Trying to Break-down Profiling, Different types of Profies, and Troubleshooting



First off, please, if you would be so willing:

To start, Happy Roasting, I hope you are enjoying your roaster, brewing some delicious coffee, mastering your cup one bean at a time. This thread, like roasting, is trying to break down the parts to make a delicious roasted coffee only to be ruined by brewing lol just kidding around, but I can’t help there. In my best attempt, I will try to simplify as much as I can, how I do some things; like approaching a new bean, using some different types of profiles and what I do when I’m not getting a roast I prefer. I’ll try to keep it broad and expansive for those who like light and dark but really this is just something I am still trying to figure out and what I say or believe is constantly evolving. In no way is my belief some concrete idea, I am always expecting to be wrong and I am sure there are things I will leave out. So hopefully, in time everything will be in order, or not, which will suck. But if you are expecting some magical coffee god to descend from the heavens with all the coffee answers, keep praying, my friends. I certainly do…into my glass of scotch. Also fore warning: to limit all the "i think"s and “i believe” etc etc, this whole thing is just that so…yea ok let’s begin.

If you can’t preheat, you’ll just have to roast a little longer…

Contrary to the title. It is really important to talk about the 5 main profiles first. The key factor is the main movement of these profiles are along the Z axis. This just means that…well why don’t I show you…

Having the middle of the curve along this axis/line is where the most significant difference of profiles will fall under. This line really affects everything from how much time you can spend in “dry” or the 1st and 2nd segments of the profile, how much momentum is being carried to first crack (which also effects the roast color and outside bean texture and evenness) and how that momentum dissipates (last segment) that either helps flavors develop or become lost.

The 4 Horsemen:
The fan in the “PNG Geometry” was terrible I highly don’t recommend.

Miss Cleo (90’s American TV show):

Charge: Helps with drying the beans, also introducing energy to really dense beans, it might be opposite for us air roasters then in a drum where more energy is needed in the beginning for denser beans than less dense. When I’m able to get nice fruit on a Ethiopian it’s by not pushing the charge/dry too hard. Fan is a magnifier for the temperature profile so as you increase fan speed you are increasing the strength of the temperature at that point. Hence why a high fan at the end can overly dry the beans and a low fan in the beginning can keep moisture like Tony Quiero stated in the SCA panel (he was only talking about the low fan in the beginning to keep moisture) (, a very good listen I highly recommend taking the time for it).

Dry: Setting up the momentum into the “yellow to brown” segment, extending this in the 320F to 360F-380*F really gives an aroma not quite like anything else but exactly like it :wink: Transitioning from a slow dry to a fast yellow is ok but I have not had good results doing the opposite. Practice understanding yellow to brown by making caramel at home. This is the exact same concept. The McGee has talked about Millard and Caramelization at The Harvard as well.

Yellow to Brown: This is a tricky phase, it’s the start of the flavor chain and is the most important for getting a drinkable roast. To slow and thing bake off but also you will get into the marshmallow/vanilla area :slight_smile: and also mesquite bbq wood. Too fast and you will get so much acid that you will create a black hole in your mouth. If you are lucky enough to pace this phase just right, you might be able to smell (very easily/shockingly) fruits, real distinct fruit notes. So far I have been solicited by blueberries, jackfruit, passion-fruit and blackberry. In the sense that I was doing something else and the smell smacked me in the face. But I have no idea how to get it. The faster through Millard/Carmelization, the less body and more ‘acid’ like notes, each profile requires balance, and by tasting after the roast you can determine where to move the profile. But when you move one part you have to move the others, or maybe know that you will have to lengthen or shorten the time you drop after your start of 1C.

Development: On a good profile it is easy to smell the development, from green/veg notes (increase end temp of yellow to brown, but you might/probably have to lengthen/shorten the Y/B phase but can keep the charge/dry/green, start with 15-30s movements), if your momentum/speed/slope is too slow in development. But if your development is fast enough (15-20ish*F/min not a lot of experience playing with this rate) those green notes will transition to something worthy.


Try the other profiles first and see what does what. This is the most important to understanding everything. Keep the fan at 80% with a 2 min cooldown. Seeing which profile get’s you close to what you like. If you like darker notes, lengthen the roast towards the end (12min) , Energy applied in the roast profile is like a see-saw. A darker roast might use a lower charge and dry temp to keep some moisture/flavors for the end to be developed for longer roasts. For lighter and darker, the charge and dry aren’t going to develop flavors but contribute to how the flavors will be developed, the job of yellow/Millard/caramelization is for flavor development. Charge and Dry are used to get the ball rolling and get the party started. If you are not getting the flavor you want in development you have to play with the Y/M/C and the dry/charge, if you found the flavor and want to keep it then you should tweak the back end of the roast, to either shorten the timing or speed it up to develop those green notes. If this sounds like you can affect the roast from both end, that’s because you can.but bigger movements are in the front and smaller ones are in the back (end).

