Judging the Tan (not racist)


Since being out of the woods, so to say, and my roasting back to normal. I’ve been trying different modifications to profiles. First being the implementation of the “Spike” and the second: playing around before the Millard.

This post is about the latter.

As some of you may be aware, Guatemalan beans have been the real crux of my roasting endeavors. The elusive milk chocolate and caramel, speak nothing to me if crisp green apples and jasmine honey. I still cannot bare the thought. Anyways, by going much lower through it all, especially hitting the “tan/yellow” phase around 400F/204C, however, extending the point to 6min I saw a new color, orange. I saw one of Sweet Maria’s videos and Tom glazed over this idea by saying “tan or orange” and I was like orange…pffff ok whatevs. But then I saw it, maybe the answer to my Guatemalan. Time will tell.

Has anybody seen the orange? What is low and slow for you and what has this “style” been good for you (or bad)?



Coincidentally, I’ve been reading the Hoos book “Modulating the flavour profile of coffee” and been experimenting with extending the Maillard.

Principally adjusting the time to 200c where yellowing begins. Also extending slightly the time between 200 and 1st crack, by reducing the rise a little.

I have a Bolivian which previously I couldn’t get anything decent out of. Now a nice viscous mouthfeel and pleasing chocolate and citrus notes. Also a Nicaraguan which is giving me the elusive buttery mouthfeel.

Playing around very slightly with point 1 is also giving nice results with a previously meh Rwandan as well as a rocking Columbian.

‘BoliMil#1| 50g’ freshly roasted by @ikawacoffee, here’s the recipe #IKAWAHOME - https://share.ikawa.support/profile_home/?CAESECyvjYRjkkuRuJ0NHJAr0z0aDkJvbGlNaWwjMXwgNTBnIgUIABD0AyIGCNcPENIPIgYI7BQQvhEiBgjjGRCyEiIGCLAeENUSIgYIwioQqhIqBQgAEL4BKgYI0RcQvwEqBgjCKhCdATABOgYImjAQ2AE=

50g batch
No preheat
Old firmware
Light/medium roast level

I haven’t had a Guatemalan since the Ikawa one that came with the roaster.

What I’m not sure of yet is what way to alter if you want a bit more chocolate/caramel relative to the fruit. Personally, I prefer fruit over the chocolate notes.


Start the Millard 200c earlier, say around 3-4 min and go slower and longer finishing at a normal but low end temp. I make it happen around the 9:30-10 min mark but this isn’t set in stone yet.

I love fruits but I also get bored easily and look for chocolate as a tasty alternate goal sometimes.



A lot of the heavier melanoidins get created by a long Maillard and these can overwhelm a pourover as they get flushed through the filter by the pour. Which is fine if you are wanting a big mouthfeel which can go nicely with chocolate/caramels but tends to lose the delicates.

It’s always good to mix it up a bit now and then though.

I’ve got a little bit of Sumatran left with which I’ll try the above profile. So far, I have preferred it with a long dry (lost the profile) which highlighted the fruits ( it’s got loads which is probably unusual for a Sumatran). So it should be an interesting comparison.


Sumatrans, I feel like they are the most versitile bean. Just the processing that leaves a ton of defects.


@stephen.pickering21 how would you alter the profile specifically? I’m about to roast some of my Nicaraguan and some insights would be appreciated.

Also in regards to Sumatrans, try roasting it a bit into the roasty notes, and let the roasty notes rest out. I’ve found after some long rest fruityness can emerge.


The Nicuraguan I roasted on the BoliMil#1 profile was a washed Laurina varietal (Bourbon Pointu) which are very small beans. Elevation was moderate at 975-1230m.

Point 1 is (200C, 3m20). I have found that by simply changing that point to (203C, 3m31) it changes the roast just a little. It’s a slight change that reduces the Maillard by hitting 1C quicker so there is a little less body. Loses slightly more moisture.

As of yet I haven’t done any cupping comparisons between the two on the same bean, though I have some Columbian resting to do that.

At this point in time I have been noticing less body between the two profiles across different beans but getting nice fruit notes on both.

So I’d speculate that say for example changing point 1 to (198C,3m10) would bring greater body without making much change to the fruits.

I’ll get round to trying that soon. Problem is that when I hit something I like and it seems repeatable I have a tendency to stick with it rather than experiment too much other than to try other beans on that profile.