looked even, better expansion generally, dry grounds smelled of berries, hot cup well integrated on mid-pallate with berries and acidity, cooled cup was still well integrated and layered flavor, but acidity became more pronounced mid-palate, on the edge of too acidic, but a great Kenyan cup. No roastiness, shockingly not baked despite the near-flatline BT.

Conclusion at this point:

More air later in the roast = better integration of flavors, better development. Roast 2 could probably be improved by using the same general profile, similar time in development, with a slightly more condensed first crack and a higher drop temp to tone down the acidity.

This was a little delayed because I've been so busy with work, but here are the notes from my second comparison. Once again, the goal was to keep everything as similar as possible, with airflow being the only difference. I got closer this time, running two roasts within 13 seconds of overall roast time, finishing within 1 degree F of each other, and with no serious differences in the overall profile. I also managed to avoid any serious flicking in the lower airflow roast (although there was the briefest little blip), and although it was a bit dryer than the higher airflow roast, it didn't have the astringent finish that I associate with rising ROR during development. In a blind comparison between the two roasts, the higher airflow cup had a bit more sweetness at the forefront of the higher airflow roast, a bit more nuance in the flavor, and a slightly richer aromatic profile, along with fewer tannins showing up in the finish. Detailed notes follow below.

---

Roasting Info: 50% Airflow Roast:

Bean: Kenya Gichugu
Roaster: USRC Sample Roaster
Charge Mass: 333g
Charge Temp: 390F
Dry/Ramp/Development: 3:15/3/2:11
FC-start temp: 370F
Finish Temp: 388F
Overall Roast Time: 8:24
Moisture Loss: 14.11%

Profile Plot:



[BT=Red (ROR = Fainter Red Line), Exhaust=Green, Gas=Blue, Fan=Yellow]

Tastify Scoresheet:



---

Roasting Info: 75% Airflow Roast:

Bean: Kenya Gichugu
Roaster: USRC Sample Roaster
Charge Mass: 333g
Charge Temp: 390F
Dry/Ramp/Development: 3:15/3:08/2:14
FC-start temp: 371F
Finish Temp: 388F
Overall Roast Time: 8:37
Moisture Loss: 14.04%

Profile Plot:



[BT=Red (ROR = Fainter Red Line), Exhaust=Green, Gas=Blue, Fan=Yellow]

Tastify Scoresheet:

So if I had to summarize our learnings so far:

- Mark: higher airflow throughout the roast gave the best cup, all else being equal, and even when the dip-and-flick has been eliminated
- Patrick: higher airflow during development gave a brighter and more flavorful cup, but not super different from the high-low-high air application
- Jeff: higher airflow through development gave a more balanced and flavorful cup
- All: the dip-and-flick behavior is harder to control with lower airflow
- All: lower airflow application gave decent cups

So at this point: more air = generally better for bringing out more pronounced flavor/acidity, particularly when applied in development phase. However, the use of lower airflow can also provide for a good, though perhaps simpler cup, with the caveat that you must control dip-and-flick behavior.

Sounds like a reasonable summary to me. I should probably do a test where I max out airflow to see if there is a point past which more airflow becomes detrimental. I have heard a few pros claim that too-high airflow strips aromatic notes out of coffee. I think I'll try that this weekend before I move on to exploring the impact of constant versus varying airflow settings.