What have I done wrong?
City Roast (to all who simply don't like Light)

Roasting Light to Dark

Hypothesis: Does roasting up to the segment determine the flavor of fruit you will get?
In this profile the temp entered in at “fruit” and developed with a heavy blueberry note. I’m not saying flavors can be made that are not already there but rather that every bean has a characteristic of each segment of the flavor wheel and by entering into that color you get that flavor it has to offer. As for complexity and getting multiple flavors I do not know yet. I don’t think it can be too easy because I’ve tried the PNG FSSF (fast start slow finish) and I did not get all the flavors if anything it tasted overdeveloped, a little burnt and off putting. I have two more test samples of my Papua New Guinea left and I will try them by adjusting the PNG geometry fan at a constant 80% and PNG SSFF (slow start fast finish) but stopping shorter because the flavor was in the ballpark but a little overdeveloped and also vague/undefined flavors but the cracks of 1C was incredible and I will drop at 15s instead of 30s from the start of the first pop.

Using the SCA flavor wheel I applied the color of flavor of the inner ring


Hey Deven do you know where to get the other three pro profiles from the lecture?

Great lecture by the I thought Jen Apodaca was amazing she explains things so clearly!


If you go in your app, on the top left, click where the three lines are, that will bring down the menu -> online library, then scroll down until you see them.


This is great, thanks for posting. You mentioned charge in attaining fruit forward flavors from your ethiopians, but do you have any other guidelines? I’ve found if particularly tricky to hit the sweet spot. Sometimes I get drop dead obvious berry flavors, and then the very next time with no noticeable change in the variables I get a more traditional coffee flavor. I’m struggling to nail these flavors consistently.


I think I was saying that Ethiopians can generally take high charge temps, from what I’ve read on a (mostly) drum roaster forum.

The effect of the fan is still very misunderstood by myself and everything seems to be contradicted with every new theory I test… But a “high” (78-80%) fan speed still seems to lead to much clearer more distinct flavors and when I try to “land” the profile gently at the end with a very low fan (75% and below) I often can char the silver skin and get very muddled, indistinct flavors. And I need to find a new and lower temp range that I can “land” a low fan end profile with.

This has always been an issue for me as well. I look for reasons but nothing seems to be the one answer. A higher humidity will roast the coffee more. Also, now I think because it’s such a small batch size that quality does really matter for consistency. And when I’ve found a profile that can get good flavors I generally roast a couple batches of the bean with that profile and see how consistent it is and how repeatable are the results. I have also seen with one other bean (Ethiopia Sidamo) that beans from a different year (and not too different in bean varietal and region) that a good profile carries over with minimal twerking. However while I am still experimenting I am going to keep buying 86-90pt. coffees.


So I’ve been twerking around trying to test and retest the above statements, still not willing to conclude the hypothesis but I do feel generally comfortable working off of them. What I am trying to do is try and create a profile and theory base that I can (somewhat) reliably create as the Control group for experimenting other profile types like the Spike series of profiles that I think are potentially much better. However I think it’s a mesh of all the “styles” of profiles that will yield the best tasting results. Time will tell.

@george.aris sorry I forgot to mention that yes the above stuff was mainly on Africans with the majority being Ethiopia, dry processed but some wet processed(that comes out more uneven in color, but similar surface texture, silver skin and bean shape).

Some things of note. Maybe it’s not the humidity (but can be), but temperature of my ambient air really makes a difference and when it gets “cold” here in Florida (low 70’s F) from the normal 85F the Marshal Etheo profile came out much too underdeveloped/peanut tasting vs berry. And I think the same as what Stephen said; extending post 1C, would be the better option. I have tried extending the dry phase but it doesn’t seem to help for this particular issue. And can be the reason why I see the potential in the Spike profiles but keep getting burn notes (But also really nice flavors, just vague). I have been trying to combine the Spike and S-curve type profiles in the profile Rao 5.30.18, it seems promising but requires more testing. I think the real key for the combo is by dropping to the temp literally 1 second after the peak spike temp. I’ll keep testing and post about it when I have more info to share. I hope this helps and inspires more people to test out theories and contribute their experiences, and to those who have done so, I look forward to more. Cheers!


Adjusting Marshal Etheo profile due to weather changes:

Answer: raise temp more in the beginning and less at the end, so for the above profile. If I need to adjust for colder weather environment, I would raise the 1st point by 15F, the next point by 10F, then 5F and the last point by 1-5 degrees. So for this profile I am still keeping in line with the shape of the profile but introducing more energy overal and trying not to change the end temp too much. Speaking of end temp adjustment, I usually am stopping the profile myself and not letting it run all the way through. But if you feel comfortable ending at an end temp, make your profile how you like. But i judge you…jk.

