Newbie roaster trying to understand my goals in roasting

Discuss roast levels and profiles for espresso, equipment for roasting coffee.
walr00s
Supporter ♡

#1: Post by walr00s »

I'm trying to produce sweeter, fruitier coffee, using natural process beans from well-regarded distributors. I know Tim Wendelboe has emphasized the importance of using shade-dried beans (30+ days of drying), but I've had a hard time finding beans that are specifically advertised that way. I've read quite a bit at this point on the subject of coffee roasting. The primary texts I've studied are Rob Hoos' Modulating The Flavor Profile of Coffee and to a lesser extent Scott Rao's Coffee Roaster's Companion. I've also seen some very informative posts here from Jim and DaveC that have sort of muddied the waters for me while calling into question some of the things I thought I knew. I've also watched a video from John Laird at the 2013 Nordic Roast conference and read many of the (free) articles on Willem Boot's website, and read or watched most anything I can find from Tim Wendelboe. The only things I feel very confident about in my knowledge of roasting are that I shouldn't be attempting to copy profiles as there is no single correct profile and that I must obey the laws of thermodynamics :P . Other than that, I feel like I'm in over my head when it comes to putting the things I've read together.

Rob Hoos seems fairly certain that the "drying phase" involves little to no chemical reactions, which implied to me that it's a good time to make sure your roast is as even as can be with high airflow and (for my roaster) lower heat application...So first question: should I be thinking about a "drying phase" at all? I can't tell if it's an oversimplification of what's happening in the bean to the point that it isn't useful. It also seems that on my roaster, this approach directly contradicts one of Rao's commandments (RoR should always decline). I have to increase heat and reduce airflow later in my roast to rapidly get to 1C and inevitably this means my RoR spikes somewhere in the 3-5 minute mark. Second question: is my approach fundamentally wrong? By applying lower heat and higher airflow early am I impeding my ability to get the flavors I want out of my roast?

I've read and heard several statements from people I respect both on these forums and in other media that state that I should be rushing to 1C for Nordic style roasts and that rushing to 1C is one of the keys to preserving sugars and sweetness in the final product (along with lowering heat during development time to reduce caramelization?) . I guess first off -- have I understood this correctly or am I oversimplifying what I was told? Secondly, I've found that on my roaster, maxing out heat and lowering airflow gets me to 1C the fastest, however it sometimes seems to result in a longer 1C, which I assume means the roast is less even? I've been assuming this entire time that we want the most even roast we can get, but I suppose I haven't expressly read that anywhere...

User avatar
happycat

#2: Post by happycat »

Welcome to roasting fun.

When I adopted Rao, my coffee got reliably sweet and balanced.

Choose one bean for awhile.

Choose one roast level for awhile.

Choose one method. Master it over a bunch of roasts. Tweak it (adjusting proportions of phases) to taste the impact.

If you jump around you won't learn much because there will be too much variation. The specific choices you make above are less important than committing to them for awhile. This will allow you to learn principles/heuristics for what changes when you tweak things.

Have fun with it... focus on the experience and not on trying to avoid error. Track your profiles. Taste with different grinds and temperatures based on the roast level you chose.
LMWDP #603

bicktrav

#3: Post by bicktrav »

walr00s wrote:I'm trying to produce sweeter, fruitier coffee, using natural process beans from well-regarded distributors. I know Tim Wendelboe has emphasized the importance of using shade-dried beans (30+ days of drying), but I've had a hard time finding beans that are specifically advertised that way. I've read quite a bit at this point on the subject of coffee roasting. The primary texts I've studied are Rob Hoos' Modulating The Flavor Profile of Coffee and to a lesser extent Scott Rao's Coffee Roaster's Companion. I've also seen some very informative posts here from Jim and DaveC that have sort of muddied the waters for me while calling into question some of the things I thought I knew. I've also watched a video from John Laird at the 2013 Nordic Roast conference and read many of the (free) articles on Willem Boot's website, and read or watched most anything I can find from Tim Wendelboe. The only things I feel very confident about in my knowledge of roasting are that I shouldn't be attempting to copy profiles as there is no single correct profile and that I must obey the laws of thermodynamics :P . Other than that, I feel like I'm in over my head when it comes to putting the things I've read together.

