Skip to content

Taste that extraction

July 30, 2011

A while back now, I had a Twitter conversation about under and over extraction, and tasting these. I ended up saying that you could taste under/over extraction and recommending an experiment (although a distinctly unscientific one). I realised I had never actually done the experiment myself, and I felt a bit daft. This post is about that experiment (which I have now done) and my new thinking from it. Thanks to the tweeters who prompted this 🙂


I used Scott Rao’s “Everything But Espresso” Gold Cup Brewing Control Charts, but a similar chart is available on the SCAE website.

The experiment:
I used a V60 and two very balanced (and tasty) coffees. I brewed 3 cups of each coffee, to 3 different brew recipes (using scales) – 60 g/l, 70 g/l & 45 g/l approximately (I recorded the various errors, but for the sake of simplicity I’ll just list the results here 🙂 )

I timed the extractions, measured the brewing water temperature and the resulting brew TDS. I tasted the brews (making note of the temperature when tasting) and made notes on my impressions – whatever words came to mind. I also tried them in comparison to the other brew recipes for that coffee.

I used the brew recipes and TDS measurements to identify each brew on the Control Chart and where it fitted in the Over Extracted(Bitter)/Under Extracted/Weak/Strong headings. I then compared the lists for the different brews, looking for patterns. From this I concluded that I tend to use the following descriptors for different extractions:

Strong Extraction: Cloying, Heavy

Weak Extraction: Watery, Cardboardy, Washed Out, Insipid

Underdeveloped Extraction: Sour, Simplistic, Brief, Simple

Overdeveloped Extraction/”Bitter” : Bitter, Not Clean

One of my concerns, and a motivator in doing this pseudo-experiment myself, was the suspicion that we often collect a whole load of different brewing problems (uneven extraction, too strong, unwanted grind distribution, etc.) under the heading “over-extracted”. I still believe this is an easy trap to fall into and one which over-simplifies what is happening, thus setting us back in trying to correct the problems.

I have also concluded that I tend to prefer coffee brewed in the lower half of the gold cup standard box, and am more tolerant of weak extractions than of strong extractions . It’s interesting to see, and it has effected the way I brew coffees – I’m now more likely to try lower dose (55 g/l -ish) and greater extraction % on some coffees (those with low acidity and “heavy” flavours).

I had previously tended to describe over-extracted coffee (i.e. A brew that progresses too far along it’s extraction, leaving it above and right of the desired point on the gold cup chart) as “sour”. My psuedo-experiment suggests that what I am tasting (the element of over-extraction I am sensitive to) is “too strong”, rather than coffee with too great an extraction %

So I’ll finish by saying – I was wrong, you can’t taste over-extraction. You can taste high extraction %, and high extraction strength – but you can only call these over-extracted if you know the intended brew ratio and extraction.


From → Uncategorized

  1. IMHO over and under extraction are horrid, meaningless terms and I just think it’s great when people get experimenting like this because it helps us all start talking in more specific ways. One certainly can’t taste over extraction if one can’t define it in a commonly understood way. I can understand how using such grey terminology helps some people in discussions, as using words rather than percentages arguably holds people’s interest longer, but as a compromise I do think it’s about time we started talking about ‘high’ extractions (yields), or ‘low’ extractions, rather than over/under, and this is where the SCAE Ideal is particularly useful. In a world of manifold variables, it gives us a constant, a reference point, upon which to pin a ‘high’ or ‘low’. I reckon one can distinguish a high or low extraction yield, so I’d say you were right in the first place. But then I haven’t done the experiment yet either, so I’m probably a bit daft too 🙂

    • Thanks mike 🙂 do give it a go yourself – very interesting. I certainly think you can (with experience) recognise differences in both extraction percentage and extraction strength. If you know the coffee a bit, and know your recipe/target, you can (legitimately) describe your brew as under or over extracted (using the definition – Too far/not far enough along your brewing ratio line on the control chart). For that definition, over extraction could also be phrased as “the extraction went too far”. But you therefore shouldn’t (generally) be thinking of other peoples brews in terms of over/under extraction it should be weak/strong & under developed/over developed relative to your preference. It’s tough to stick to that, particularly since most of us (certainly myself) have an implicit assumption people will be using a 60g/l brew recipe and aiming for a spot in the middle of the brew control chart. If the recipe is a given, the terms over extraction, too strong and over developed really will apply equally and always occur together.

  2. Did you use ExtractMojo, Roland? Just wondering how you calculated TDS – and hence how I can too 🙂

    • No – I’m afraid ExtractMoJo is a bit pricey for me 😉 I used a TDS meter. They are rather more approximate, but sufficient for this task (and vastly cheaper!) A web search should give some easily available options if you fancy trying it.

  3. Cool. Been wondering about cheaper alternatives for a while now.. Thanks for the info… got a £40-er coming 🙂

  4. Glad to find your blog today via a retweet from @hasbean.

    two thoughts:
    1) in order to explore over-extraction, under-extraction and “good” extraction in this sort of way, you need a brewing method which helps you to extract *evenly*. Without getting into details, I would argue that a V60 is an inherently difficult dripper to accomplish this with.

    2) consider this: in the real world, ALL coffee is over-extracted, under-extracted, and well-extracted at the SAME time.
    If you took the most perfectly medium-rare steak, and dissected it into tiny bits and collected them by similar color, you’d have a pile of “overcooked” bits, bits that are “properly” cooked, and (perhaps) bits that may be more rare than you’d really want to eat on its own. This obviously depends on your preferences, so steak is perhaps not the best analogy to use, but you get my point. All coffee is a mixture of extraction levels, depending on the size and shape of particles, depending on the brewing geometry and flow-dynamic, and depending on the grind size in general. So the best coffee extraction is one that is either the best-tasting balance of the spectrum of extractions present, or that has the highest proportion of “peak” extraction, depending on your brewing philosophy.

    Keep at it! Will be following your blog!

    • Thanks Nick! On point 1, whilst I agree there are some reasons to be suspicious of how even the extraction of a V60 is, I’m not aware of any firm evidence of this being true. Personally, I’ve found it a very consistent and reliable brew method for me – although the main reason I used it was that it allowed me to produce multiple brews in a very short time. That wasn’t true of the other brew methods I had available at home. On point 2, I’m with you 🙂 the impetus for this post was partly my frustration at the tendency to generalise brewing issues into “over-extracted” or “under-extracted”. I’d love to see more detailed research on brewing, extraction and particle size distribution – and I’m sure that this research will lead to tastier (and more consistently tasty) coffee. At the same time, that the Gold Cup model is a simplification doesn’t detract from it’s usefulness to me. I think it’s a good tool to help us move more people beyond the “too strong”/”too weak” attitude to brewed coffee.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: