Whale wrote:
Just to reiterate my query, is there a flavor/effect that can be attributed to a particular profile (temperature vs time curve?

Convex curve tend to make the coffee taste something like...
Concave-straight curve make coffee taste more ...
Flat/straight curve tend to produce.... flavors.
Flat-convex curve will ....
Concave-convex (S) curve ...

Honestly, I don't think anyone here knows the answer to the above question. IF they claim they do, they most probably are completely FOS. And what might be true of one particular bean would likely not be true of others.

Given an "ideal roaster," one with lots of controllable heat generating capacity, well ventilated with a lot of control over the amount and timing of ventilation -- well, then you could probably test these ideas. But most real roasters aren't that finely controllable, and as a result, it is hard enough to try to get your roasts to "behave" to fit a few desired time/temperature parameters that testing these ideas on any one roaster would be damn near impossible.

Whale wrote:
If there is little appreciable difference between the pre-first crack profile (shape of the curve) I would tend to think that to achieve Ken's recommendation to slow down the roast prior to first crack than a straight-convex curve would be the easiest to "program" (if someone has such luxury).

Do not get me wrong. I will try all the above profile with varying time period for myself to find which I like most, if I taste a difference at all. That is, I will try to the limit of the frankenroaster that I will end up building.

Another factor is how much time one has to test every possible thing you can do to a roast profile. I certainly don't have the time for that, and there are so many other variables that it would be hard to ever be sure that differences you were tasting were the result of what you did with the roaster.

Still, it IS true that the beans that you use to make espresso, including whatever you did to them in the roasting process, will have hugely more impact on the drinks you make than with most any barista technique, or changes in equipment beyond a certain basic level.

ken

Ken,
Thanks for your post and starting this thread. Perfect timing for a novice like me. I've roasted eight 1/4lb. batches so far and my times between 1 and 2c are short compared to what you reccommend. I've been hitting 1c between 8:50 and 10:50 depending on bean and when I turn up the heat gun.
I know you've intentionally avoided citing specific roaster info, but do your broad stroke times apply to the heatgun dog bowl method as well. If so, roughly at what point in the roast should I back off temp. and how much ,to extend time between 1c and 2c? Is there a bean color or other indicator I should watch for? If it's by time, how long before 1c or should I go by so a certain time into the roast?
This is precisely the type of stuff I'm trying to learn about before buying a hottop. That's why I'm starting out with a hands on, manual method.
I would very much appreciate any other info or tips about profiling a roast and what different parts, times, etc. affect flavor and how. Hope I'm not asking for too much, but your post came at it from a perspective that helped me understand more so far than all the other vague instructions and articles I've read so far, so I'm hoping to speed up my learning curve through your expertise and interest in helping others become better roasters.

Thank you,

Glenn K.

Ken Fox wrote:At first, moisture departs the bean during the so-called "drying phase." You only need to worry about this with a roaster having the tendency to roast "too fast, " in which case you might choose to apply less heat during the first 3 or 4 minutes of the roast. Roasters that have trouble roasting "fast enough" don't need any attention given to the drying phase, as it will happen, all on its own.

I appreciate the post and its startlingly clear content, but take issue with this portion if Ken's post. It has been my experience that both reaching the drying phase and being able to control the length of time in the drying phase is of equal importance to any other stage of the roasting process.

another_jim (and others) posited that the roasting process can be broken down into 3 "ramps" or "phases" with approximately equal amounts of time in each ramp:

1. Drying - timed from drop-in until the beans change color from a raw, light green to muted light yellow.
2. Ramp to 1C - timed from the end of Drying to the beginning of 1C. By "beginning", I'm excluding the early random pops, and considering the point where the time between pops is less than 1 second.
3. Ramp to finish - timed from the beginning of 1C to the end of roast, including time spent in 2C

Given a consistent bean type, load and my humble stopwatch, I've found that:

* where C = 1st pops of 2C

1. A = 7 minutes, B = 3 minutes, and C = 3 minutes results in a flat, relatively uninteresting brewed cup.
2. A 7, B 7, C 7 tasted like bad toast
3. A 7, B 7, C 3 tasted like cardboard

I continued testing different combinations, and came to the conclusion that the combination of "equal ratio" and total roast time of 10-12 minutes works well for my roaster, which works out to about 3 to 4 minutes spent in each ramp. IMHO there IS such a thing as "not fast enough".

-s

Sherman,
I'm in the process of trying to figure all this stuff out for HG/DB method. Actually between roasts as we speak,(so to speak). Your breakdown of the phases/times is helpful.
What setting are you using on the HG? Do you change settings or move gun closer and farther away at certain points? How far away do you hold it from the beans?
This info would be very helpful in establishing a decent first profile to then experiment with.
Any other knowledge you'd be willing to share re: heat gun roasting would be much appreciated.
I use a Wagner HT3500 with 12 digital heat settings so I have room to play with temp. variance if it will help.