Rob Hoos seems fairly certain that the "drying phase" involves little to no chemical reactions, which implied to me that it's a good time to make sure your roast is as even as can be with high airflow and (for my roaster) lower heat application...So first question: should I be thinking about a "drying phase" at all? I can't tell if it's an oversimplification of what's happening in the bean to the point that it isn't useful. It also seems that on my roaster, this approach directly contradicts one of Rao's commandments (RoR should always decline). I have to increase heat and reduce airflow later in my roast to rapidly get to 1C and inevitably this means my RoR spikes somewhere in the 3-5 minute mark. Second question: is my approach fundamentally wrong? By applying lower heat and higher airflow early am I impeding my ability to get the flavors I want out of my roast?

I've read and heard several statements from people I respect both on these forums and in other media that state that I should be rushing to 1C for Nordic style roasts and that rushing to 1C is one of the keys to preserving sugars and sweetness in the final product (along with lowering heat during development time to reduce caramelization?) . I guess first off -- have I understood this correctly or am I oversimplifying what I was told? Secondly, I've found that on my roaster, maxing out heat and lowering airflow gets me to 1C the fastest, however it sometimes seems to result in a longer 1C, which I assume means the roast is less even? I've been assuming this entire time that we want the most even roast we can get, but I suppose I haven't expressly read that anywhere...
These are great questions. Forgive me if I missed this in your post, but what roaster are you using?

On the subject of Rao vs. Hoos, I strongly favor Rao. This is not a knock on Hoos, who is a wonderful resource; it's more a compliment to Rao, whose expertise astounds me. People mostly talk about his declining RoR advice (which I subscribe to), but his advice on warm-ups and between batch protocols is arguably more valuable. Using Rao's methods, I can trace curves to within 3 seconds on all marks: dry end, first crack, drop (see pic). Before I used his methods, I'd be lucky to get within 30 seconds of my background total roast time. That kind of consistency is crucial; if you can't trace curves, you can't alter variables in isolation, which makes it very difficult to determine cause and effect.

Apart from Rao's consistency guidance, I buy into his theory on avoiding flicks/crashes. I've performed blind cuppings of roasts with crashed RoRs vs smooth RoRs, and I've been able to tell the difference. Crashed coffees are not as sweet; the acidity it less pleasant. That said, the difference is small. It's not the kind of thing the average drinker would notice, so I get why people think he spends undue time on it. Scott has also backed away some from his earlier, more draconian stance on crashes; now, unless the crash is severe, he doesn't (necessarily) demand that you fix it. Likewise, Rao has moved away from his strict advice on DTR. Today, he'll basically say to ignore DTR and focus on curve shape. I agree with the evolution in his thinking. Roasts certainly don't all need to be in the 20-25% range the way CRC suggests. I've had wonderful coffee dropped at 12%, 15%. 20% and so on.

Rob Hoos has some great information in his book, but without data -- the book contains no curves -- I don't find it as persuasive. That said, he circles around an idea I mostly agree with: shortening/lengthening roast times regulates brightness/sharpness (I'll quibble with the way he talks about each phase in isolation, though). I don't agree with his theory that drop temp correlates with sweetness - don't agree with it at all, actually. In my experience, high drop temps equate to roasty notes, but the flavor impact of low drop temps is far less scrutable, particularly when you consider that low drop temps rarely occur in isolation; they often happen alongside underdevelopment or stalled RoRs, which makes determining causality almost impossible.

Both Rao and Hoos instruct that there is little-to-no flavor development happening during the dry phase, and I agree with that. It's more about establishing momentum.

Regarding your airflow/heat question, usually with drum roasting you have either steady airflow throughout your roast (a set-it-and-forget-it approach), or you use less air initially, then increase airflow. You almost invariably start with more heat and drop your gas levels as the roast progresses (I'm tempted to strike the word "almost" from this sentence). It sounds like you're doing the opposite: more air/less heat initially, then less air/more heat at the end of the roast. If that's the case, you should reverse your approach. Your RoR should look like a flipped/reversed Nike symbol: a spike initially, then a slow, steady descent until you drop the batch.

Nordic roasts are their own thing. I won't weigh on them much, but I don't agree that rushing to first crack preserves sweetness. In my experience, that approach leads to very bright coffees, not necessarily very sweet coffees (although it doesn't prevent the coffees from being sweet, either). I agree with the idea that you want less heat during development to avoid excess caramelization; that's doubly true for natural processed coffees, which rocket in temp during development if you use too much heat.

I don't know enough about Wendelboe's advice to comment -- but I enjoy his Aeropress method!

Hope all of this helps! Let me know if you need clarity on any of the above!

