The title of this thread is a question, but I thought by leaving the question mark off, I could stir up more discussion. I'm not asking how I can become a great roaster as much as I'm asking you how you think you can become a great roaster. It's a question of dogma, but in this case, some peoples beliefs truly are better than others...

More and more, the vast majority of my free time revolves around either roasting or thinking about roasting or studying what others have learned roasting. All things coffee envelope a living passion, more than a hobby for me and I'm sure quite a few of the rest of us as well. One thing I constantly ponder is how many paths are there to being a great roaster. More than trial and error, or practice, practice, practice, but what can we as amateurs do if we want to produce the quality roasts we cup from the pros more consistently? Is there dogma? If it were basic knowledge that could be distilled into text form, someone could name their price for a book on solid beginning thru advanced roasting methods. I imagine such a book would start out by telling you if you can't cup well, you'll never roast any better than your cupping skills.

This is mainly just a mental rant, that I'd like others to contribute their thoughts to. I often think when I jumped up from the manual gadgetry to the Behmor to the Quest, then quickly added thermometry and a laptop, that the process of increasing ones gathered information, I had initially assumed it would have made roasting excellent coffee easier or more reproducible. But the more I think about it ( and I've thought about it for a long time) my opinion is roasters like myself can end up chasing our tails and never zero in on exactly what it is that makes us improve at roasting. I've been roasting since I started home-brewing beer, 10 years now, and I'll be humble enough to admit that every time I think I've got the bigger picture in focus, the closer I look, the more blurry spots get.

I know that learning to cup better each time and being able to recognize subtleties more clearly must be a backbone of the process. But I can write notes on roasting profiles and methods for years and still end up with fuzzy and sometimes inconsistent information.

If I were to take two amateur roasters like myself and split their methods into two pools, I'd have one person divorce themselves from laptops and thermocouples and online roasting forums, buy larger quantities of one very common, accessible specialty coffee and only roast that one coffee, 25 pounds if necessary, with fastidious cupping notes, nosing their way thru the roast process, before even thinking of moving on to another bean. And the second person doing what I do, buy as much 90+ greens as I can get my hands on, read everything I can find, gather data points every 0.5 second and rarely roast more than 1-2 pounds of a given variety before moving on to try something new. I'm willing to bet in 2 years, the person who did the former would be able to out roast the later with any given coffee.

But my point to the whole thing is I'm thoroughly stuck in the later camp, not just out of fear of boredom or wanting to try various coffees nor a particular addiction to gadgetry, but some strange conviction that experiencing more roasts and roast variety will shorten the learning curve. I enjoy roasting more than any other hobby and I love doing it the way I do; maybe that's why I'm stubbornly committed to changing too drastically.

Thats my flawed dogma. What's yours?

Thank you, Tom.
I am pretty much in the same boat as you are. I spend most of my free time either in roasting, espresso brewing or in tinkering with the equipment.
I agree that you cannot become a better roaster than you are able to discern the effect of whatever you do during the roasting process. Whether this is done in formal cupping or simply in brewing is only a formality.

I strongly believe that roasting is like most trade or craft; It is learned by practice. One can learn the basics, the terminology and a bit of techniques in a book, forum, videos,... or wherever you get find the information but there are so many different variables and they all behave together in so many different ways, in so many different situations that one cannot (I believe) expect to acheive the desired outcome without experimenting.

Every bean is different from the beginning and it changes with time. Every roaster is different by design. The roasting process will be different for each roaster, even for the same basic concept or design. This is why it is nearly impossible to take information from someone else's roast and apply it to your set-up and expect the exact same result.
Of course, the more similar the set-up the more likely the exchange is possible. The Quest M3 community is taking advantage of this to great success I believe. But having a different roaster and reading some of what is posted, it is clear that some of the statements that are being made are only applicable to the very specific roaster and are not true in a general sense.

What makes it hard in establishing what is good and desirable for coffee roasting, is that what might be good for a given situation is not necessarily good for another situation. The more one is familiar, and successful, with various setup and conditions and the more likely the interpretation of any given phenomenon will be accurate and meanigful.

I am a firm believer in trying different things but in order to learn you have to be successful with most of them. If you try to roast a monsoon coffee for a few roasts and move on to another coffee without being successful with it then you may have not learned all that you could from the experience. The monsoon example is chosen here because what you think you know is good for roasting a central American coffee is mostly not applicable to the Indian variety.

