Sometimes the most fundamental and important messages seem to get lost in the details. Some of what is said online and otherwise about roasting, and about what I have said about roasting, is mistaken. I’d love to set the record straight on the fundamentals and my beliefs, and there’s no better time to do it than just before my upcoming beginners’ online roasting class.

Roast Time
There is no one “correct” duration for a roast. However, depending on your machine, gas pressure, and batch size, there is a reasonable range of roast times. For example, one can’t say “an eight-minute roast is too fast” without knowing the context. Eight minutes is not too fast when roasting 3kg in a Probat P12, but definitely too fast when roasting 12kg in a P12. It’s all in the context.

I don’t recommend a specific roast duration; instead, I always teach that optimal roast duration is dependent on the ratio of batch size to burner output.

Roast Color

I’ve never told a client how light or dark to roast: to me, that is a personal and business decision. While i’d love to see most roasters roast lighter, and while I personally choose to roast and consume extremely light roasts, light roasting can be a poor business decision for many. I also believe roasters should roast only as light as they have the skill to do successfully; if you often underdevelop coffee, then I recommend roasting a little darker until you have figured out a system to improve development of lighter roasts.

The Best Roasting Machines

I do not recommend any particular brands, at least not publicly. I’ve said kind words publicly about a couple of brands, but that was not a recommendation to buy their machines. If I were to publicly name my top three roasting machines, everyone reading this would be very surprised by at least two of them. When clients ask me what machine to buy, I explain the pros and cons of various machines, and ask them to talk about their preferred roast style and budget. We also consider factors such as which brands offer service in the client’s country. There is no machine that is ideal for everyone. While it’s perfectly sensible for Tim Wendelboe to roast on a Loring, a machine well-suited to light roasting, it’s equally sensible for a second-wave chain to choose a classic-drum roaster such as a Probat. One should consider budget, service, roast style, ease of use, preferred degree of automation, reliability, and several other factors when choosing a machine.

Data Collection

It would be inadvisable to look at a curve I post online and try to copy the numbers using your roasting machine. I may hit first crack at 375f (190c) and drop a batch at 405f (207c) but on your machine the equivalent numbers may be 10f (5c) higher (for example), depending on our relative probe calibrations and the environmental temperatures at those moments in the roasts (ET readings affect BT readings.)

Having a reasonably good probe (2.5mm—3mm diameter, ungrounded is my preference) in a good location, using Cropster or Artisan, and learning to read curves is critical. Prior to data-logging software, specialty roasters as a group made little progress for decades. After data logging became popular, roasters’ learning curves went vertical. Please do not assume that your machine’s manufacturer has set you up for proper data collection. If they are not offering a 3mm (ish) probe and an ET probe, both well located, at the least, please talk to them about it. I find many manufacturers to not be very interested in data collection and presentation, but customer demand for better data collection has altered their decisions significantly over the past five years.

Baked Roasts

It’s taken years, but I’ve won over many roasters to the understanding that baked roasts are caused by hard ROR crashes, not by slow roasting. Some roasters intentionally bake coffee to decrease acidity, though I recommend other methods to accomplish that. Baked coffee is generally less sweet, more hollow-seeming, and often has hints of straw and flatter acidity.

DTR

Sometimes I wish I had never invented the concept of Development Time Ratio. Despite it being just one concept in a 100-page book, 99% of comments about the book have focused on DTR. DTR is often misunderstood, so let’s address that: the book wasn’t written for only those who roast on very lightly (that I am fond of light roasts); it was written for all of the world’s roasters. If you think 15% is the perfect DTR for your middle-of-first-crack drops, please know that you are in the 1% of the world’s very light roasters. It may be appropriate for you, but I would have done a disservice had I written a book full of advice that excludes 99% of the world’s roasters.

DTR is useful as a QC tool, as a target, and as an indicator of a balanced roast curve. However, if your ROR is crashing and flicking all over the place, your DTR doesn’t matter. DTR is also not a good reason to drop a batch; please drop batches based on color or bean temperature. STEP ONE in roasting is to control and smooth your RORs. Step two is to worry about DTR and everything else.

SMOOTH RORs

I’ve saved the most important issue for last: First, please trust that 99% of roasters’ RORs are not smooth enough to eliminate all roast defects. I know this because for some reason people ping me weekly on Instagram showing me their “smooth” curves, but less than 1% of those curves have in fact been smooth. (PS Please don’t DM me on IG with your curves, my inbox is a disaster :0. Thanks.)

It’s very difficult to master smooth RORs. So difficult, that when roasters tell me they don’t believe smooth RORs are good, I know they haven’t actually mastered—and I do mean mastered—smooth RORs. I know this because I’ve consulted for about 600 roasters and 99% of them have been happier with their roasting once their RORs became pretty smooth. Only 10% of those clients are what I would call ‘masters’ of smooth RORs. Mastery takes time and practice, like it does in any pursuit. By definition, it cannot be a formula or something you attain in a few weeks or months.

Many of these comments may be controversial to some people. That’s great news— disagreement leads to our mutual learning, but only if there's engagement. I openly invite contrary opinions and comments so we can talk about it and find some common ground. Thank you.