★ Helpful

User avatar
Chert

#4: Post by Chert »

bicktrav wrote:These are great questions. Forgive me if I missed this in your post, but what roaster are you using?
]
OP profile indicates Sandbox Smart Roast R1. I see nothing substantial about it on HB and I can't open the google-searched website.
LMWDP #198

Nunas
Supporter ♡

#5: Post by Nunas »

Try here: https://www.bellabarista.co.uk/sandbox- ... -deal.html For some reason, I could not find it either, even by drilling down via the menus. But I found this Google hit works.

User avatar
Peppersass
Supporter ❤

#6: Post by Peppersass »

I'm not an expert roaster, and what I've learned on my Quest M3 (M3s drum) and Ikawa Pro V3 sample roaster won't fully translate to your roaster, here are some quick thoughts:
  • First and foremost, the factor that most determines whether you get to your roasting goals is the quality of the greens. Just because the greens supplier assigns tasting notes like "berry", "citrus", "sweet", etc. to their Ethiopian natural doesn't necessarily mean you're going to get those notes or even a great tasting cup. This doesn't mean that spending a lot of money on greens guarantees success, but my experience has been that the best greens usually come at a higher price.
  • A good way to assess greens quality is to buy them from a commercial roaster who sells both green and roasted beans for the same coffee. Recently, I bought a bag of roasted Ethiopia Guji Kayon Mountain Natural along with a pound of the same greens from Klatch. The roasted coffee was an excellent medium-light roast with nice fruit notes and sweetness. That told me the greens I bought would produce the same results if I could crack the code. I used a roast profile that works pretty well on my roaster for espresso, and found that if I roasted to about the same color as Klatch I got similar results.
  • Speaking of naturals, while they often feature more fruity flavors, in my experience they're harder to roast well than washed coffees. Natural greens that are less than top-quality often have funky, astringent flavors with medium or light roast profiles. You might get some high-quality washed greens and compare results with your natural greens.
  • Speaking of color, I think it's a good idea to get a color measuring device. I used to use the original Tonino, but it was somewhat cumbersome to use. I switched to the relatively inexpensive Roast Vision color meter a while back and found that it's much quicker and easier to use, and the measurements are very reproducible. The readings may not exactly match the numbers produced by an expensive commercial color meter, but they're in the ballpark and consistent enough to give excellent relative measurements, which is all that matters to me.
  • In terms of color, I find that for most coffees I've roasted, the "sweet" spot is between medium-light and light. If I go too far towards medium, I lose fruit and sweetness. If I go too light, I get grassy, vegetal, astringent flavors.
  • Are you roasting for brew or espresso? Despite many commercial roasters no longer doing SO espresso roasts, implying that their SOs work well for any brew method, I think the roasting approach is different, especially for naturals. You can get away with a very fast roast for brew, but likely you'll need a longer roast for espresso. My sense is that most brew methods are more efficient at extraction than espresso because of the much greater water contact time. Sometimes you can get around this with flow profiling (ultra-fine grind and long, slow preinfusion), but you'll get more sweetness in espresso with a longer roast time.
  • With what appears to be more of a rotating basket than a drum, I'd say there's a lot more convection than conduction going on in your roaster. I think this is true of many small-drum roasters, especially those with perforated drums, perhaps making them a bit more like a fluid bed roaster than a large drum roaster. That might mean adjustments in airflow could be just as important as adjustments in heat. For example, most profiles for my Pro V3 gradually increase input heat throughout the roast, though the curve gradually gets less steep (causing declining RoR), and in some profiles flattens out about when 1C starts. In a typical brew profile, airflow decreases throughout the roast (also contributing to declining RoR), then increases during cool down to blow off chaff. Many drum profiles I see increase airflow during the roast, especially during 1C to blow off chaff, which I suspect in a fluid bed or small-drum roaster with a highly-perforated drum might lead to a big spike during 1C. Again, mine is a different roaster than yours, but my experience with the small, perforated-drum Quest suggests that "just leave airflow steady and increase during 1C" may not be the best approach.

User avatar
Peppersass
Supporter ❤

#7: Post by Peppersass »

Nunas wrote:Try here: https://www.bellabarista.co.uk/sandbox- ... -deal.html For some reason, I could not find it either, even by drilling down via the menus. But I found this Google hit works.
I posted before viewing the video. Some questions:

1. It sounds like you're trying to construct your own profile. Have you tried any of the profiles built into the app? What do they do with respect to heat and airflow? Are the results not acceptable? I've mostly been working with profiles available from Ikawa. At first I modified some of the profiles to shorten the roast, but found it's better to just manually drop it early based on DT (not DTR.) I'm not going to heavily edit profiles or create my own until I get more familiar with the available profiles.