Some will learn faster than others, some will become really good and some will learn slow and some will never pick it up. I might very well be in the latter group, but I keep on trying because I like doing it and I like the coffee that I produce.

I do both methods. I buy 30+ lbs of stunning coffees when they come out (Tchembe, Konga), and also buy several orders, 2-3 lbs, of beans that are new to me. As a newbie, most beans are new to me, and I have lots of beans and origins I haven't tried yet.

So on most weeks I'll roast a batch of my true and tested beans that I've nailed and mastered (meaning I like it a lot and it matches the cupping notes of the coffee) in various styles, and at the same time roast a new bean that I'm discovering. Two batches a week, a Tchembe and a new bean, or a Konga and a new bean.

What I focus on now is repeatability via PID and a routine setup. I've had beginners luck on some beans in the past, but having kept no precise notes, failed to reproduce them. Now I take advantage of Artisan's notes, and write the plan and PID parameters, drop-in temp and ramp/soak segement etc. in the notes section for each roast. It only takes a couple of minutes, but well worth it.

Having just had an intro to cupping, my next challenge is to cup each new roast, and write the cupping notes, also into Artisan. So far I've only been printing and marking what I consider a "nailed" profile for a bean.

Roast as many times as possible (smallest batch). Make detailed notes for every roast, and taste multiple cups of every roast only after they've been rested. Straight from the roaster is no way to judge.

Taste other roasters' coffees, looking for the positives not negatives, then be inspired to improve your own roasts.

I think the only rule of thumb that has held true in my experience is to make adjustments as smoothly as possible trusting that you are far more likely to F things up than fix them. For that reason, I think PID's are a terrible thing with their on/off hysteresis.

Stalling definitely flattens the cup.

There are always anomalies that just baffle me, especially this one: Why did the unwieldy, uncontrollable sample roast come out significantly better than the perfectly controlled production roast with the same general profile?

How much does the roast add to the coffee?
-- This summer, George Howell was speaking at the new Coffee Fest, and his answer was getting more extreme: all caramelization masks the coffee taste, and therefore roasting can never improve coffee, only reveal or screw up its inherent taste.
-- but what about Kenya that tastes better when roasted to cloves and blackcurrants than to champagne and pink grapefruit? Is it an indifferent coffee, not worth the bother?

I'm beginning to think that George Howell is neither completely right nor wrong on this. There are some coffees that can be improved by the roast, and they aren't just the poor ones; while other coffees can only be screwed up. (BTW: I'm becoming quite sure that all Geishas are in the screw up only category)

I think you want to to have enough of the coffee to come to an opinion on this, and to then roast it to its perfect level. For my part, I would start looking for a new coffee after this point; unless I really loved it.

But one person's diminishing returns is another's perfection; and for winning roasting competitions or bragging rights, more exploration is probably needed.

Having a roaster that can be very well controlled. Has proper agitation to keep evenness. It's heat application has enough flexibility(speed to ET) and doesn't abuse the beans. And then being one with that roaster.

More and more I think the higher end of greens have several favorite roasts. Both at level and profile. I can roast a bean several different ways and have a different favorite depending on my taste desires at time of choice. It might be my mood or if it's a hot day or cold one or what time of day it is, what I might be eating while with the cup etc. ect. ect. I have to have several choices available at all times and every once in a while I'll even roast something that has an interesting result at a rather dark level for an option.

Many Kenyans can both be more citrus or more berry depending on how roasted, among other differences. I just can't see that you can get 10 people to all pick the same roast as their favorite. Then the next day would all 10 necessarily all pick the same as their favorite that day. You can certainly screw up a roast with basic roasting defects bringing out generally undesirable characteristics. The challenge is having the knowledge and experience needed to be able to key in on a characteristic and then be able to consider several characteristics with the best desired balances. And then ultimately to repeat.

The point at which I started learning anything was the when I got a roaster that was controllable(quest), then paired it with something that took all(or most) of the roasting notes for me(artisan). It allowed me to be both "noseing" threw the roast, and taking detailed roast notes at the same time. (sort of both worlds your talking about)

Also I think you can experiment with a bunch of new coffees, and have a big stockpile of one or two beans to really play around with. I buy a large stock pile of beans from time to time(Nyeri Tuga AB, Kochere, Konga,etc) and really work at dialing in a profile... all of those beans i actually really struggled with, but hopefully I learned a thing or two along the way. At the same time I had smaller stashes of other beans I could play with, so i didn't get burned out.

