Have you converged on a few roast profiles?

Discuss roast levels and profiles for espresso, equipment for roasting coffee.
User avatar
Brewzologist
Supporter ♡
Posts: 1165
Joined: 7 years ago

#1: Post by Brewzologist »

It's with a bit of humility that I write this post, but I'm also wondering what other seasoned roasters have experienced. Do you find yourself converging on a small set of roast profiles you use most of the time? Why and why not? And is this a rut I should try to get out of?

By way of example, I needed some coffee so I banged out roasts of Guatemala, El Salvador, Yemen and a Sumatra/Brazilian blend this weekend. I had different end points in mind for each, but otherwise was kind of on auto-pilot during each roast. Started wondering about them being too similar this morning and lo-and-behold they were:


skink91
Posts: 161
Joined: 1 year ago

#2: Post by skink91 »

I don't roast myself, but this is interesting to me from the standpoint as merely a "consumer" of other's roasts - because I am at a place right now as merely a "drinker" of questioning whether I should be experimenting more with other roasts/regions/etc. outside those I currently prefer.

Do you find this is a similar/analogous scenario to going to your favorite restaurant and only ordering your current favorite thing vs. trying something new but risking missing out on the one(s) you already know you love?

User avatar
baldheadracing
Team HB
Posts: 6225
Joined: 9 years ago

#3: Post by baldheadracing »

I can't read curves, but I do recall Tim Wendelboe saying a few years ago (pre-Loring, so Probat UG) that he roasted a new-to-him green to six default profiles, cupped them, and then picked the one(s) he liked for tweaking. However, there is not a huge variance in the greens that he roasts; you're not going to see wet-hulled Indo's or pulped-natural Brazils, or even very many Naturals.

I only have two roast profiles, but I only have three greens - Aida Batlle's washed, a decaf, and an "I spent so much money on this green that I'm terrified to roast it" freezer-dweller.
-"Good quality brings happiness as you use it" - Nobuho Miya, Kamasada

tinroofrusted
Supporter ♡
Posts: 28
Joined: 2 years ago

#4: Post by tinroofrusted »

I would be overjoyed to be able to produce four roasts with the consistency your graphs show. My roasts are not nearly so consistent. (I roast on a Stir Crazy / Turbo Oven combo setup). May I ask how you are able to display four different roasts on a single graph like you posted? I assume you did that in Artisan, correct? Could you post the steps?

Thanks.

User avatar
Brewzologist (original poster)
Supporter ♡
Posts: 1165
Joined: 7 years ago

#5: Post by Brewzologist (original poster) »

skink91 wrote:I don't roast myself, but this is interesting to me from the standpoint as merely a "consumer" of other's roasts - because I am at a place right now as merely a "drinker" of questioning whether I should be experimenting more with other roasts/regions/etc. outside those I currently prefer.

Do you find this is a similar/analogous scenario to going to your favorite restaurant and only ordering your current favorite thing vs. trying something new but risking missing out on the one(s) you already know you love?
Definitely experiment with coffees from different regions, different roast levels, and different brew methods (e.g filter, espresso.) It might expand the list of what you prefer, or if nothing else make you confident in what you already like. For me the closest restaurant analogy is like trying different cuts of steak each at different degrees of doneness.
tinroofrusted wrote:I would be overjoyed to be able to produce four roasts with the consistency your graphs show. My roasts are not nearly so consistent. (I roast on a Stir Crazy / Turbo Oven combo setup). May I ask how you are able to display four different roasts on a single graph like you posted? I assume you did that in Artisan, correct? Could you post the steps?
Thanks.
Thank you for the kind words. Have a look at the Artisan Comparator. Very easy to use.

Milligan
Supporter ❤
Posts: 1507
Joined: 2 years ago

#6: Post by Milligan »

I've settled on a few ways to go about a roast on each specific machine. I'm still learning the USRC, but with the Cormorant I kind of know what to expect. Each bean is a bit different. At least for my machine, there is a rather narrow window to get through FC without either stalling too much to get to the desired roast level or having some rises. That keeps me within certain roast profiles. I plan to experiment with the diffuser at some point though, so there is still some poking around to do.

Marcelnl
Posts: 3831
Joined: 10 years ago

#7: Post by Marcelnl »

I came to a similar conclusion a while ago, I mostly vary the drop temp and less frequently the duration to FC but mostly that latter parameter is driven by the bean.

The comparison tool is great indeed, I used it a while ago and noticed very little differences across roasts of each bean.
My guess is that environmental impact allows for most of the variance as I'm using a fairly rigid time based 'protocol'.

edit: I have arrived at that protocol also to avoid stalls etc
LMWDP #483

User avatar
LBIespresso
Supporter ❤
Posts: 1226
Joined: 7 years ago

#8: Post by LBIespresso »

I am kind of in the same boat though clearly don't qualify as seasoned with less than 300 roasts. I find that I vary charge temp going lower for naturals and decaf. But charge to yellow and yellow to FC are the same for all of my roasts. I vary my roast with drop temp and time spent in FC as well as ROR at drop but I shoot for 20ish at start of FC and just below 10 at drop. And yet, my curves all look similar to yours.

This comes from what I have learned from Scott Rao. I am now looking to better understand the approach of Rob Hoos, which seems more flexible and varied in the approach. I find all of this talk of extending one phase and getting through other phases quickly fascinating and finally feel like I can steer my Cormorant better. Maybe someday I will feel like I am ready to dig into Neal's concepts.
LMWDP #580

GDM528
Posts: 837
Joined: 2 years ago

#9: Post by GDM528 »

Brewzologist wrote:Started wondering about them being too similar
So, do they taste the same? Presuming they've rested long enough.

Technically none of the roasts are the same, because each used different greens. So by using the same roast profile you've minimized the other influences on the tasting differences. 'Design of Experiment' salutes you!

I've been using a linearly declining RoR profile when first comparing new batches of greens. Is that a reasonable choice?

No two roasts of a given green are the same after that, as I toggle between linear RoR and stepped profile shapes.

User avatar
Brewzologist (original poster)
Supporter ♡
Posts: 1165
Joined: 7 years ago

#10: Post by Brewzologist (original poster) »

GDM528 wrote:So, do they taste the same? Presuming they've rested long enough.
Technically none of the roasts are the same, because each used different greens. So by using the same roast profile you've minimized the other influences on the tasting differences. 'Design of Experiment' salutes you!

I've been using a linearly declining RoR profile when first comparing new batches of greens. Is that a reasonable choice?

No two roasts of a given green are the same after that, as I toggle between linear RoR and stepped profile shapes.
As you know of course, they do not taste the same. 8) I enjoy the terroir of each coffee which is why I roast different ones. But you are correct that by minimizing variables it can in theory remove the roast itself from the tasting differences. The reality is obviously a lot more complicated because each green responds differently to the same roast profile. And as I said I did vary the end points (FC time and drop temp) too a bit for each green based on my preferences for that coffee.

I also typically use a linear declining RoR as a default profile both for new and existing greens. I might go more hot/fast or low/slow but it's still the same general profile, excepting Sumatrans which also often include a gas-dip. I've gotten away from earlier experiments with non-linear profiles, like spending more time in Maillard ala Hoos 'Modulating' book, etc.

So perhaps a clearer question to you and others is whether they have converged on a smaller set of profiles like me? Or if they still experiment or regularly use substantially different profile approaches.

Post Reply