Where are we with pressure profiling?

David Schomer’s comments on pressure profiling this week haven’t been particularly well received it seems.

This is quite frustrating – he’s raising an interesting point, but has done so in a way that allows it to be torn apart due to his presentation. You could say he’s unable to back it up, he’s making such sweeping statements topped up with a self confidence easily labelled as arrogance. 1

My experience with pressure profiling is also fairly limited. Back in August ’08 I had a rather wonderful custom built profiler leant to me for a while, incredibly generously I should add, by John Ermacoff (who – despite not being a coffee industry person – has one of the most interesitng flickr accounts ever!).

People who knew about my experiments with that profiler have often asked why I haven’t really posted much about it online. In fact there is very little posted anywhere about pressure profiling, apart from many people being excited about its possibilities and about how various manufacturers are bringing machines to market to fulfill people’s interest.

Tim Wendelboe’s post has probably been my favourite on it, as I identified with a lot of what he said. If you haven’t read it then I’d suggest going there now.

My own experience’s with profiling could easily be summed up with a sentence stolen from that particular post:

It is also very easy to make really bad tasting shots.

Pressure profiling is immensely frustrating. You change a profile and you change the flavour, there is no denying the impact of changing pressure during the shot on the taste in the cup. Please note that this doesn’t mean the cup is improved, only different to before. Most of the time it is worse. What starts to break your head is wondering if it is worse because the profile is wrong, or because you got the dose/grind/brew weight wrong for that particular profile. Extraction analysis would help a little bit, but you’ll get lost incredibly fast.

Dialling in with tasting is also very difficult. Shot variance is often confusing when it comes to trying to diagnose problems via acidity/astringency/body/bitterness/balance/etc, and espresso palate fatigue was a major issue for me. I’d just start to feel like I was making some headway when I would hit the tasting wall.

What is worth noting is that despite the number of machines out there that can now profile to some extent, we still lack a coherence on even the basics of creating a profile for a certain coffee. I’d be willing to guess that most baristas using them are having a similar experience to me. I’d also hazard a guess that those using them in commercial environments (rather than lab/roastery/training room ones) are using very, very simple profiles so that they stand a chance of dialling in and repeating.

La Marzocco Profiling Chart: Interesting to note that the suggested profiles make no correlation with flavour, only suggestions to make notes on flavour - something you could say about varying dose or brew time.

Does this mean I think we should write off pressure profiling? Should it be dismissed in a Schomer-esque sweeping statement?

No. I think there is, somewhere in it, some mileage and benefit. However, to get at it will require a bit of cohesive work from the industry. A little crowdsourcing would be very useful here. I don’t know where would be a good hub, or what format that should be (forum/messageboard etc).

All of this is a long rambling answer to a barely asked question earlier in the post: Why didn’t I post anything about pressure profiling and my experiences with it?

I’ll be honest and admit that having near infinite control over the profile (which John’s machine offered, along with automation too) left me confused and intimidated. I hope that I’ve learned a little about espresso in the last two years (I certainly feel like I have), and I hope that I could now come back to it with a little more structure and forethought when it comes to experiments. However – at the time I felt I had nothing to say. I had found no repeatable trends, I had no theories, I’d played with a bunch of other people’s profiles (Andy Schecter and Greg Scace have both done way more work on this than me – and I was grateful for their input). I am curious whether other people out there are having a similar experience?

If we don’t make some headway, as a group, then we’ll probably end up abandoning it – too much work, too hard to replicate, too difficult to implement into a busy bar – and we might have missed something potentially very useful. So I will end this with a plea to those who are experimenting to share more – even if all they are sharing is a feeling of confusion and frustration – as all stand to benefit from a little shared experience.

  1. What actually annoys me more is the pseudo science is his writing. ” the preservation of these unstable molecular structures that constitute coffee flavor/aroma” being an example. Ah yes – these unnamed, mystery structures of wonder, so unstable that they’ve somehow survived temperatures in excess of 400F during roasting, but still so unstable that 1F variance in water temp ruins all!  I should also add that in person he’s been nothing but friendly to me, so I feel a little mean writing this.  ↩︎

47 Comments

  1. I’ll note up-front that my purchase order for the Strada EP has been with La Marzocco for a couple months now.

    It seems to me that there has been little discussion about pressure profiling because very few people have not only had the opportunity to play with a PP machine, but even fewer have the experience of using a PP machine more than just a couple of hours.

