I was reading this article by Rob Hoos on vey small batch roasting on the Bullet on Aillio's Web site, and he seemed to stress out that the goal of sample roasting is just to be able to get a good enough roast to evaluate whether one wants to buy those beans or not. Whereas, when one roasts for drinking, it's a separate process.

It seems that many roast at home with "sample roasters," for drinking (not cupping and buying large quantities that will then be roasted using a more developed roast profile). And I've read for example that one thing people like about the Ikawa Pro is that it's easy to find roast profiles from beans producers. Are those meant for cupping or drinking? And if the former, how easy is it to go from one to the other?

Awesome question. I believe the Ikawa Pro is best suited for its intended purpose, sampling greens. It really shines quickly getting coffee to light roast. There have been good results going darker, profiling, etc as well but for the immense cost of the machine there are better options if quality of cup and output are top of the list.

A major issue that doesn't come up often is servicing. The Ikawa Pro needs to be sent back to Ikawa for a full deep cleaning every year or so with regular use. This isn't a big deal if it is a business expense, but can be pricey for home use. Home users should consider cost per pound of coffee roasted on the machine and this cleaning can add significantly to that cost. The cleaning is roughly $200-300. Some may be able to get by without the cleaning. I'm not sure, this is just what Ikawa has relayed to me.

I feel like the allure of profile sharing has caused a bit of over excitement among us. I've wasted more beans trying out dozens of "pro" profiles than I do dialing in the few that I've spent time modifying myself. I think the smattering of profiles all over the internet causes more frustration than benefit especially for someone learning.

I love getting in new beans and being able to hit them with my few sample profiles I've become very familiar with and have dialed in (after well over 100 roasts.) The ikawa will be very valuable for testing like that once I get my larger roaster. I will not waste time profiling for taste on the Ikawa though. That needs to be done on the "production" roaster.

Lastly, DO REMEMBER that Rob Hoos looks at pretty much everything from a commercial view point. An important aspect of coffee taste profiling is pulling beans at many different levels of roast and timing. This is by far best done on a drum roaster with a trier to pull samples instead of doing 10-15 different roasts on an Ikawa. Many people would be happy with tinkering on an Ikawa at home though. Just depends on what is important to you.

Side note: Roasting for green buying is completely different than for drinking. For example, with green buying I want to know what JUST the beans offer. So, roast to around 390-400F and note the acidity, flavor, off notes, body, and aroma (wet and dry). Then if interested, roast to around 415-420 to see developed acidity and body. A lot of coffees aren't going to be very pleasant for casual drinking at 390-400F or would need to be profiled to really shine.

There are a few users on here with the Ikawa Pro and Home, so there will probably be some great discussion around this.

Having a Ikawa Pro, I find it hard to imagine a reason to sent it back for cleaning. I can imagine a shop wanting to do that because I might be unwilling to trust users with cleaning it because of how easy it is to get liquid on the electronics while cleaning.

Also, if the capability to get a roast you want is possible on an air roaster, I find it had to imagine I can't duplicate it on the Pro. I also expect some other much less expensive competitor will appear with equally good control and better software and maybe a third probe to get an even better idea of what's going on that is able to out do the Pro for 1/3 the price. Truthfully I'm surprised it's not happened yet, but maybe it's to small a market.

He's saying that's the goal of sample roasting, not that those roasters are only good for sample roasting.

ira wrote:Having a Ikawa Pro, I find it hard to imagine a reason to sent it back for cleaning. I can imagine a shop wanting to do that because I might be unwilling to trust users with cleaning it because of how easy it is to get liquid on the electronics while cleaning.

Also, if the capability to get a roast you want is possible on an air roaster, I find it had to imagine I can't duplicate it on the Pro. I also expect some other much less expensive competitor will appear with equally good control and better software and maybe a third probe to get an even better idea of what's going on that is able to out do the Pro for 1/3 the price. Truthfully I'm surprised it's not happened yet, but maybe it's to small a market.

From what Ikawa told me the service includes disassembly, replacing the temp probe, replacing the bottom of the roast chamber, cleaning the fans, and a few other details. It is pretty involved.

In my mind you can sample roast..... On any size machine. The only difference is the size of your sample.

Obviously you don't want to have too large of a batch when sample roasting.

I understand that roasters can request and get provided with small samples of beans from producers. These are like 1 lb or such. If you want to be able to try it a couple of different ways then a hundred grams is a good size sample. Probably no more than 200 g.

Now to drink for me personally, and only roast once a week..... I roast 1-1.5 lb at a time. I don't want any roaster that can't roast at least 1 lb. So I'm really on the borderline of being able to use a true sample roaster for consumption. I mean, I could but I'm not roasting twice a week, I'm not even roasting multiple batches back to back... That's a waste of my time.

I vividly remember starting out with a popcorn popper and having to roast every other day when roasting 80 g at a time. That was awful experience. Because you had to roast if you wanted to have coffee the next day. Not interested in doing that ever again. So I couldn't care less about a little rooster like the ikawa.

Frenchman wrote:... It seems that many roast at home with "sample roasters,"

I can't think of anyone who roasts at home with a sample roaster.

Small roasters that can roast samples, yes, plenty of folks have that, but a multi-barrel sample roaster like a Probat BRZ ... never.

This is a fascinating discussion topic.