2. Can you post some curves from your roasts?

3. Do you have the cooling tray? Seems to me that just taking the beans out of the roaster, as in the video, wouldn't cool the beans fast enough, which may result in roasting well past the target color and development. That could squeeze the fruit and sweetness out of a natural.

4. When are you marking 1C? After several years of roasting, and reading Roa's ideas about using ETRoR to mark 1C, I concluded that I've been marking 1C too late. I used to wait past the first few outliers until I got a more continuous set of pops. Now I mark when I hear the first pop. Since doing that my roasts have been fruitier and sweeter.

walr00s (original poster)
Supporter ♡

#8: Post by walr00s (original poster) »

bicktrav wrote:These are great questions. Forgive me if I missed this in your post, but what roaster are you using?
Really appreciate the detailed response. I am using the Sandbox Smart Roaster
bicktray wrote:I can trace curves to within 3 seconds on all marks: dry end, first crack, drop (see pic).That kind of consistency is crucial; if you can't trace curves, you can't alter variables in isolation, which makes it very difficult to determine cause and effect.
On my Sandbox, I feel like a lot of that magic is sort of handled for me. It tends to hold your hand through the controls so that reproducing roasts is fairly simple. The only major adjustment I've found myself having to account for on a regular basis is first roast vs. everything after in a session. First roasts tended to behave quite a bit differently. For that reason, I essentially have two of every profile I use, one for first roast and one for everything after.
bicktray wrote: I've performed blind cuppings of roasts with crashed RoRs vs smooth RoRs, and I've been able to tell the difference. Crashed coffees are not as sweet; the acidity it less pleasant. That said, the difference is small. It's not the kind of thing the average drinker would notice
Afraid I fall into the average drinker category here. I think my roaster or maybe my method prevent most of the crashes, and all of the flicks (only roasting 100g at a time so maybe there's not enough bean mass to produce a real flick at 1C?). I've had a couple of crashes from RoR of 3-4 to 0 or -1, and cupping them I can't really tell the difference between subsequent batches roasted with the same profile but without the crash for whatever reason. My roaster measures bean temp based on the air coming off the bean mass rather than with a probe directly in the bean mass, and I think Dave Corbey stated that there can be some false appearances of crashes due to this when he reviewed the machine last year.
bicktray wrote:Rob Hoos has some great information in his book, but without data -- the book contains no curves -- I don't find it as persuasive.

Interesting, I feel basically the inverse. Hoos gave cupping data as a basis for every trend/recommendation he provided across multiple varieties and process types, whereas Rao seemed to give out a lot of advice based on his experience alone with no empirical data. I respect that they both have probably forgotten more about roasting than I will ever know.
bicktray wrote:That said, he circles around an idea I mostly agree with: shortening/lengthening roast times regulates brightness/sharpness (I'll quibble with the way he talks about each phase in isolation, though). I don't agree with his theory that drop temp correlates with sweetness - don't agree with it at all, actually. In my experience, high drop temps equate to roasty notes, but the flavor impact of low drop temps is far less scrutable, particularly when you consider that low drop temps rarely occur in isolation; they often happen alongside underdevelopment or stalled RoRs, which makes determining causality almost impossible.
In my studies of these texts, I failed to fully recognize the importance of this stated relationship. So if I'm getting roasty notes that I don't care for, I should start by attempting to lower drop temps? But if the drop temps are lower, and there's other flavor problems it may be addressed due to stalled RoR or...? Sorry think I'm still failing to understand the full implications of this relationship between drop temp and flavor.
bicktray wrote: Regarding your airflow/heat question, usually with drum roasting you have either steady airflow throughout your roast (a set-it-and-forget-it approach), or you use less air initially, then increase airflow. You almost invariably start with more heat and drop your gas levels as the roast progresses (I'm tempted to strike the word "almost" from this sentence). It sounds like you're doing the opposite: more air/less heat initially, then less air/more heat at the end of the roast. If that's the case, you should reverse your approach. Your RoR should look like a flipped/reversed Nike symbol: a spike initially, then a slow, steady descent until you drop the batch.
I'm starting at lower heat (70% power) and high airflow (100% fan), then during what Hoos calls the Maillard phase (~155C measured bean temp almost all the way to 1C), I raise to 100% power and lower fan to ~70% (been experimenting with this). I'm then ending (development phase) with very low power (40-55% depending on bean) and max fan again. As I said, if I start with high power and low airflow and continue until I get fairly close to 200C, 1C ends up very spread out. I'm guessing that means my bean mass has a higher variance in internal temperature? That seems bad to me but I suppose it isn't necessarily? I've experimented with a gradual ramp in airflow and high heat the entire time, and it does seem to reduce the duration of 1C, but it also seems to greatly slow (60s+) how quickly I get to 1C.