    PP presents the potential for a whole new paradigm in espresso production. The variations (and complications) are infinite. Guaranteed there will be many more failed shots than successful ones – it is the nature of research and development.

    You’re right to be confused and intimidated. It is confusing and intimidating. PP taken to the extreme of the LM Strada is wildly over the top and probably impractical for daily bar use. However, the EP model offers the potential for a level of automation that makes sense for a production environment.

    There’s nothing that can be deduced now or anytime soon. Schomer’s remarks are so off that they’re laughable and should be quickly dismissed. There is a lot of testing that remains to be done.

  2. Wow, lets not hang Schomer just yet. He is legit and has a track-record of worthy writings, so without him speaking out on PP we would perhaps not be having the discussions that we need to in order for the community to challenge this topic.
    I see a lot of frustrating challenges with repeatability, dialling and diagnostics to find a constant that we can use as a reference point in our pursuit to make the best coffee possible and if it was easy, everybody would be doing it.
    Thanks for keeping it alive James.
    geir oglend

  3. A whole lot rings true in this post of yours, thanks. A couple comments:

    (1) In addition to your comment about the limitations of LM’s “guide” about profiling, reports about positive changes ascribed to pressure profile manipulation are completely worthless if the shot volume (ie, grams of beverage) has been changed, too. This is because, as we all know, if one pulls a shorter (fewer grams) or longer (more grams) shot from a given dose, the flavor and texture will change. Some people, in reporting fantastic shots from a profiling machine, have been so vague about their shot parameters that it is impossible to know what other variables have been changed.

    (2) Aside from concentration, a key difference between espresso and brewed coffee is that the high pressure espresso extraction emulsifies oils and produces the famous (and hackneyed) “polyphasic colloidal foam.” FWIW, I haven’t seen much mention of the fact that if one spends a significant proportion of the extraction a lower than normal pressure, one’s beverage becomes more and more like brewed coffee and less like espresso.

  4. Geir said, “lets not hang Schomer just yet.”

    Why not? I was ready to hang him back in 2001.
    Hahahaha, just kidding. ;-)

  5. James, Thanks for posting your comments with regards to Davids post. Don’t worry about questioning Davids opinion as I think and hope this is the dialog he intended us to get going. The post by David was much to short and wordy to be considered anything more than compelling, I suspect.

    My hope is that we can find new terminology to reflect what is actually occuring with this perceived pressure profiling. Pressure manipulation, adjustable / variabale pressure, Manual pump pressure adjustment….. Heck I dont know. So far, Jepy, and LM have prototypes of profiling machines, and there are others who may or may not be further along in research, however the only true profiling machine available to purchase at this time is from Cimbali…… Hmmm, seems nobody is talking bout that one. Anyhow, I think and agree, that we still have work to do, to determine the true value, if any to pressure manipulation while extracting espresso.

  6. First off, what do you mean by profiling…some of my vocab isn’t up to par…I get what pressure you are talking about, but why pressure profiling, I don’t really understand.

    Not trying to be pessimistic, but It feel to me as if yes, this change in pressure will change the espresso, but I don’t know if it could be not noticeable. I’m NOT saying that this is wrong, obviously I have little to no experience with this. But, what I’m getting at is that you will taste a certain characteristic that you have in your mind. If you think more or less body, chances are, you will feel that because It’s already engrained in your mind.

    Just a thought. But once the ideal pressure is set, it seems like a good idea in order to make your coffee unique and flavorful. Consistency is one of the most important characteristics of coffee to me.

    Perhaps this is what makes one café different from the other.

    Now, once again, just thoughts. But this is how coffee has gotten this far right? : )
    Experiments

  7. I’m really curious about the La Cimbali. Has anyone actually used one? I saw it in Milan at HOST, but they weren’t letting anyone actually touch it, and I assume it was full of stuff they were waiting on patents for. Has it been released?

    Agreed on needing better terminology for the process of varying the pressure exerted on the coffee at different times during the extraction.

    Is anyone tying pressure into flow rate, rather than time?

  8. There are an increasing number of machines that will allow you to vary the pump pressure hitting the cake at different stages of extraction. Original pressure profiles – i.e. lever machines – used a varying pressure as a result of the spring inside the lever expanding. This was replaced by rotary pumps that delivered consistent pressure throughout the shot – sometimes with a mechanism to influence how quickly they reached that pressure (such as flow restrictors, which create a form of preinfusion).

