Thanks to the expertise here I think I can say that I've developed half-decent skills on being able to take a new coffee and pull a likable shot from it on the equipment that I have. I'm blessed by having access to some great local roasters, but curiosity is leading me to try roasting.

One of the puzzles for me is that roasting throws another set of variables into the mix; basket, dose, grind, temperature are all pretty well controlled, and I'm hoping that an upgrade to a K10 will improve grind even further. I can take a new bag and get a drinkable shot within half a dozen shots and a good one reliably within another half dozen. Twelve shots at around 20 g -- call it 250 g -- a little less than the 12 oz "pounds" sold in San Francisco these days and I can be pretty confident that I'm at least close to what the coffee can do, at least for my taste preferences. The second bag often results in "tweaks" that result in a cup that is even a little better.

I'm planning on trying out an air popper with thermocouple monitoring to start, but am concerned that the batch-to-batch variance in ~100 g batches (four or five shots, it seems) won't let me separate out my inexpert barista skills from changes in my roasting technique.

Is this a valid concern? If so, how can this be mitigated?

When I get around to "upgrading" the roasting rig, what suggestions do people have for a minimum batch size that speeds the learning process?

People will disagree with me; but I'm old school on this -- you can't learn to roast by pulling espresso shots from your home roasts. Brew them instead. Then you'll have plenty of coffee for judging the roast (modern commercial sample roasters only do between 100 and 150 grams).

A roast is good for espresso if, when brewed and left to cool to about 95F, it tastes more sweet than bitter or sour.

The fastest way to learn is to practice lighter roasts, stopped roughly thirty seconds after the first crack ends, when the roast starts smelling of caramel, and no longer has irritating grassy or vinegary aromatics. For these to taste clear and smooth, with no cutting, astringent or ashy flavors, you need to get all the early phases of the roast right. Once you have that, doing excellent medium and darker roasts for espresso is a cakewalk.

Look at it this way. Roast a 100 gram batch. Brew 10 to 20 grams, keep notes, and see how consistent you can keep your roasts. Then make shots of the rest and see if how consistent you can get in the four to five shots you'll have left.

Jeff wrote:I'm planning on trying out an air popper with thermocouple monitoring to start, but am concerned that the batch-to-batch variance in ~100 g batches (four or five shots, it seems) won't let me separate out my inexpert barista skills from changes in my roasting technique.

Valid concern. I found it very challenging to monitor the thermocouple output on a multimeter and tweak roast settings. Things got much easier when I hooked things up to a computer to plot it. I don't want to dissuade you from taking up roasting by raising the barrier to entry, and you don't need to start with this setup, but its something to keep in mind. It might make more sense to roast by sight/sound/smell, then jump straight to computer monitoring rather than TC/multimeter. YMMV.

another_jim wrote:People will disagree with me; but I'm old school on this -- you can't learn to roast by pulling espresso shots from your home roasts. Brew them instead. Then you'll have plenty of coffee for judging the roast (modern commercial sample roasters only do between 100 and 150 grams).

I'm an espresso novice, but completely agree with that. Espresso is so... overwhelming, its hard for me to pinpoint subtleties when my palate is being bombarded. Brew is the equivalent of tasting it in slow motion.

another_jim wrote:People will disagree with me; but I'm old school on this -- you can't learn to roast by pulling espresso shots from your home roasts. Brew them instead. Then you'll have plenty of coffee for judging the roast (modern commercial sample roasters only do between 100 and 150 grams).

A roast is good for espresso if, when brewed and left to cool to about 95F, it tastes more sweet than bitter or sour.

The fastest way to learn is to practice lighter roasts, stopped roughly thirty seconds after the first crack ends, when the roast starts smelling of caramel, and no longer has irritating grassy or vinegary aromatics. For these to taste clear and smooth, with no cutting, astringent or ashy flavors, you need to get all the early phases of the roast right. Once you have that, doing excellent medium and darker roasts for espresso is a cakewalk.

Look at it this way. Roast a 100 gram batch. Brew 10 to 20 grams, keep notes, and see how consistent you can keep your roasts. Then make shots of the rest and see if how consistent you can get in the four to five shots you'll have left.

Thanks Jim. Very useful. As I've bought a 12 lbs sampler from SweetMaria's to start with and only have 1 (maybe 2) pounds of each coffee, this is hard to apply for me now (roast for cupping, adjust/re-do, then be left with no more coffee), but I will definitely follow this advice once I buy a bigger batch of the same greens.

I'm still intrigued as to where this knowledge comes from. Is this all know-how from your own experience? I've read Kenneth Davids' book on home roasting, but still miss all these nuggets I find by reading multiple forum threads (here, SweetMaria's, coffeed).

another_jim wrote:People will disagree with me; but I'm old school on this -- you can't learn to roast by pulling espresso shots from your home roasts. Brew them instead.

I had opposite experience. Most of my roasts when brewed are very good and I don't mind drinking them. I tried trick with letting them cool down and they are nice and tasty. But, when made as espresso they stink.

I now firmly think that the only way to evaluate coffee for espresso is to brew it as espresso... It makes sense to me.

Louis wrote:I'm still intrigued as to where this knowledge comes from.

Hobbyists share their their experience. The first edition of Ken David's book was based on popcorn roasters, and he's trying to teach b people to get something drinkable without too much fine tuning. This remains true for the subsequent editions, although he now addresses common home roasting gear as well. The forums attract people who take it further by customizing their machines and taking professional courses.