So, stepping back a bit, the context (that I'm sure everyone knows) is that if you are sample roasting in a coffee roastery, you have probably gotten a tonne of green samples from your suppliers. In Australia, they would usually be 50-100g, but Rob seems to say that up to 200g is common in North America. Or maybe he gets favourable treatment! So you have 5-10 things that are green and you need them brown to sample them. You also probably only get one shot at roasting them. So if you take the coffee too dark, it is going to be hard to make buying decisions, since dark roasts tend to taste less distinctive from each other. So there's a natural incentive to err on the side of roasting lighter. Every roast is going to be a guess about how the coffee will perform.

In an ideal world, I'd have thought that roasters would want their roasts ... well, their filter roasts, at least, to do the best possible job of showcasing the distinctive qualities of the green they have. So, if they could, they would dial in the sample roasts to what an ideal filter roast would be and go with that. Whether or not this is how roasters come up with their sample roast profiles I don't know. I suspect that the answer is probably yeah, maybe in an ideal world, but in actual fact none of the settings on a sample roaster translate to a larger production roaster, so you are always coming up with roaster-specific profiles, and the ideal filter roast may be the nominal goal for a sample roast.

But the plot kind of thickens! Because those that sample roast are not actually limited to green buyers. Green sellers sample roast, too. Some roasters won't have any sort of sample roaster at all and they ask their sellers to send roasted samples to them. Now, when that happens, you'd kind of think that the sellers are going to have many similar requests and, so, are going to have the chance to really dial their sample roasts in for a particular coffee. Or you'd think that what they would do is to do a production sized roast, freeze it, and send samples. (I'm told that at least one reason why production roasts may be easier to control than sample roasts is because your'e roasting so much more coffee and therefore you get much better coverage of the bean temperature probe than you do with a sample roast, so it's easier to get good information. At least for some types of roasters.) So do green sellers come up with better quality sample roasts than green buyers? I'd say maybe. I've had roasted samples from a bunch of green sellers in Australia and they're usually not great roasts. I suppose that part of the problem is that I don't think there are really a lot of great systems around for automating and repeating roasts. The other problem is that the fast speed of sample roasts means that you really have a very small end point between underdone, perfect and overdone, so it's probably pretty easy to get it a little wrong.

Frenchman wrote:It seems that many roast at home with "sample roasters," for drinking (not cupping and buying large quantities that will then be roasted using a more developed roast profile). And I've read for example that one thing people like about the Ikawa Pro is that it's easy to find roast profiles from beans producers. Are those meant for cupping or drinking? And if the former, how easy is it to go from one to the other?

So with the background above ...

I guess that one nominally ought to think that a sample roast probably should be conceived with the idea that it might at least be delicious and drinkable, but that it will err on the side of being light. I mean, the closer it is to an aromatic style production roast, the better (as opposed to a more traditional "espresso" roast, where the aroma is probably necessarily compromised.)

I think the most practical answer for you is that you're not going to have much of an idea of what any profiles are intended to do. For stuff like the ikawa, it can roast light and it can roast dark; it's going to come down to what the individual user wants. And no roaster is going to tell you that they make underdeveloped, grassy and sour roasts. Nor are they going to tell you that they take things too dark and they're bitter, aromatically dull and unpleasant. They're all going to think their roasts are good, and you'll have no idea. I was utterly staggered that Rob (and the SCA) recommends a sample roast colour that is so incredibly dark. I like coffee light, but those roast levels look like they would be suitable for traditional espresso.

And, I mean, part of it is going to come down to what roast level you consider to be good for drinking.

And as to "sample" roasts v production roasts, see further:

https://www.scottrao.com/blog/2021/2/18 ... -white-lie

I can add that, personally, I've been caught out on sample roasts. I tried a sample roast kenyan coffee and it was a little dark and tasted a bit dry and hay like, so I assumed that that was poor roasting and that it was baked. It was a little bit, but when I received the green, I found that it was very, very easy to develop this undesirable flavour in it. So if I had had a better sample roast, I might not have bought it.

Sample roasts for bean quality should be always be done identically. It should be a light roast in a profile with which you are so familiar you can repeat it in your sleep, and cup it in your sleep. The object is simply to find the best bean in a set of competing samples, and to say yes or no to buying it.

What can be called "trial roasts" are to find the best roast for a certain use of a certain bean; for instance, as a the body component in an espresso blend, or as a medium roast for brewing, etc. etc. Here you should have a "go to" profile for each of these uses that you can fine tune later.

There is no real distinction between these two uses; and no reason to use two different roasters. If you are a home roaster, you will typically use the same roaster to do the final production roasts as well, since you are not producing large lots; a commercial roaster will have one or more large roasters for their production needs.

Depending on your own psychology, you can use a roaster like the Ikawa, with a ramp/soak PID controls, so you can program the profile, or you can use a manual roaster like the Quest, where you monitor bean and environmental temperature and set the heat. I had a home brew, ramp soak controlled air roaster once, but use the manual Quast now. That's just personal psychology. Once I I've got the cupping and purposing done, I paste a label on each of my beans with the profile to be used. One reason I gave up on ramp/soak profiles and went to manual roasting is that my profiling includes smell sign posts, as well as time and temperature ones. It will be a while before automation based on "electronic noses" will be cheap enough for anyone except industrial roasters using Neuhaus Neotech or Probat gear.