    Pressure has a massive influence on all sensory aspects of the cup – but we really don’t understand how because we lack systematic experimentation as an industry.

  9. I wonder how PP machines like the Strada compare to the traditional (fixed, as long as the paramaters remain the same – amount of coffee/grind etc.) pressure profile of a manual lever machine (I know you are experienced with a beautiful Victoria Arduina Athena :-) ).

  10. You mean if they exerted identical profiles? The tricky bit would also be matching the temp profile of a lever machine – though the Athena is more stable than many I’ve seen.

  11. Good point Andy. “FWIW, I haven’t seen much mention of the fact that if one spends a significant proportion of the extraction a lower than normal pressure, one’s beverage becomes more and more like brewed coffee and less like espresso.”
    Yes and all we need is a 15-20000 dollar coffee brewer. I know that you and others like John and Greg has spent a lot of time/research and money on PP with great success and honours.
    Maybe we need to change the temperature during the shot and not so much the pressure and see were that takes us?
    Correct me if I’m wrong, did PP concept arrive from trying to copy the Piston/Lever profiling?? Were it would have a soft pre-infusion and then ramp up and change pressure as the piston would be slowly pushing the column of water though the coffee bed??
    I love this concept and any body who has had a properly prepared espresso from a lever machine will probably agree. And I must ad that the coffee was not hot like it is from an regular espresso shot all things being equal.
    That’s my rant for the year. sorry James.
    geir.
    BTW this is the reason we like the 3group Strada MP over the more expensive version so if this whole PP thing is Rubbish we still have a regular espresso machine with PID and pre-infusion.

  12. We have put in an order for an EP version of the Strada. Like others have mentioned, playing on the MP Strada we managed to create some gorgeous results but are worried about how it could work commercially in our cafe. We are purchasing one for the same reasons that Tim Wendelboe mentions. Once you have come across results with certain coffees that seemed unattainable previously we didn’t want to go back.
    I agree that it is going to be very frustrating at times and that we will get lost chasing our tail quite often, but feel that it Is worth pursuing. Once we have installed the EP Strada id love to start chronicling our results/ parameters used and be part of a messageboard/forum as you have mentioned.

  13. “What starts to break your head is wondering if it is worse because the profile is wrong, or because you got the dose/grind/brew weight wrong for that particular profile. ”

    This is why I have been entirely unable to get excited about this new generation of machines. The energy required to chase a wild goose through kilos of a new blend to get dialled in when temp, pressure, grind and dose are all for the changing is simply exhausting.

    I am a firm believer in the importance of a varying pressure profile, and anyone who wants to see its effects directly just needs to get hold of a Pavoni for a day. This is why I have a lever machine in the shop – its profile puts its own accent on the brew. I can’t alter it to do my bidding and amen to that; I don’t want my machine asking me questions that cause a headache – just shut up and get on with your job already.

    I’m certainly glad that this technology exists and that the major players in the market get excited enough about the top 1% of espresso to put serious R&D dollar into these machines but I really can’t see their commercial purpose beyond the bragging rights.

  14. Hi James! I was sitting on the floor with technicians of Cimbali in Italy more then one year ago, drawing and discussing about pressure adjustment on the machine. (Not even knowing about the strada yet! La Marzocco is not very common in Austria and I didn’t follow any blogs those days) Anyway – my intention was to get them to make a machine with a pump that copys the pressure of a lever machine. Purely for the reason that I found my coffees always better on those lever machines. I assumed it was due to the higher pressure of the stretching spring at the beginning after releasing the lever (more then 10 bar) and the lower pressure at the end of the shot. Which made sense to me due to the fact that there is more to extract at the beginning of a shot then at the end…material is almost empty – or better – we don’t want too much out of the material. But no need explaining this to you.
    Anyway, one year later, last spring, I was back in Italy at Cimbali and they proudly prsented me the one prototype (that was also in London later) and I adjusting it together with the technician. It’s complicated, you have to get into the menu on the display (“mice-cinema” as we call it :-D) and set 4 phases (time and pressure).
    So I set the first phase for 1 second and 4 bar – like a pre-infusion on a lever.
    The second phase with 9 bar and about 7 seconds, then 7 bar for the third phase and 5 bar for the rest of the shot.
    On the other group we still had the pressure adjusted to 9bar constantly.
    Compared the two shots with each other I have to say that the one with the 4 different pressure phases that should copy the sinking pressure of the lever machine was more acid, but also much brighter in flavors and lively. Deffinitely much better then the “normal” shot. I don’t remember what blend was in the grinder but we deffinitely got the “better extraction” with the lever-copy!

