Level: Alchemist

Reading Time: 15 minutes

Coming from a coffee roasting background, it's difficult to get away from the roasting precision requirements and theory minutia.  I keep telling myself, its not as important with cacao beans, but I still get a little lost trying to plan out what I'm going to do.

What permutations of the different roast profiles would you suggest, in general, to explore cacao roasting?  This is all I can think of:
short + light
short + dark
long + light
long + dark

And how would those be defined, again, in general?  Assuming a behmor charged with ~2lbs - is a short roast 14-18 minutes?  Long roast 22-26 minutes? 

I just drilled through the behmor's drum, and stuck a skinny thermocouple in there, so I can measure temps like a drum, with that in mind, what's a light/dark roast?  light would be 240-260?  Dark 300+?

It's been a while since I had roasted, and chose P1 instead of the usual P2.  Attached P1, set to 20 minutes.  I cut it short because there was quite a bit of popping.  I'm not sure what I should try next, I only have 2 pounds left.  The output was designed for my coffee roaster, several fields go unused, please ignore those. 


Let’s stop right there at the beginning.  For goodness sakes, don’t abandon your coffee roasting experience.  It is virtually 100% transferable to cocoa.  I too started down this road as a coffee roaster and everything I learned and continue to learn about coffee affects and informs how I think about cocoa roasting.

Why do you keep telling yourself the precision and minutia are not as important?   I would put it more like this.  At this moment in time we do not yet know the minutia and don’t have the precision in our roasters to make use of the details if we had them.

To that end, I do roast with solid and definable bean temperature profiles, and both of my go to roasters (my Royal #5 and a homemade one) both have 100% control and have dual thermocouples so I can monitor both air and bean temperature.  So far I have held off presenting a large amount of what I think I know.  It has not been for any greedy reason or wanting to keep the knowledge to myself, but because without the equipment to pull off the advice I give, it is mostly useless.

Well, that is over.

2017 I hereby pronounce the year of cocoa roasting.  It Is time to get serious.  The Gourmia is out that allows you measure bean temperature and air temperature, and allows you to control power.  You have taken my advice to modify your Behmor so you can get bean temperatures and you are recording the data.  The stage is now set to dig deep, geek out, and get really serious about roasting and determine some of that minutia. 

It is pretty well established that coffee roasting has 4 distinct phases in the roast.  The details and names may change but they are basically this.

  1. ·         Drying – Ambient to Yellowing (70-220 F)
  2. ·         Maillard phase – Yellow to 1st crack (220-400)
  3. ·         Flavor development – 1st crack to end of roast (eor) (varies 410-470)
  4. ·         Degree  of caramelization- end temperature.

I have found cocoa is virtually exactly the same except the temperatures are lower, and there is probably fewer if any Maillard reactions due to the lack of reducing sugars (because they were consumed during fermentation).  The profile shapes are the same though.  This is how I approach cocoa profiles.

  1. ·         Drying – Ambient to water release (70-200 F)
  2. ·         Pre-flavor development phase – aroma to near pops (200-235 F)
  3. ·         Flavor development – Aroma change to end of roast (eor) (varies 235-275 F)
  4. ·         Degree of roast - end temperature.

Similar?  Yes.  The same?  No.

I need to put out a disclaimer, and a very bold claim/opinion.  There are lots of people out there roasting in convection ovens at very low temperatures for extended lengths of time.  In some cases they barely get out of the drying phase, or stop in the middle of the pre-flavor development phase, yet make good chocolate.  My only  real comment to that is that if that is what they like, more power to them.  To MY tastes, they are leaving so much on the table.  This style of roasting is how I like chocolate to taste and how I have tried to get people to roast from the very beginning, whether they knew it or not.  Can you make good, or even exceptional chocolate not following the above?  Sure.  Who am I to argue that other ways won’t work or might be more suitable to others taste?  Some people like a 25 minute coffee roast (standard is 12-18 minutes) but the best I can say about those roasts are they are usually inoffensive and most of the time not dynamic (read bland).  I feel the same way about ultra long or ultra low temperature cocoa roasting.  It often results in a chocolate that is inoffensive yet bland to my taste. 

Ok, that was a long disclaimer.

Next.  I don’t have answers. But I have hints of answers.  In this Year of Roasting I am going to start dissecting profiles and how they very specifically affect the final flavor of chocolate.  As I write this I don’t have data I can present but do have 14 years’ experience fine tuning how I roast and mentally filing data away.  And on a more concrete note, I currently have 12 different chocolates of the same origin made with profiles constructed so only one variable at a time is changed and can be correlated directly with another roast.  I’ve tasted them all non-critically and they ARE different and none were even close to bad, and one was utterly exceptional.  For instance, I set up these 6 roasts for direct comparison.  In all cases I left the degree of roast the same at 265 F.  Just as I tend to like a medium roast coffee, I tend to like a medium roast cocoa.

  Here is the plan.

  • 1 & 4 will test if subtle changes in profile will affect the flavor.  My gut feeling is no.  Less than 1 minute will probably be moot and beyond the tasting ability of virtually everyone.
  • 2 & 5 will test the effect of the drying phase.  First indications that a 2 minute difference makes minimal impact.  I have another bean with a 8 and 18 minute drying time.
  • 3 & 4 evaluates how drawing out the flavor development phase changes the taste. Will it enhance it, mute it, modify the acidity, bitterness, chocolate level or what?  There IS a difference (per the non-critical tasting) but I have no prediction yet.
  • 2 & 6 will test how a final hard push effects the flavor.  Will it taste more or less roasted?  Flatter or more dynamic?  Or too subtle to notice?
  • 3 & 4  will further test the impact of the drying phase when it is combined with a final push.
  • The sets 1,2 & 3, and 3 & 5 will compare the overall affect profile shape has in that in each set the roast times and end temperature are the same but the profile shape is different.  My experience is that shape is critical and will document why you can think you roasted the same (18 minutes to 265 F for instance) but your chocolate is seemingly inexplicably and significantly different.

Once I have some correlations I will test them with other beans and attempt to predict how the flavor will change before I taste it, at which point we will have predictive roasting.

So, what should you do?  Don’t think ‘short & dark’ or ‘long & dark’.  It is going to be more complex than that.  It is more complex than that.  Think in terms of times and phases.  Only then can you start to define the actual effects each portion of the profile makes on the resulting flavors in your chocolate.

My utmost hope and goal is to be able to eventually present concrete statements that are effectively universal across all quality beans.  Something like this: 

  • ·         When the pre-flavor phase is lengthened, acidity does X.
  • ·         When the flavor phase is shortened, bitterness does Y
  • ·         The EOR temperature changes the flavor in Z ways if the total time is the same, and in W ways if shortened/lengthened.

That is the goal at least. There is an adage in coffee roasted.  “Roasting isn’t about turning coffee brown.  It is about HOW you turn it brown”.  It is about profiles and how you get to the EOR.  Its the journey, not the destination.  I have no doubt the same is 100% true for cocoa (excepting of course there is no color change).

Oh, as a final answer to your question about light, medium and dark roasts, I hope by now you see that is just inadequate.  It is only one of at least 5 variables that will end up affecting the flavor of your final roast.  That said, I’m hesitant to take any roast beyond 275 F or under 235 F.  What should you try next?  I would roast again on P2 or P3, and stopping at the same EOR temperature, and see if you can make correlations.