I think one underlying question I have been always asking myself is: How do I adjust this profile? Or any profile for that matter. Often times it is easier for me to give good advice to others but when I ask the same question I get a mental block. Irony at its finest…anyways, the more I think about this the more I think the variables come into play i.e. Moisture content(free or fixed in the bean*), environmental temperature, bean age, varietal and growing region. (If I left something out please say so). But I think chemical processes happen in consistant ways, so if you get good tastes out of a profile adjusting similar beans, using that same base shape would be wise.

One thing these variables of the bean have in common is moisture and heat needed to remove that moisture, and by doing so create flavors. By extending a point by time and keeping the same temperature. Heat is spread out. When we increase the temperature but keep the time more heat is applied but doesn’t give much time for those free molecules to react out. For instance if I speed up the begining and make a whole bunch of acid flavors but don’t give them the time to react with or break down, then I’m left with a very tart cup. However at the end of the profile. If I move a point hotter in temperature, in a way, the roasted bean is much more sensitive to temp changes and molecules more likely to break down completley and char/burn, but if I extend out that same point I am not adding too much overall heat and so I can keep moving up but in slower, gentle increments. And my likelihood of getting into the “burn” temp range is lower.

If there is a “green or unpleasant” note the answer is much more likely to increase the heat of the profile while there still is momentum. And the safest way to do that would be towards the beginning of the profile rather towards the end.

I do not think there is a way to create “new” flavors, but there are lots and lots of flavor protential in a speciality grade, quality bean, heat is applied in certain ways can lead to different flavors, I can’t stress enough that this is just simply a guess and will take lots of time to understand whether it is possible or not. The end result is pretty much determined by profile speed, momentum and overall total heat of the profile. So just to reiterate what was posted previously where you end up in that color bar is pretty much what flavor you are going to get, because going flat (where the momentum slowly whittles out, 3rd point on the profile) isn’t going to alter the flavor into a new flavor but develop that flavor into itself. Also to note that color bar moves slightly from bean to bean and is dependent on similarities form bean to bean and changes more on the beans differences. So where that 3rd point is, is what the flavor that’s going to be developed. At least in the Ethiopians and Rawandan I’ve done that are Dry Processed.

*To briefly explain what free and fixed moisture is (as I understand it):

Free moisture doesn’t cost anything…another amazing joke…It is moisture or water molecules that that in the interstitial tissues of the bean they are not part of larger molecules that require heat to break down but simiply just need to be heated to be removed.

Fixed moisture is moisture attached or part of larger molecules that break down to smaller molecules by the removal of a water molecule or a hydogen. When you remove a fixed moisture you are creating a free moisture, however since this happens at a higher temp then free moisture, the removal is easier, but not non-existant.


I think the idea of making adjustments to this profile by getting more heat in earlier makes sense. Tapering the adjustments is interesting and I will try this out.

I live in Scotland (brrr…) so I find that I always have to raise the base Etheo profile compared to Florida. But there are also some beans that seem to need more heat regardless of location. I have a natural Ethiopian Guji just now that really needs it or it comes out underdeveloped. So I will give it a tapered boost and see how it goes.


Yes exactly, I got hooked up on humidity so much last time I encountered this problem, couple with less of a familiarity with how to adjust (still somewhat unsure but more confident, if that can make sense). I would look at Spike profiles and think “well thats going to be too hot” and for me and my weather, it is(!), I don’t know why I was fixated on the concept and not testing against that the consistency of this roaster is infallable (in a way). Then again with trying to fan out my own thoughts, I didn’t have a lot of beans to test other profiles. Hopefully/I feel like, I am near the end of understanding this profile type and how to use it. Time of course will tell. I appreciate the feedback and I’m glad you’ve had good results, for a minute I was getting kind of worried that somehow the results I was getting was an outlyer. If your Guji is wet processed use the MELFS3.4 profile if it’s a DP, make sure you are not flattening out the end and that it still has an 18-12F (I would say 15 is a safe bet) ROR at the end. I just did a couple more blank trial a where I was testing the feedback of the probe/thermometer I’m using compared with the numbers and fan I’m seeing on the app and it seems the temp catches up in the end so front of profile adjustments seems to be the prevailing thought, that experiment was just a self confirmation and I’ll post how it went in a little bit. Lastly a key point to the adjustment is keeping an avg declining ROR, just keep that in mind. I’ve been testing out the golden ration (1.61) but I’m out of beans and need a break for a bit.