    I am realistc that it would be a difficult and time consuming target for the technicians/and salesmen in the future to adjust each machine for the individual coffees in every coffeebar (they don’t do it yet either) – so this systhem will not be for the masses. But For some of us this is a great opportunity to explore our coffees. And beeing realistic this will not be a mass product, but as my colleague Georg said who was with me the last time: AUDI also has a “quattro” Modell. Not for the masses, but for some people just the right thing :-)

    So the machine from Cimbali should be released 2012 as they have their 100year anniversary. If everything works out until then.

    I just got a used synesso for my school, but the model with the paddel that let’s me pre-infuse… I thought I’ll have a paddle for pressure adjustment, too, but nope. Anyway, still have a lot to play here, too ;-)

    Greetings from Vienna!
    Johanna

  15. Can’t we get some intelligence on this by asking long-time manual lever machine users? Surely at least some of them have figured out how to “dial in” their pressure profiles for different coffees, how to understand which effects are down to grind and dose and which to pressure? Or are manual lever machines such different beasts that this wouldn’t help? Or, perhaps, are all manual lever users struggling to get something like the standard pressure profile of a pump machine?

  16. > I am a firm believer in the importance of a varying pressure profile,
    > and anyone who wants to see its effects directly just needs to get
    > hold of a Pavoni for a day. This is why I have a lever machine in
    > the shop – its profile puts its own accent on the brew.

    The problem with what you’re saying is that (as others have mentioned) the “effects of the Pavoni” may be due to a combination of factors: pressure profile, temperature profile, method of water infusion, etc.
    It’s not that hard to put a profiling pump on a conventional machine, and then pull shots using a Pavoni-like pressure profile. But the coffee doesn’t suddenly become magical.
    It appears that there is more to lever machine shots than just the pressure profile. This would include the romance of the lever, which may significantly alter people’s perception of the coffee…. :-)

  17. The lever/piston topic is well covered over on Home Barista.
    The problem with most of the lever/piston machines is that the group picks up water from the same boiler as used for steaming milk, which is IMO too hot for coffee to be brewed.
    So when I said properly prepared espresso I was referring to lever machine with either HX or the temp turned down to around 3/4 bar. Presently I have a single la Cimbali Eleva from the 70s that makes a wonderful espresso and tastes different than when brewed on a conventional pump machine. My opinion aside from the romance of the lever, the flavour is better/different because of the volume and column of water being pushed through the coffee slowly and evenly with thorough saturation and seemingly less coffee.
    geir

  18. I was interested by Tim’s comment :
    “A profile programmed for one coffee on a Wednesday does not necessarily work with the same coffee on Saturday. The coffee degasses and this changes the taste a lot. I had to totally change the pressure profile from Wednesday to Saturday even if I was using the same coffee from the same roasted batch.”

    Now certainly this makes sense, as the coffee will change as it degasses etc. I’m interested though that it occurs to me now I don’t think I’ve noticed anyone make the same observation regarding temperature.
    I definately find that as our coffee ages, I need to increase the boiler temp on our GB5. As the coffee stales, fewer flavours are as easily available, and so more heat energy is needed to extract them. To be clear I’m talking about an increase of normally just 0.1 degree C in the boiler temp each day.