Also I would guess the best way to shorten the learning curve would be to have someone highly skilled mentoring you. I say this cause when I train people about espresso extraction, or just extraction in general, I think about how hard it would be for them to learn how to pull a good shot if they were to just teach them selves. Not that it would be impossible, but for sure harder and more frustrating.

A large part of home roasting is learning how to blend (for espresso for me). Learning how to cup and identifying beans by their origin are basic skills. For me, this is not easy and I did miserable in i.d.ing the beans in the 2nd HB coffee roasting contest. Once a year I visit Counter Culture in Raleigh-Durham for their Fri. am cupping, but they no longer present their coffees as an unknown. Trying different commercial (or a competent home roaster) roasts and knowing what goes into them helps. I learned that I like a small amount of Indonesian from trying Hairbender from Stumptown in NYC. Other than that, experience. Roast one bean enough times until you are comfortable with it and can replicate what you have. Living in the Bay area makes it easier as you can go back to Sweet Maria and buy small or large quantities rather than buy mail order and not know what you end up with in the cup.
Sourcing good beans is a large part of home roasting. Most of my beans come from SM and the rest for the buying coop. Some of the posters here buy from wholesalers, distributors etc for which I lack connections, ability and am sure would end up with grade c beans.

Marvin

Wow. What a great question. How does one get to where they ask such great questions

So much of what i've learned along the way seems to be useless. And to some extent it is. It took me 5 years to make a decent shot of espresso. I taught my wife in 5 months. Do I know more than her? Yes. Does it make a difference in the cup? sometimes.

Do all those years with a heatgun and a dogbowl matter? Did I learn something in those years that someone else couldn't learn in a couple of roasting classes? Hopefully. But I'm sure that someone apprenticing under a good, established, roaster could learn more than i have learned in lot less time. So, to become a great roaster, apprentice for a great roaster. (is that even possible? Is there such a thing as an apprentice roaster at Bird Rock Coffee or Ecco Cafe?)

I keep the wild hope that I will "stumble" upon some magic in the roasting process that will make my coffee better than every else's coffee. Do [this] and all your roasts will be amazing. [flatline ET for 30 seconds after the drying phase, or crank the heat between 350-360, etc]. But I'm busy enough changing my finishing temp by 5 degrees, or extending the first ramp by 30 seconds. I never get around to any crazy experimentation that will probably not pan out but will also probably teach me a lot about what not to do to coffee. (except for that one time i tried to deep fry coffee. It definitely improved the body )

I do believe that roasters in general (me included) tend to focus a lot on temperature to the exclusion of airflow. (although technically heat transfer, but airflow is something we can control and temp + airflow is a better approximation of heat transfer than temperature alone). I have a good understanding of how temperatures and times affect my roast. But I haven't yet scratched the surface on airflow.

Ultimately, I think that developing a feedback loop is the way to improvement. (there's probably a fancier name for that). That's why instrumentation is important. If you don't know how the coffee has roasted in the first place, you can't change what you want the next time. For most of us this is thermocouples and dataloggers. But what about roasting notes or even memory? How do we log airflow? Then it's feedback - tasting the coffee and deciding how to do it differently. This is where my wife is invaluable - she's got a better palate than me and will point out the defects in my roast much more quickly than i can identify them. I finally realize why so many roasters have cuppings open to the public - anyone can taste a defect, and sometimes the roaster is too close to the coffee to notice it.

Do we waste time on this journey? Yes, i probably could have instrumented my roasters earlier. I spent a couple of years just making coffee brown because I lacked the data to see small differences between my roasts. Sometimes i just got lucky and had a good roast. Sometimes not. I would have loved to start out with the setup I have now. Each time i've upgraded, i realized how the old equipment was holding me back. This is true with roasting and with brewing.

I guess that's why another question worth asking is: Even if I am not a great roaster, do I enjoy this? Am I happy with myself and my roasting on a day-to-day basis, or am I disappointed? As a good friend once said to me: "life's a garden, dig it!"*


*Ok, I probably need to rent Joe Dirt again.

For most of us this is thermocouples and dataloggers. But what about roasting notes or even memory? How do we log airflow?

Like this:

The craziest %#*$ing thing I've seen all day...