Cupping: From Raw to Ready

Today at the roastery we had a very interesting cupping. We had pulled out a sample from the roast every minute, starting five minutes in and ending at around 15 minutes. This is not a particularly new idea – full credit to Tom at Sweet Marias. His video of it here is worth watching, especially as I am not really going to talk too much about how each bowl tasted.


I didn’t expect to find some stages of coffee roasting to produce such unpleasant effects, or flavours in the cup. Just sniffing the cupping bowls was enough to put me off! Hard to describe, and once water hits it is very different to the aromas you’ll get when using the trier during the roast.

I think it is definitely a worthwhile experience. The only word of caution for those thinking about it is to do with grinding raw/barely roasted coffee. It is incredibly tough stuff, and I was glad to have a VTA6 running full speed to drop the coffee into!

I posted about the little experiment online and got an interesting tweet in response. I had a spare moment this evening, so I went upstairs and rebrewed everything to see what the numbers would say.

Each bowl was done at 60g/l and had the same grind and steep time. I then poured each bowl (rather messily I might add) through a dry v60 paper into another bowl to let it cool. (There is a certain joy in making an absolute mess doing this stuff, knowing there is no one to tell you off!)

I then measured each one in the Extract Mojo. A few very important points before we look at the graph. This is based on a single cupping bowl, and a single experiment. There is plenty of room for error here. Secondly – the refractometer measures the liquid’s refractive index. Software is required to convert this to a strength of coffee liquid. That software has not been created to do accurate calculations for green/barely roasted coffee so the data shouldn’t be considered accurate. (This should be considered no more serious than Tim and I experimenting with the K-ONE written up here!)

Also – I spilled a lot!

Even so – we get an interesting line:

Extraction in a cupping bowl versus roast time

I should add that the numbers (in terms of time) are not accurate and the final sample is more than a minute ahead of the one before it, perhaps explaining the leap. First crack is evident in the jump at around 11 minutes (I realise now this is actually 12 minutes as we’re missing a sample from earlier on – apologies!)

I haven’t tried this particular product (Nescafe Green blend), but having done this cupping today I am fairly sure that Nescafe may well have created something even more disgusting than their regular instant coffee. Impressive work….

7 Comments

  1. In my view Necafe Green Blend tastes watery. Nothing spectacular (obviously) but nothing inherently disgusting beyond being instant. Though I did have a churning stomach after I tried it, whether this is due to the green bean element or the nature of the product I honestly can’t say.

  2. Good post James!

    This has been on my mind since Vinces presentation in the NBC in september. I didn’t really know how to express this thought, but I always imagined that given enough brewing time and consistent temperature, you might get to similar numbers on the chart with a green/barely roasted coffee as you would with a fully roasted batch (with regular brewing parameters of course).

    I’d tasted a barel roasted batch once before and it had comparative underdeveloped tastes in it as you’d find in underextracted coffee. How did the flavour develop in your cups? Astringent -> sweet -> bitter?

    Anyway, a very nice experiment!

  3. We did this after my “Advanced Coffee” training at Java City a couple years ago – it really helped me see what we’re adding to the coffee during the roasting process, and the balance we’re looking for.

    What surprised me the most was how obscenely nutty the coffee was before the roast starts to get it under control around first crack. Then, as the acidity starts to balance out, the coffee comes alive. Roast much longer though, and everything beautiful and lively about the coffee dies and is replaced with a flavor created almost exclusively by a machine. You go from an open, untamed field to an industrial, smoky city.

    Identifying the balance between those extremes is the art.

  4. James,
    I have done a great deal of setting brewers to achieve an acceptable extraction. I always have a look at the roast colour, as this gives me an idea about how readily it will extract. generally, the paler it is the harder to extract, except when you get to dark roast with oils apparent on the surface of the bean, where the trend reverses…. somewhat borne out by your graph.

  5. James,

    Thanks for giving trying this experiment! I’ll be the first to admit that I’m a neophyte when it comes to the science of extraction. While I see the obvious need for tools like Mojo on the roasting/production side of the coffee equation, I’m still struggling to get my head around the way that some people seem to want to use it to codify fairly rigid brewing techniques. Even with a significant margin of error, it’s clear from your chart that extraction has a linear relation to roast. Again, not exactly news– but something that isn’t discussed much on the retail side. To me, as a barista, it indicates that we need to be _more_ flexible in brewing technique. Not only more flexible from one varietal to the next, but in approaching each lot. The more I’m around coffee, the more I realize that it is less like baking (very predictable results with uniform ingredients, exact weights and measures, specific temperatures) and more like cooking with seasonal proteins that require much more subjectivity to tease out the best tastes.

  6. Take a Capao Espresso bean and break it, you will smell Honey Smacks.

Submit a comment