    Man I wish I had a chance to play with a Strada

  19. Hello James et all,
    My name is Scott Guglielmino, and through my work at La Marzocco USA I feel that I might be in a unique position to be able to comment on what I personally see as results from pressure profiling. However before I venture a comment I would like to first express a couple of disclaimers. First of which is that my comments come from my own personal experience using a prototype version of the Strada EP in testing. While I have pulled shots numbering in the thousands of this machine, the primary goal was not to create a theory of pressure profiling but instead to test the machine’s construction. That being said, I do believe I have come away from the experience with at least a limited person theory of pressure profiling. Additionally I refuse to take part in an online referendum on any one coffee professional’s views on the technology. Of course I also need to state that my comments are mine alone, and not to be taken as the viewpoint of my employer.
    My personal, very limited theory of pressure profiling as the following basic assumptions; first of which is that by manipulating pressure in a variable, controlled manner I am able to change specific flavors in coffee. As you stated in your post I also find that by manipulating pressure I am able to make what I like to call lateral changes in a coffee’s flavor. Making lateral changes in a coffee’s flavor (that is to say changes that are neither good nor bad just different) are rather meaningless without a clear target or reason. My second stipulation is that my goal in an espresso (or any other brew method) is a beverage that is the best representation of the coffee I am preparing. My third stipulation is that every coffee (of an acceptable quality and roast) does not equally perform in the espresso process when given a linear pressure of 9 bars through the extraction. That is to say that I have had coffee’s which I am unable to reproduce a flavor profile in espresso that is comparable to when that coffee is cupped or brewed in various other fashions.
    Following these basic assumptions I am able to then state my goal/theory of pressure profiling (I hoped to never write those words). By manipulating pressure I am able to make lateral changes in a coffee’s flavor profile that allow me to best prepare a coffee, results that I would not be able to achieve with a static brew pressure of 9 bars.
    While there are many specific points I would like to address I feel that I may have already shared far too much of my own personal viewpoint with little past subjective tasting as factual basis for it. It is not my goal to be correct about pressure profiling; I simply wish to share my observations of what I have learned from my uncontrolled tests of a new process. I would invite anyone who wishes to discuss my thoughts in a private manner to feel free to email me. My email address is Scott@lamarzoccousa.com
    Thank you all for spending the time to read my thoughts on this subject,
    Scott
    I would like to add that the first generation of espresso machines brewed at a pressure around 1.5B. For the purpose of this conversation I wish to stipulate that espresso can be defined as “a brew obtained by percolation of hot water under pressure though a cake of roasted ground coffee, where the energy of the water pressure is spent within the cake.” (Illy & Viani, 1995)

  20. I spent half a day on Greg Scace’s pressure-profiling machine a few years ago, and like the others, my assessment was: different, not better. Nothing I’ve seen has changed this assessment.

    I do, however, like the idea of using a Strada EP profiling group to set the initial ramp-up pressure. In other words, an inline gicleur (flow-restrictor) works with the plumbing (particularly the volume of plumbing between the gicleur and the coffee cake) to establish a pressure ramp profile. A Strada EP group would allow a customized ramp-up curve, which in my experience, can yield notable quality effects.

    My greatest frustration through the whole pressure-profiling thing has been the assumption that pressure profiling means “more control,” and that this is a good thing. I’m reminded of experiences with keyboard-hero baristas who blog and wax poetic about nuance and such, but who dose their espresso with horrible unevenness and demonstrate little actual skill. More control is great when applied in places where it is helpful. But just as the human body has autonomic systems that don’t require conscious control, maybe there are things in espresso (and coffee in general) that we simply don’t need to be fiddling with. Otherwise, as others have already stated, it just gives you more ways to fuck things up.

    The tricky part is knowing the difference.

  21. I’ve been using a Slayer for around a year now. I realise it works differently and isn’t necessarily “pressure profiling”

    Just for people who don’t know, it controls the flowrate on the middle stage and this is fixed while pulling a shot. Its not a pressure setting but pressure is determined by the resistance on the puck in relation to the set flow rate. Set the flowrate to around 60mls/min and you see the coffee start to drip from the basket at around 20 seconds and a pressure of around 4 bar.

    To generalise, at this setting we find it just accentuates what is already there. If the coffee was roasted just a little too dark it will taste overly ashy, acidic coffee can become blindingly bright. On the other hand a nice sweet, balanced espresso becomes even more so. Temperature settings also make bigger differences and mistakes are also amplified.

    I’ve read a few people saying the slow ramp up helps with channeling and other problems but we have found the opposite. Shots channel much more easily the slower the ramp up.

    In the end I think its the coffee and roasting that makes a much bigger difference but I definitely think its a good thing.

    Sometimes.

  22. Hi there:

    My original interest in pressure profiling was in trying to figure out why so many people raved about lever machine-produced espresso. One main difference between spring levers and pump machines is the variable pre-infusion and declining brew pressure produced by a spring lever, compared to fixed pre-infusion and constant pressure. Since it was easier to vary pressure than to vary temperature in my Linea, I built a system to vary pressure. I’m gonna respond to you and throw in my two bits for Schomer in the same post.

