Ethiopian Kaffa Anderacha. 100g for 9m20s at 65%. Cracking started at around minute 6. When I was happy, I went to shake it outside in the sieve. I was aiming at a light roast but realizing that it’s not a very even method, I went and made sure even the greener beans had time to get a little tan. I think I’ll use 50g instead next time.
Advice is super welcome.
I just prepared a cup using the James Hoffman daily driver aeropress recipe as soon as it was cool. I know I should let it degas but I wanted to experience the freshest cup possible as a base for later comparison. As is, to my understanding, it tastes planty, caramely and neither too acidic, astringent or bitter. Really nice cup.
You must log in or register to comment.
Looks uneven indeed, but the most important thing is that it tasted good. Maybe you can continuously stir it next time? A roaster basically stirs continuously
I really did, but I suppose I could stir even more. Maybe lowering the amount will help.
I’ve done some roasting in a cast iron, more for the fun of it since it was really difficult to nail. But I remember it taking way longer, at least 15 minutes for first crack/light roast. Stirring constantly for 15 minutes was not the most thrilling thing… It was still uneven but not as much as it looks in your picture. Maybe turn down the heat a bit.
And yeah I agree, the drastic change in taste from freshly roasted to a week old was surprising.
I tried 4 or 5 pan roasted batches but never got the consistency I was looking for. Highly recommend trying the thrifted bread maker with heat gun method!
I will keep an eye for a used bread maker, but I’m just not sure I need a heat gun, so unless I find a use and ask one for Christmas I’ll be doing pan roasts for a while. Still, that method was a fun rabbit hole to explore.
I personally would up the amount to about 200-250g. If you’re going to sit stirring for the whole roast you may as well make it worth the time. With adding more beans, you can lower the heat of the pan when starting the roast and then you wont over-roast anything that that directly touching. That should bring the colour uniform roast. 9m is a good baseline for roast time. You want to aim for about 1.5-2min for the roast after you hear cracking at an even rate. So you want to heat to bring first crack at 7-7:30minute mark. You won’t know how much heat to add until you keep doing more roasts and just trial and error.
The more you roast, its easier to be consistent when aiming for a specific total roast time you aim for. A 9 min roast will be different (but not better or worse) Than a 15min. So just pick a goal and work with that. If you have beans that came out the same colour but one was at 9min and the other at 15, you will find quite a taste difference.
Remember that taste is subjective, and really just try to enjoy the experience as much as your end product.
Batch 3. 100g for 15m 30s at 48% bumped up to 50% near the end. First crack began at 12m 30s, so my next batch will be done at least at 50% power. It’s a little more uniform. Net loss of 17g (17%). I found a comfortable and efficient agitation technique. My second batch was more of a medium roast and I feel like I lost some fruit and chocolate notes, so I’m trying to get back there a little while keeping a pleasant bitterness.
Batch 4. 100g for 12m at 52%. First crack (after 4-5 outliers beginning at minute 3) at 7m. Net loss of 15g (15%). I was aiming for a medium-er roast since the acidity was a bit much. Next time, I’m keeping everything the same, but I will try to use a lid and stir with occasional tossing motions for the first 3 to 5 minutes.
Second try. 50g for 9m at 55%. I let the pan preheat more thoroughly which threw me off. Next time I will heat at not more than 50% with the full preheat. Cracking started at 3m which was a little scary and unexpected. I stirred A LOT. Net loss is 10g (20%).
Haven’t tasted this batch, but the original one feels optimal right about now. I just think the first batch is a little underroasted, so I’ll judge the next batch and see if that’s my new approximate color target.