Here's a picture of one of my profiles, not sure how useful it is in a vacuum, was for an Ethiopian Natural grown above 1500m:


bicktray wrote: Nordic roasts are their own thing. I won't weigh on them much, but I don't agree that rushing to first crack preserves sweetness. In my experience, that approach leads to very bright coffees, not necessarily very sweet coffees (although it doesn't prevent the coffees from being sweet, either). I agree with the idea that you want less heat during development to avoid excess caramelization; that's doubly true for natural processed coffees, which rocket in temp during development if you use too much heat.
That's very helpful. It sounds like maybe I've been over-focusing on getting to 1C quickly and I need to focus more on heat applied during development time or just development time as a whole.

Again, thank you for the detailed response, really appreciate it.

walr00s (original poster)
Supporter ♡

#9: Post by walr00s (original poster) »

Peppersass wrote:I posted before viewing the video. Some questions:

1. It sounds like you're trying to construct your own profile. Have you tried any of the profiles built into the app? What do they do with respect to heat and airflow? Are the results not acceptable? I've mostly been working with profiles available from Ikawa. At first I modified some of the profiles to shorten the roast, but found it's better to just manually drop it early based on DT (not DTR.) I'm not going to heavily edit profiles or create my own until I get more familiar with the available profiles.

2. Can you post some curves from your roasts?

3. Do you have the cooling tray? Seems to me that just taking the beans out of the roaster, as in the video, wouldn't cool the beans fast enough, which may result in roasting well past the target color and development. That could squeeze the fruit and sweetness out of a natural.

4. When are you marking 1C? After several years of roasting, and reading Roa's ideas about using ETRoR to mark 1C, I concluded that I've been marking 1C too late. I used to wait past the first few outliers until I got a more continuous set of pops. Now I mark when I hear the first pop. Since doing that my roasts have been fruitier and sweeter.
Really appreciate you taking the time to give me some advice. I've read your other post and will attempt to make adjustments to how I approach naturals accordingly. I'm going to try buying greens from a local roaster who has a Costa Rican Natural that I am really enjoying right now and attempt to replicate their roast. I believe they buy all of their greens from Royal. I have bought mine from Sweet Maria's, Coffee Bean Coral, and one order from The Captains' Coffee.

1. I tried both the light and medium profiles provided within the basic app. The light profile came out so-so, but I found that making some adjustments to heat and airflow in each stage of the roast has improved my results. The medium resulted in what I'm guessing is "baked" beans where the taste was flat, roasty, and generic.

2. Sure. I don't want to spam. Most of my profiles end up looking very similar. I haven't been playing with Development Time as much as drying phase and the approach to 1C. The ones I've let go for more than ~60s past 1C have ended up being way too dark for my tastes, but it sounds like I've maybe been applying too much heat during DT for those roasts. Here's a few:




3.I do have the cooling tray. I roast in my garage, where it's usually 90-105F right now, but the tray does get the beans down to ambient temp in under 60s.

4. Could you link Roa's ideas about using ETRoR? Is it in The Coffee Roaster's Companion? I know Wendelboe essentially uses ET and BT to keep his roasts consistent, since he can't hear 1C in the Loring (or maybe that was for the Probat...). I actually don't really mark 1C at all, but instead manually adjust the heat setting when I hear the second set of 2 consecutive-pops within ~1s. So you're marking 1C right at the first pop? Does how consistently the popping sound occurs play a role at all? I've been trying to even out my 1C to all occur within the same ~30s interval, rather than spread over 60s+, but it seems like maybe I'm overvaluing this and need to consider letting it go in favor of trying other things.

User avatar
mkane
Supporter ♡

#10: Post by mkane »

Roast and roast some more! Figure out how your roaster act's when you twist the dials. Until then don't
stress over taste because lot's of it won't taste very good.
I too subscribe to a declining RoR. I also think too much airflow in the beginning wastes lot's of heat and strips flavor.

Were 1300 roasts in on our BC-1 and still haven't figured it out. At some point in time we may just try almonds as there everywhere around here.