    For you, Mr.Schomer, because I know that you come here for inspiration and insight:

    First, I got from your blog that you have decided that a few hours of informal testing is all that is needed for you to decide that pr3essure profiling is bunk. I say that’s fine., but I have these questions. Why do people like lever machines? They don’t have flat temperature profiles that fit your model of constant temperature, and they have an unusual declining pressure profile during the shot. There must be something there, because many good industry pros have high opinions of these machines. The difference has to be related to only a few things. Either temperature variation or pressure variation within the shot makes them stand out from modern pump machines. Could be one. Could be the other or a combination of both. Could also be that the people who like levers are deluding themselves. Can you please comment on this?

    Second, it’s pretty obvious to anyone that has ever fooled around with pre-infusion rates and times that messing with these variables has a pretty noticeable affect on the cup. This is certainly pressure profiling because in order to change preinfusion rates snd times, one must change the pressure conditions within the coffee cake. It does not matter if the profiling is done mechanically, as in a spring-loaded E-61 preinfusion chamber, through flow restrictions like installing 0.6mm gicleurs in LM double boiler machines, or through electronic controls. It’s still profiling. Denying that there is any benefit in profiling means that to you there is no benefit in any pre-infusion scheme at all. I think that there are plenty of folks who will disagree with you on that.

    My philosophy on espresso machine design and modification is that I’ve always worked toward reproducibility in machine parameters, and in trying to decouple effects of one parameter from those of another. It’s the ability to reproduce brewing conditions from shot to shot that is of utmost importance. These days we’re only successful for certain conditions. We’re only good now at producing brew temperatures with little shot to shot variance for cases in which the brew temperature is relatively constant at the top of the coffee cake during the brewing process. (ignoring time-dependent spatial temperature differences within the coffee cake). We don’t know if that’s the best profile and lots of folks argue that it is not, without any reproducible data (not their fault). However, I think we can make a pretty good case that consistency is good. WRT pressure and pressure profiling- regardless of whether or not fancy profiles means anything, one benefit of electronic controls on brewing pressure is that brewing pressure becomes more consistent. For example, mechanical preinfusion rates are dependent on grind fineness and size distribution, as fines migration and overall particle size influence resistance to flow. for a given gicleur size, pressure force builds more quickly when grind particles are small and when fines migrate quickly to the bottom of the cake. For the example here, brew pressure at the cake and grind fineness are related parameters. They are decoupled when electronic controls are implemented. If pressure rise is dependent on a pre-programmed profile using a pressure transducer for feedback control, then you will produce the same profile regardless of particle size and the two parameter are now decoupled and pressure will be more consistent from shot to shot. Additionally, the machine designer now has the choice of adjusting pressure with respect to volumetric flow or with respect to time. There are reasonable arguments for both.

    New topic in my reply / rant – I quickly discovered that I very much liked the consistency of having controlled pressure decoupled from grind size. I also quickly learned that I could confuse myself mightily and I agree with James that it’s really easy to make lousy coffee thru lousy profiles. I think that the folks who are really gonna make hay on this are gonna be people that are very disciplined in their approach to testing and good tasters to boot. Andy Schecter deserves credit here for working on methods of correlating taste to mass. Another guy that folks ought to get tapped into is Jim Schulman. He’s passionate. I think he’s a good taster, and he’s a good statistician, which we sorely need. He can help design the right testing methodology that will minimize the required work.

    -Greg Scace

  23. I think that your second paragraph is right on the money.

  24. Not meaning to nitpick – and correct me if I am wrong here – but most temperature probes lack an accuracy of 0.1C. Equally I’ve never seen a machine hold temperature to within 0.2C for the duration of a shot. I’d be very surprised to see a bump of 0.1C on a PID controller make a significant or noticeable difference on the shot.

    I am currently of the opinion that the minimum required change to brew temp to be noticeable, and consistent outside of other areas of preparation is about 1C. By that I mean that rarely have I seen anyone capable of replication of other aspects of their brew recipe that would match the variance of 0.5C in brew water.

    I’m not saying people don’t have skills, more that brew temperature is perhaps a little overestimated these days because it is easy to control and easy to measure.

  25. Hi Scott,

    I don’t think saying that varying a coffee’s flavour in the cup using pressure – be it lateral or not – needs any disclaimers.