Hope to hear from you,

Glenn K.

Sherman wrote:I appreciate the post and its startlingly clear content, but take issue with this portion if Ken's post. It has been my experience that both reaching the drying phase and being able to control the length of time in the drying phase is of equal importance to any other stage of the roasting process.

another_jim (and others) posited that the roasting process can be broken down into 3 "ramps" or "phases" with approximately equal amounts of time in each ramp:

1. Drying - timed from drop-in until the beans change color from a raw, light green to muted light yellow.
2. Ramp to 1C - timed from the end of Drying to the beginning of 1C. By "beginning", I'm excluding the early random pops, and considering the point where the time between pops is less than 1 second.
3. Ramp to finish - timed from the beginning of 1C to the end of roast, including time spent in 2C

Given a consistent bean type, load and my humble stopwatch, I've found that:

* where C = 1st pops of 2C

1. A = 7 minutes, B = 3 minutes, and C = 3 minutes results in a flat, relatively uninteresting brewed cup.
2. A 7, B 7, C 7 tasted like bad toast
3. A 7, B 7, C 3 tasted like cardboard

I continued testing different combinations, and came to the conclusion that the combination of "equal ratio" and total roast time of 10-12 minutes works well for my roaster, which works out to about 3 to 4 minutes spent in each ramp. IMHO there IS such a thing as "not fast enough".

-s

Hi Sherman,

I guess I'd just have to defy you to tell me when the "drying phase" was over in my sample roaster. A long time ago (YEARS ago) Jim S. and I had some conversations about this subject, about the beans turning a uniform yellow color, and how I was going to use this in my roasting. I had never looked for this before at the time, so I started to look for it, by shining a bright flashlight in through the large tryer hole of my roaster (I now have a permanently mounted LED light that is attached to the smoke hood above my roaster and which points inside the drum). The plain and simple fact was that there WAS NO such period of color uniformity that arrived at a time when it could effect my roast timings, because uniformity of bean color occurs in my roaster just before the onset of first crack; before then there is a fairly wide ranging mixture of bean colors.

My best guess is that this phenomenon is much more likely to occur in a roaster that is principally roasting with hot air (fluid bed, whatever you want to call it). I also think it is highly subjective, even if present and observable. The cracks, on the other hand, are really obvious -- anyone with normal hearing can detect them, and you could get two people with roasting experience, who have never used a particular roaster together, put them in front of the roaster, ask them to write down when the cracks started, and the times they would give would be within 5, at most 10, seconds of each other.

None of the foregoing says a thing about how important is or is not the first 3, 4, 5 or 6 minutes of the roasting process, as compared to any other part of the roasting process, in determining the final results. I just do not believe that one can generalize, across the many different types of roasters, about the importance of this part of the roast, using visual cues as you and/or Jim have described. And my point in writing this post was to come up with a "roaster-independent" set of guidelines that can be incorporated into the thought process of home roasters, that they can use to learn how to think about this process, of roasting coffee.

ken

slowhand wrote:Ken,
I've roasted eight 1/4lb. batches so far and my times between 1 and 2c are short compared to what you reccommend. I've been hitting 1c between 8:50 and 10:50 depending on bean and when I turn up the heat gun.
I know you've intentionally avoided citing specific roaster info, but do your broad stroke times apply to the heatgun dog bowl method as well. If so, roughly at what point in the roast should I back off temp. and how much ,to extend time between 1c and 2c? Is there a bean color or other indicator I should watch for? If it's by time, how long before 1c or should I go by so a certain time into the roast?
This is precisely the type of stuff I'm trying to learn about before buying a hottop. That's why I'm starting out with a hands on, manual method.
I would very much appreciate any other info or tips about profiling a roast and what different parts, times, etc. affect flavor and how. Hope I'm not asking for too much, but your post came at it from a perspective that helped me understand more so far than all the other vague instructions and articles I've read so far, so I'm hoping to speed up my learning curve through your expertise and interest in helping others become better roasters.

Thank you,

Glenn K.

Hi Glenn,

I have to plead a total lack of experience with HG/DD roasting. Intuitively, it should behave like other "fluid bed/air roasters," however there are some obvious differences in that it is so very much operator dependent, and also because the beans are not really contained in some sort of enclosed vessel.

Taking the above for what it is worth (quite possibly,not much), I'd assume that a HG/DD roast would go a bit faster than what I am accustomed to in my sample drum roaster. I don't know what the implications are for the duration of time between onset 1st crack and the end of the roast, however I'd assume that you still need "enough" time for the beans to develop. Whether enough time is 4 or 5 minutes, as in my experience, or is less with your method, I don't know and the only way I can think of to find out would be to test it. If the results of such testing are as obvious as what they were for me and Jim when we tested the results with my roaster, then you won't have to spend very much time tasting the coffees you are comparing because it will be wholly obvious.

