James Hoffmann's writing on coffee
Coffeegeek article on crema
Coffee Collective Blog post
The Coffee Collective
Congrats on your first video blog post. I’ll definitely give this a go. This seems to be an extension of your advice for skimming the top off French press coffee – which I now always do. Looking forward to the next vid blog post.
Love the blog love the shirt!
Nice one James ( & Coffee Collective kids ). Something we find helpful at TW’s is serving our espresso in larger cups to allow and encourage people to stir/swirl their espresso – thereby mixing the crema into the rest of the espresso, balancing the flavours and hopefully opening up some of the aromas. Filling an espresso to the rim of the cup is like filling your wine to the brim. No room for swirling and no room for your nose to get stuck right in there…
Oh and a big shout out to the Coffee Collective!
Whenever people come to the roastery we also serve the espresso with spoon, despite the lack of available sugar, and try to encourage them to stir. Stirring is the way, but removing the crema is surprisingly deliciouser.
you are crazy man!!!
why don’t you just drink filter/french press/syphon coffee instead…..
leave the ‘spro alone!
“don’t touch my crema™”
I can’t say that I’ve always agreed with you Mr. Hoffmann, but I have been waiting for something like this for a long time.
I should add that in my case, it’s not because I prefer espresso sans-crema, or because I have some kind of dislike for it, but rather I tend to question things that average folk obsess over. Let me illustrate my point:
I have a very good friend. He’s a fantastic chef (and I do mean chef, as in has worked in some of Montreal’s top restaurants.) He loves coffee.
Any discussion with him about espresso though, and he invariably concentrates solely on crema. I’ve heard other coffee-likers do the same. However this particular friend uses pre-ground, roasted in italy, shipped over in a boat coffee (often illy), and presses a shot (I won’t criticize the machine even if it’s cheap) in about 7 seconds. For all intents and purposes, the crema looks beautiful on his coffees. But they taste well… let’s just say there’s a lot to be desired. (I did use the pepper mill question on him last time. I also chose not flog the horse, but the point was taken – sadly the next step seems to be to use an available blade grinder. Oh well.)
Some other thoughts:
In a closed system like an espresso machine where the only way for liquid and gas to come out during the extraction is uni-directional (ie. into your cup), is not crema the same phenomenon as bloom; bloom being the gas in the beans escaping and combining with coffee bean particulate such that it looks like creamy foam. Again, in the espresso system, bloom has nowhere to go but out the same way as the coffee and ends up as crema.
I don’t really yet care what happens to it, but I do think that there remains an overly obsessive focus, especially with those who really don’t know about coffee, on crema. As usual however, I believe there’s probably some kind of middle ground. We just need to bounce off extremes to find it.
I do like Americanos and I’ve never noticed one way or the other what my coffee shop is doing to the shots before hot water is added; however, there’s no reason why we can’t experiment is there?
Thanks for that nice video. But when it finished, I was left not knowing what to do with my daily espresso. Should I always get rid of the crema? Or is the next step to find out, how much crema is maybe necessary for a good espresso. I guess no one in his café will shot espressos in future and the put away the crema? How are you doing it from now on? And isn´t crema an optical proof of quality which does not have to taste good? Mayber the more crema you get, the better the espresso will taste, when you put it away … to many questions … : (
Moments after watching this and testing it out for myself, a customer walks in and orders a double espresso. I tells him; “I’ve had a revelation mate, and you’re going to be my guinea pig to test a theory. I will give you two espressos and one, you will just stir and drink, the other, you will first taste just the crema and then drink the espresso after you have removed the crema” Bemused, he humours me, follows the instructions and confirms the hypothesis that; crema does indeed taste bad and an espresso tastes better without it.
For some espresso enthusiasts out there this must be akin to being told that the world is not flat and the sun revolves around our spherical earth. Nonsense!! What crap!! Crema is awesome!! people will say; but how will it pan-out? Will it catch on? Will people start to scoop their crema off, or will some places start to serve espresso sans crema?
What a can-o-worms this is, but thanks Coffee Collective for opening it, and thanks James for kicking it over.
If this is what you’re leading off with, that’s a hella lot of pressure for the next four vidcasts this week ;-)
Just tested this with black cat classic at our standard dose (18+g, double ridged bskt, spouted pf) and it was a wash – three of us tried it and came up with three distinct opinions. I found that while the crema is grainy, its not horrible tasting (a little sweet in fact) but nothing you’d seek out, either. Resulting cup w/o crema seems lacking and if anything, edging toward more bitter/licorice. Oddly, the current cherry vanilla bc notes were indistinct in both components when tasted separately, but were there when served as a “whole” shot.
Not trying this at the moment, but I imagine if we updosed for a triple naked the difference would be more in line with your findings.
Using provided specs for our guest espressos, the verve streetlevel does not agree with you, but kenya chania does benefit from a skim. So, perhaps the theory is somewhat coffee/roast/recipe dependent?
Something else to play with in our spare time ;-)
Despite the non-desirability of crema in the finished drink, is not crema (not just the presence of it but also its quality) not still a good indicator of a properly pulled shot of espresso?
I like to think of all of the characteristics one is looking for when pulling a shot as, first and foremost, indicators that things are going along as they should. The final, and most important, test is the drinking.
How does this play out in milk based drinks? And if it’s a great improvement there, what does that mean for the future of latte art? Just a thought.
aw….yes. But why? Or is that saved for tomorrow’s video? ;)
Had a try of this technique today, I can definatley see where you are coming from, and the taste of crema on it’s own does give a somewhat unpleasant taste. I wonder though how much of the overall cup taste is enhanced by the crema and how much it is ruined. I tried a few techniques, and while the removal of crema tends to give a smoother taste, I find that removing a part of the crema and mixing the rest in actually works as a good medium.
Thanks for the Video Blog – keep it up.
Agreed, crema is illogically sought after as something required in espresso for “great flavour.” Anyone who simply skims the top of a well pulled shot will immediately realize what Mr. Hoffmann has so eloquently expressed. While crema may not have a positive role in flavour it is a great indicator of a properly pulled shot of espresso as others have noted.
I enjoyed watching your video discussion of crema. What I find interesting about your discussion is that it mirrors the act of removing the crust from a cupping bowl. Even once those grounds have sunk, that residual crema/crust is better removed than left in the cup. This also echoes the practice outlined in plunger coffee that was mentioned by Henry previously. I guess what it does highlight is commonalities between coffee making methods and that these things when thought of in more practical terms result in the ‘coffee collective’ example.
Is kneeling at the alter of crema nothing but the worship of a false god ? Maybe I wouldn’t go that far, but one thing is true, it made a whole lot of people go out a drink coffee a different way and that’s always good.
All the best,
I agree that crema, compared to the liquid beneath, is bitter.
As for powdery and ashy, while I have experienced those things, I wouldn’t have said they were the norm.
I found the espresso minus the crema somewhat lacking. It was cleaner, brighter, but also less satisfying.
This of course is only a quick comparison with one espresso blend, but my gut feeling would be that while the bitterness of the crema in isolation may be unpleasant, it’s contribution to the overall beverage is worthwhile.
I don’t really like tonic water on it’s own, but throw some gin, ice and maybe a wedge of lemon in there, and your speaking my language.
I definitely think any effort to improve, what I consider to be a pretty mediocre beverage, the americano has to be commended. If you can start bringing some of the flavours of a good filter brew to it… great.
My silly question – what affects the taste of the crema mostly?
Often the first drops of a shot can be really awful, harsh and bitter. If these drops build the top (and taste) of the crema and rising up like the bloom of a french press then I don’t wonder about the bad taste?
So, does crema always tasting bad, especially for some “start dump/center cut” shots or can it be influenced by some preinfusion or slower flow rate/pressure ramp?
Thanks for your intense blogging… (and sorry for my bad spelling).
Definitely a point I have always stressed that the creama is very harsh and bitter. Love that this is getting some voice. The great thing is that I can remember reading an article fromt he collective when they first started playing with that Kenya as Espresso, and the article was all about how creama does not mean good coffee. Cause when they first pulled it they were getting very dark oily shots,but delicious flavours. Cheers. On another note though I do think creama is not completely lost, but that is for another time. ;)
Great post. I also feel a little stupid for the fact that I never thought about tasting the crema separately.
Have you tried smelling the deliverance between the two cups?
definitely starting with a punch James!
Few things, most of it already mentioned above:
– crema on it’s own might not be the best, but it changes how the coffee feels in your mouth. Some people will say that espresso without crema has low body and isn’t thick enough
– crema and crust are a bit different when it comes to removing them from the cup. You remove the crust as it still extracts in the brew, I doubt crema extracts, it just sits on the top is slowly disappears
– crema is not always an indication of a properly pulled shot, the same goes for awesome looking naked extractions, the fact that the shot looks good does not guarantee a good taste
– I haven’t done too much experimenting myself, but is the ‘crema is rubbish’ statement true for all coffees?
Tried skimming and tasting the crema with a couple shots this morning. I guess I’m with the camp that says, “This may be a good thing for shots that are already too bitter, but many shots BENEFIT (in balance and complexity) from a slight bitterness.”
Along these lines, as others have said, there are numerous mixed drinks where bitters are added deliberately.
Also, I would add that tasting the crema separately is not quite “fair”: since the mass of crema is small (compared to the rest of the espresso), its bitterness may be nicely incorporated into the whole IF you mix before tasting.
Some great thoughts. Interesting opinion.
I will be trying this shortly!
I wonder whether or not this is a case by case issue.
Not all crema tastes ashy etc. I have tasted crema alone that was sweet….although not as sweet as the liquid… given that the gas takes up space that could be used to impart flavor.
I would wager that any sweet substance with natural bitters would be made less sweet an more bitter due to the addition of co2.
I wonder also if you not get the same effect by simply stirring so as to “damage” the crema and incorporate the foams solids into the liquid beneath.
Btw…the last thing any barista wants is to to serve a mediocre drink…but also the last thing they want to do is skim every shot for customers! Until this is proven (since it is subjective it will never be “proven”) and there is a machine to produce a crema-less shot I would not leak this one to the press!
I tried this first thing this morning and the initial taste of crema off the top is quite strong and overwhelming and as predicted the espresso, when skimmed, was sweeter. Out of curiosity I tasted the residual crema which had been skimmed after it had sat for around 45 seconds and it had completely changed in flavor and viscosity. It was much sweeter and the body was getting more syrupy. Could that first taste of crema be misleading in the amount of bitterness or harshness it actually leaves in the cup?
Jim, you’re deconstructing espresso, which is a fun lab exercise. Here is a fun exercise that we do in our beginner espresso lab:
Take a “good” shot and break it up into 4 separate extractions. In our training sessions, we do this exercise to point out what flavor and body compounds get extracted at what point during the extraction period of a shot. We basically split a 40 second shot into four, ten second extractions, obviously over extracting the shot for emphasis. You may want to do 5 – 8 second extractions.
Have four shot glasses at the ready and swap them at equal intervals, catching the flow of one “good” shot of espresso into four separate glasses. Now, taste all of them separately. Separately, they all taste like ass. As a complete shot, they taste great.
As a control, you can do this with a double spout pouring a single shot from one spout and dividing the other shot into four extractions. Taste the single to make sure it is a “good shot”, and then taste the four separated extractions.
The crema, like the contents of any of these 4 shot glasses, lends flavor and mouth feel to the entire shot. But, also, like the contents of any of these 4 shot glasses, is raw and one sided alone, and needs the contents of the other glasses the create the full experience of “the shot”.
In any event, my motto is “you need to leave room in the world of coffee for preference”, so if you prefer the espresso that you source, roast and blend with the crema removed and you suggest that your customers drink it that way, then I support you, my friend! I prefer my Espresso Toscano with crema intact, but I have enjoyed tasting it with the crema removed this afternoon, if just for the experience. A great deconstruction exercise, thanks for suggesting it.
Two things: 1) The crew at Intelli Broadway were playing around with line-pressure infusion on their Hydra, and one of the comments I heard was that “the crema was a lot sweeter/more pleasant”. This is a second hand description, but it’s from someone whose palate I trust (and with whom I’ve worked a bar), so I’m inclined to believe it. So is it possible that allowing some of the CO2 to leave the grounds under less harsh conditions is a solution to crema being rubbish? 2) I’ve always thought I was strange for preferring my espresso after it has rested for a while, the crema settling down and the whole thing cooling off. It’s sweeter and the flavor clarity is better. All my career I’ve worked on Lineas, with no active pre-infusion, so again: is pre-infusion the solution to this? Not gicleur-restriced infusion, but full-on preinfusion, E-61 or line-pressure style.
Bonus: How does crema taste from lever machines? Anyone have experience?
Nice one James.
FWIW… after testing this with various folks, I think that “skimmed” shots have great value in evaluation of espresso (a la cupping) and with some coffees yield a “beverage experience” that is enhanced over a more traditional shot — but for my taste I don’t believe this is a universal improvement when it comes to simply drinking espresso. As noted elsewhere, I will be doing this in the future when evaluating a new coffee or blend – but will continue to drink old school shots otherwise.
Oh… and thanks for the advice on swirling the shot when using a naked portafilter. Works like a charm.
This technique, as others have pointed out, is quite specific to certain coffees. I tried it with two blends at work today, one a blend with 20% robusta, and that one certainly benefited from the skimming. The second blend I tried, however, which is an all-arabica and low on finale by design, was thrown completely out of whack by the crema-skim.
This whole technique, then, would need to be tested with lots of coffees to get real results. African coffees, especially, seem to my eyes to need the crema to be mixed in to be complete as an espresso.
an interesting and thought-provoking experiment. Three observations: First, the resulting beverage is now no longer espresso, it lacks the mouthfeel and a good portion of the balance/complexity of a “traditional” espresso. Second, while some crema is bitter, some is not. And lastly, is bitterness a bad thing? Does it balance the drink, or is it just a poor shot? Or is it a perfectly pulled shot on espresso that is too fresh? Does too fresh = too much CO2, and too much resultant bitterness?
For me, I want crema on my espresso. If I want something different, there’s many ways to brew coffee.
– crema and crust are a bit different when it comes to removing them from the cup. You remove the crust as it still extracts in the brew, I doubt crema extracts, it just sits on the top is slowly disappears.
Crema and crust are two different things yes, the crema exists within the crust. Crema comes from extraction whether it is fairly passive as in cupping or plunger or an espresso machine. If crema doesn’t extract where then does it come from ?
You can extract less crema but it is still an extraction none the less. You will find that a paper filter withholds the crema during extraction and does exactly as this blog outlines. You should experiment more I think you get hooked on all this stuff.
I meant that if you leave crust in the brew it will extract more and spoil the brew (in an ordinary pour-over method when brewing in a cup), if you leave the crema (which in my opinion is something specific to espresso) it won’t spoil the espresso, it will simply disappear after a while (not sure what happens there, I guess the CO2 escapes right?).
Great post and I agree about the taste if you only try the crema alone! The only common taste of all coffee is bitterness and most of it is known to be in the crema. So far so good… But, you never drink only the crema! I mean how good does bitter Schweppes taste if you drink it alone without the spirits and a sweetener? It´s the combination that makes it good, not the substance alone…
However, if you are looking for a clean cup of course you should take the crema away. But then you should use another brewing technique. Aeropress, filter, frensh press or why not Brazil cupping?
Conclusion: Crema is bitter and not so good – we´ve known this for ages. But with a great brewed espresso, the combination – crema and coffee- can taste like heaven.
At this time I prefer Aeropress as the no.1 brewing method. Clean cup, nice strength and sweet as honey if you do it right. Sounds like you are looking for something like that.
Getting rid of the crema from espresso? It´s like saying “hey I would like to order a regular filtercoffee, but please make it with crema – because I like the espresso taste”.
You should not use a Formula 1 car as a taxi. It´s two completely different things.
I will try this when I get my next bag. I don’t understand the view that chopping the crema off an espresso means it’s no longer an espresso… surely espresso the drink is the result of espresso the brewing method?! It’s like saying taking the white head off a pint of Guiness means the black stuff underneath is no longer Guinness.
In another change of habit, recently I’ve been allowing (espresso and press) to properly rest and cool down before drinking. The array of flavours that become apparent as the drink cools is eye-opening, I had always seen coffee as something that should be drunk quickly after brewing.
We’ve wrestled with the crema issue for a bit now, too. To me, its like the head on a beer. It is a sign that your using at least relatively fresh coffee, but that’s about it.
Anyhow…while we’re pushing the limits, try this one on–Aeropress filters fit pretty well in a synesso double basket or Simonelli “competition” basket. Rinse it, press it firmly into the basket, then dose on top of it as usual. Makes for a really interesting (and mostly crema-free) shot and cleanup is a breeze.
It does for espresso what cloth or paper filters do for french press. Espresso minus the particulate.
Machine wise:Started with a Gaggia Espresso Stock. Then modified it to the hilt. Only thing missing was pressure gauge. Then bought a Nuovo Simonelli, then finally Quickmill Alexia/PID’d (5yr span).
Bean wise: Started buying pre-ground Illy, then store bought Illy in bean form, then sourced out
local roasters, bought green beans local then bought from Sweet Maria’s,etc. then bought an air roaster, then bought a behmor (3yr span) *grinder: solis graduated to Rancilio looking to move up*
I only drink lattes. I have noticed that larger Crema quantity had a negative effect on taste.
I only came to this conclusion after getting the Alexia. My first few shots were very different.
They were just plain thick. No bubbling mass. They did contain crema but it was very different.
The crema is harsh at times and other times not. As i have noticed it’s very dependent on the type
of beans/roast used. I can really enjoy the crema on some shots when all the factors are right and
other times it’s something i’m drinking just to get to the middle of the glass so i can taste the good stuff.
This is all fantastic stuff. I will use and test as i hope to open a shop in the near future.
If I take an espresso and use a spoon to scoop off the crema, is it still an espresso? Well, if I take an apple and use a knife to peel off the skin, is it still an apple? Yes, they both have a different mouthfeel, but neither becomes something else.
Seeing the reaction to crema-skimming makes me realize how people must have reacted back when machines first started producing crema in the first place. In the present case and the past case, it’s just not what people expected or felt an espresso should taste like.
This reminds me of the first time I met Heather Perry. At the USBC 2008 in Minneapolis a few of us were having a discussion about the rules of the competition. Heather stated that she knew her espresso did not have good crema but she would gladly sacrifice the visual points in order to get the flavor points that were more important. hmm. I don’t think at the time she was necessarily saying “crema is bad, ew” but she knew her espresso and she knew what factors made it taste good or bad regardless of what some rules said an espresso must be in order to taste good. This blew my mind at the time. In a competition setting the winners are the ones who understand the rules enough to know which ones to follow and which ones need to be broken. In real life, understanding the rules is about being able to communicate with our customers and being able to give them something they each can enjoy. This is why I love coffee, it makes us think too much.
I don’t really know – keep experimenting I hope. I think crema is here to stay, I don’t think it will ever be removed from the process – it is an inevitable byproduct of brewing under pressure with fresh coffee.
I think if people take the time to occasionally re-asses their coffee then that is enough.
Crema is an indicator of fresh coffee, pulled relatively correctly. I can produce beautiful looking crema by pulling a truly terrible shot – awful, awful tasting espresso but very pretty.
The information offered by crema really is relatively limited, and I think it is something we worry way too much about.
Personally I would speculate that it is to do with the amount of fines suspended in the foam (that we see as mottling/tiger striping.) These probably don’t taste great in isolation and would certainly explain the decrease in both bitterness and body that most people are reporting.
Thanks for this Jim.
I have tried this during the week, and I really feel like this benefits cappuccinos more than a straight shot of espresso. In the past I have always tried to “aim” for the spot in the cup where there is no crema on top. When skimming off the crema, it creates a uniformily sweet and creamy cap. I tried it with some customers and they all agreed that the capps were smoother.
I do have to say that I miss the mouthfeel in a straight shot though…
Hmmm…so i tried to skim the crema off in a latte as well as stir thoroughly.
Oddly enough, i didn’t like either option.
My shots are pulled straight into a heated cup timed at 25-30 seconds for subsequent latte.
To me the taste was much better then skimming. I tried 5 different beans and roast variations.
All of which tasted better without skimming.
James this proved to be true for us. Our team was blown away by the results . . . now onto asking ourselves what we are going to do in changing protocols.
Perhaps, crema was only ever a surrogate marker for a good espresso?
(Please excuse the digression / talking down, if this is familiar territory).
In biomedical research, quite often a ‘surrogate’ marker, or endpoint, is used/measured, rather than the thing we’re really interested in. This may be for several reasons – it may be easier to measure, cheaper to do so, or it may produce results faster than waiting for the real endpoint of interest. (Drug companies often like this type of study, as it shows up promising data sooner).
A good example would be much of the research into drug treatments for HIV/AIDS. In this case, the real endpoint, (the index we’re interested in), is usually mortality, deaths. But it can take a long time, decades, to collect that sort of data. So instead, we can measure another index, in this case the number of CD4 T cells, in the patient’s blood. We do this before and after the treatment, and look for an improvement. Then, we hope that an improvement in this index will correlate with an improvement in the real index, the one we care about. Often, only time will tell, though we can suggest mechanisms why there might be a correlation in advance.
Crema on espresso may be like this – it correlates with fresh beans, clean equipment, reasonable skills at the machine – but it doesn’t tell us what the cup will taste like. Confounding factors, like the extreme of ‘perfect crema’ devices, make it an imperfect index.
So when we have access to the real index (when we can taste the coffee), we shouldn’t care about the surrogate. Perhaps crema has its place, for instance in telediagnostics and internet forums where tasting the espresso is impossible – and I’m also quite attached to it – but perhaps its importance has been taken way out of proportion.
The crema, like the contents of any of these 4 shot glasses, lends flavor and mouth feel to the entire shot. But, also, like the contents of any of these 4 shot glasses, is raw and one sided alone, and needs the contents of the other glasses the create the full experience of “the shot”.
A researcher in the US found that CO2 tastes sour and has a specific taste receptor on the tongue. So, because of the dissolved CO2 in an espresso crema it tastes more bitter.
Interestingly, when cupping I remove the grounds and foam from the brews I’m testing somewhere between 3 and 5 minutes to prevent the cup becoming bitter as it cools.
An espresso without crema is like a decapitated man and plays a fundamental role in the definition of espresso. When we take the crema out of the espresso, aren’t we applying an additional “manual brewing” method in the overall brewing process? If so, should we still call it espresso? Crema is also a decorative element, an important detail in Gastronomy.
Ey, just some ideas for the future if we want to continue further beyond on experimenting, why we do not blend espressos on the cup? e.g. transfer the best tasting crema of one espresso into a decapitated espresso.
HI James, Is stirring in the crema similar to swirling the crema in the cup before drinking. I seam to get similar results, though the stirring is more romantic and classy. Maybe my taste is not refined. Thoughts?
I think the answer to this debate is based soley on opinion. I t really depends on how you like your espresso. For example, some may like the thinner espresso made in a more authentic italian way in a moka pot with very little if any crema. On the other hand, many like the more viscous shots, thick with crema pulled on 9 bar espresso machines. We often forget that the most important thing about coffee is taste, so this debate is not black or white.
mind blowing if it’s true. I am going to try it very soon. I am with many of the other people who commented on this, I think it what it comes down to is preference. I don’t know yet for sure but I have a feeling there will be more depth in flavor with the crema and a “cleaner” cup without it. And if I want a “clean” cup I will brew a Chemex.
Espresso is a method of extraction.
Maybe the grinder is creating too many fines which float up with the crema? Was the coffee ground with a good grinder?
You must be logged in to post a comment.