drgary wrote:Michael:

He also writes something to the effect that even using all of his techniques, we are not at the point where even a professional roaster can achieve identical batches of the same coffee.

Which raises the question for repeating a roast: Is coming closer on the phase times more critical than matching the btROR curve.

In Artisan 2.4, there is a new feature called Comparator that is designed to compare roasts of the same bean, and in the Web Ranking Report you can can see the phase times depicted. Along with the new Autosave feature that allows you capture the bean name, it's easy to pick out roasts to compare.

Oh good more espresso.

Here's a related thought. Do we need to have a steadily declining BT ROR or not have a suddenly crashing or flicking ROR? The logic behind this would be wanting the bean to thoroughly develop to its center and to avoid interrupting a cascade of complex chemical changes and have those changes go instead to overcooking the sugars, which would transform them into baked flavor with loss of sweetness.

I've seen profiles of professional Nordic roasters where the BT ROR during ramp is pretty level, not steadily declining. The example he shows in the book is a steadily declining ROR with a soft (i.e., mildly sloping) crash at the very end.

drgary wrote:Here's a related thought. Do we need to have a steadily declining BT ROR or not have a suddenly crashing or flicking ROR? The logic behind this would be wanting the bean to thoroughly develop to its center and to avoid interrupting a cascade of complex chemical changes and have those changes go instead to overcooking the sugars, which would transform them into baked flavor with loss of sweetness.

I've seen profiles of professional Nordic roasters where the BT ROR during ramp is pretty level, not steadily declining. The example he shows in the book is a steadily declining ROR with a soft (i.e., mildly sloping) crash at the very end.

Very interesting topic here!!!

paraphrasing drgary's quote wrote: 1. Do we need to have:
a. a steadily declining BT ROR or
b. not have a suddenly crashing or flicking ROR?

2. The logic behind this would be:
a. wanting the bean to thoroughly develop to its center and
b. to avoid interrupting a cascade of complex chemical changes and have those changes go instead to overcooking the sugars, which would transform them into baked flavor with loss of sweetness.

I find these thought provoking as (1a) implies a roast where BT ROR is steadily declining from start (well almost start) to end. However, a (1b) suddenly crashing or flicking ROR may also mean that it is steadily declining from start, and (may) not preclude that the bean has already been (2a) thoroughly develop to its center.

(2b) is also a very interesting and may have to be based on whether a (1b) suddenly crashing or flicking ROR would adversely interrupt a cascade of complex chemical changes and have those changes go instead to overcooking the sugars (the latter implies a problem brought about by flicking rather than crashing).

These are indeed thought provoking. Thanks for the sharing.

And along those lines, how much crash or flick is too much? And is it more important that straying from an even decline is minimal overall even with some bumps. I have been using the Profile Analyzer to draw the straight line, and to help see where the adjustments should be made next roast. It also shows you how much and how long you stray off course. Most accurate with no smoothing. As to figuring out how close is close enough, I haven't been rigorous enough, yet I am guessing that there are people here willing too be.

CarefreeBuzzBuzz wrote:And along those lines, how much crash or flick is too much?

Or at which point of the roast these crash or flick takes place and for what amount of duration

And yes, definitely more espressos lol

CarefreeBuzzBuzz wrote:Which raises the question for repeating a roast: Is coming closer on the phase times more critical than matching the btROR curve..

Matching the phase milestones affects the flavor profile, steadily declining ROR relates to roast defects.

I have a very hard time discerning even large changes in phase milestones. For example, I did a 500g roast of Panama Elida on my Cormorant. I didn't preheat enough and, using all the gas I had, wound up with a 12:30 roast. My normal roast on the big roaster is 8:00. The Cormorant roast didn't hit dry until over 7:00. But the RoR was steady and declining all the way. The roast tasted wonderful. Even at this extreme difference, it tasted like Panama Elida. The acidity was rounder, the body was syrupier and the cup was creamier over all, but it was yummy. Most of my customers, unless they were A/Bing both coffees at the same time, would likely not notice the difference.

OTOH, I have had experiences wear I got distracted during a roast and let the end of the roast flick because I didn't turn off the gas in time. Even though those roasts hit the same milestones, although a bit hotter at drop, it was noticeably roastier in the cup and a bit dry.

I live for the day when I have the time to sit around and try dozens roast profiles to find that magic sweet spot where I get the perfect tangerine/bergamot note. But the reality is, by the time I figure that out, the coffee would have aged more and it might not be available any more. I settle for 0 roasts defects and tried and true profiles based on density, moisture, variety and terroir.

I have a question.... Does everyone log and try to replicate:
1. the amount of time to cool (to x temp), and
2. the target cooled temperature

And whether these had affected the cup

drgary wrote:I don't know if I want more precision. I like that Scott Rao is identifying things to think about and to try. Precision would be hard to achieve with the great variety of roasters out there, and this includes the size of the roaster. His book is obviously aiming at commercial, production roasting, mostly on machines much larger than we are using at home. The kinds of takeaways I'm getting include thinking about what my gas settings actually mean. I haven't previously thought about the BTU rating of my burners or that when I'm adjusting heat between numerical settings the difference in heat applied is not linear. I also hadn't known I could identify the start of first crack by watching the probe in the exhaust path. Now I'm going to reconnect that to the Phidget. There's plenty more when I have time to go over the book and think about it. I very much agree with the following:



Finally, this book really is for people who have been roasting for awhile and are familiar with some or most of the terms and techniques we've been discussing, including how we're trying to understand roast profile readings presented by roasting software.

I've been confused about using exhaust temp as a cue to mark 1C. Is exhaust temp simply another way of saying external temp? Are the terms interchangeable? Since reading Best Practices, I've started monitoring ET RoR, and while there is typically a dip/spike before 1C, it is often minimal and sometimes not clear at all. The way Rao describes it in the book makes me feel like it should be more pronounced. Am I monitoring it wrong here?

hannson wrote:I have a question.... Does everyone log and try to replicate:
1. the amount of time to cool (to x temp), and
2. the target cooled temperature

And whether these had affected the cup

I highly recommend you try for yourself. My roaster takes a good 5 minutes to get an 8# roast down to 100*F. I've pulled samples from the cooling bin and cooled them in 45s and can't really taste a difference.

Almico wrote: I highly recommend you try for yourself. My roaster takes a good 5 minutes to get an 8# roast down to 100*F. I've pulled samples from the cooling bin and cooled them in 45s and can't really taste a difference.

cool!!!! I'd try it