I referred in this and in an earlier post to the study that Jim and I did on this subject, with an Ethiopian bean roasted 2 ways, in one case with 2.5 minutes between onset 1st crack and onset 2nd crack, and the other with 4 or 5 minutes in between. I've paticipated in a number of simultaneous blind shot tasting comparison experiments, and this is the only one I ever participated in where the results were so obvious that you couldn't miss them. Usually, the differences are hard to pick up, and if real take a number of shot test pairs to detect. In this case, it was hugely obvious with the first shot pair and every shot pair we tasted afterwards was equally obvious. We were able to tell which shot was made from which coffee sample with 100% accuracy, in a blind tasting. In order to do that, the differences are ENORMOUS, so you should be able to easily test this yourself, without any need for a fancy experimental design.

I wish I could give you better answers to your questions than I have, however I would rather not answer your questions than to give you wrong information.

ken

Ken Fox wrote:Tying this all Together: This is the part that only you can do for yourself and that no one else can help you with. In order to become a successful home roaster, you are going to have to roast a lot of coffee, including many different types. You are going to have to roast a very large number of batches, and to play around with the variables, listed above, that you can control. You will have to look at the beans after you roasted them for any obvious flaws you may have caused, and most importantly, you are going to have to taste the coffee that you roasted, trying to correlate the results you get in your cup with the variables you experimented with on that particular roast or set of roasts. The specifics of exactly how you will need to modify these variables to get the results that you want will only become clear with your own hard work and experience. No one can give you more than general guidelines unless they personally have a lot of experience using the same roasting device that you are using, and in addition, know your taste.

ken

This is the paragraph IMO that comes from roasting wisdom and is the most important. I too when starting kept trying to find the shortcut to success. About all that one can help with to start is heading one in the right direction as the rest of Kens post does. Questions around details are then best asked and answered from experience.

Ken,

I'd love the chance - not because I think that I'd see anything different, but just to be exposed to other roasting methods. By the same token, send me a PM if or when you find yourself in Chicago, and I'd be happy to give you the opportunity to test out my setup if you so desire.

I'm not doubting that, in your roaster, the results are as you suggest. I just don't think that drying time is a device-independent variable, and the basis for my opinion is experience on my setup; nothing more, nothing less.

I did have a chance to see Jim's M3, and I can appreciate that, for that particular roaster at least, there is less concern for a drying phase, as it appears to naturally ramp in a manner that produces good result. If nothing else, this should be evidence toward the idea that the drying ramp is device-dependent.

-s.

Sherman wrote:Ken,

I'd love the chance - not because I think that I'd see anything different, but just to be exposed to other roasting methods. By the same token, send me a PM if or when you find yourself in Chicago, and I'd be happy to give you the opportunity to test out my setup if you so desire.

I'm not doubting that, in your roaster, the results are as you suggest. I just don't think that drying time is a device-independent variable, and the basis for my opinion is experience on my setup; nothing more, nothing less.

I did have a chance to see Jim's M3, and I can appreciate that, for that particular roaster at least, there is less concern for a drying phase, as it appears to naturally ramp in a manner that produces good result. If nothing else, this should be evidence toward the idea that the drying ramp is device-dependent.

-s.

Hi Sherman,

I'd love to see your setup if the opportunity presents . . . . .

Note that I'm not doubting that every part of the roast is important -- I'm sure that they are. And, I suspect that my usual roast profile more or less supports what you are probably doing -- after 3-4 minutes, max, I ramp up the flame height in my sample roaster, to try to get to first crack by what I'd call a desirable time frame (~9 minutes, give or take a little). I do pay attention to the length of the drying period -- I let the roaster run on a lower flame height for around 3 to 3.5 minutes, then I crank up the heat, but not so high as would scorch the beans.

So, these ideas and practices are not by any means mutually exclusive. I've just found on my own setup that I can't go by bean color early on in the roast, because the bean colors are very inhomogeneous until shortly before the onset of 1st crack, which I regard as an easier and more reliable "signpost" along the way, than bean color changes, at least with my equipment.

rgds,

ken

Ken Fox wrote: First Crack should begin by 9 or at most 10 minutes if at all possible, and if it starts before 6 or 7 minutes, it probably indicates a roast that going too fast to have good results. The first crack should start slowly then pick up speed, then appear to peter out slowly.

Thanks for the post. This seems to be where I need to do some experimenting. Even roasting 12oz in the Behmor, I usually hit 1C somewhere around 14min. I've gotten pretty good at stretching the 1st crack to the second, but if I am going to hit 1C under 10min, I'm probably going to have to go down to an 8 oz load on the 1lb setting. Could this explain why I don't seem to get the brighter tones I think I should?