    I’d be very interested to hear how you feel that one can use pressure to specifically impact a certain characteristic in coffee. It would be a shame if any information disappeared into email and away from the public domain where it could be multiplied and improved upon.

    And are you saying then that the Aeropress does in fact make espresso?

  26. I hate to bring up measurement again – but it would be interesting to know if the profile that bumped acidity did so because it changes the extraction or produced a good looking underextraction?

  27. I played with that machine (a bit) at CoffeeFest Seattle last month, but I didn’t get much time to mess with the pressure profiling. I had a sales person trying to sell it to me for our cupping room. It’s got some VERY interesting aspects to it that I’d love to have a few days to mess around with.

  28. good point. When I get a mojo I’ll look at that.

  29. Hi James,
    You are correct that moving information out of the public sphere could be a disservice. I wanted to limit my input in order to avoid the perception that I am simply trying to promote a product. Instead I wish to only comment on a method of preparation.
    While I can describe my process for dialing in, there is a certain amount of the process which can only be described as searching. This same searching is also how I go about adjusting temperature in most cases (I taste the coffee and if it doesn’t taste “right” I adjust it). What I believe you are asking me to offer is a clear cause and effect of the pressure manipulations made, which I am unable to provide. Metaphorically that is to say I can shoot an arrow at a target and after seeing the arrow hit the target I can change how I shoot the arrow next time in order more accurately hit the bull’s-eye. However I cannot explain the physics of the arrows’ flight.
    In regards to your last comment on the definition of espresso I cited. After rereading my post it is apparent that my complete citation did not survive. “a brew obtained by percolation of hot water under pressure though a cake of roasted ground coffee, where the energy of the water pressure is spent within the cake.” I cited as the definition of espresso given in Illy, Andrea, and Rinantonio Viani. Espresso Coffee: the Chemistry of Quality. London: Academic, 1995. Print. Without commenting on if Aeropress produces “espresso” under any given definition, I simply wish to accept for the sake of this conversation that the beverage produced by a pressure profiling machine is in fact espresso.
    Thank you,
    Scott

  30. Funny that there has been no mention of texture in this discussion. Sorry to harp on about levers but it is the only area where I’ve had experience of variable pressure profiles. With all things being as equal as they ever can be on a manual lever machine, nailing the right pressure curve can give the velvetiest yogurt shot whilst retaining subtlety and detail the combintion of which I’ve never been able to match on a regular rotary pump machine.

    It may be more productive to aim for a quality such as this rather than endlessly looking for an elusive idea of how this coffee can taste. Maybe you want the coffee to taste a certain way, but maybe it doesn’t have it to give. How can you know? From cupping? If you enjoyed that flavour so much then relax with an immersion brew. There seems to be an awful lot of fighting the ingredients, when perhaps if one predicts and accepts that one’s equipment will always cololur the end product then you can finally say this is how this espresso tastes in this shop. I like it and I’m gonna stop messing.

  31. > Shots channel much more easily the slower the ramp up.

    That is the opposite of my experience.
    One thing I’ve found is that the slower the ramp, the faster the flow rate — once the pump comes up to normal pressure. The resulting faster flow may look like channeling, but it’s not.
    So, to maintain a “normal” flow rate, one has to grind a little finer when employing a slow rampup.

  32. Sure you’re quite right that the machine can’t hold temp within 0.2C over the duration of a shot, but surely raising the boiler temp, even by 0.1 degree, will raise the average temp of the shot.

    Simple reality is, I’ve never gotten a hold of a probe to measure the brew temp at the puck, but I know that if my boiler temp is even 0.1 or 0.2 below the ideal, my shots will taste more almondy and lose balance, 0.3 or 0.4 of a degree and they’ll become noticably over-acidic. Likewise if I’m a little hotter than ideal there’s just a little sharpness, half a degree hotter and the shots will become really astringent.
    Maybe it’s just a weird thing about the blend I’m using, that it has a very tight sweet spot temp-wise (though the blend does have multiple sweet spots, and I’ve noticed a similar margin for error around each of them).

    Don’t know if that sounds a little mad to you, but I’ve gotten to know my blend very well. As an interesting caveat, there have been a few days now and then where I’ve forgotten to change the temp, and the reason I’ve been alerted to it is that regulars have commented that something tastes different.
    I certainly don’t claim to be an amazing barista with mad skills, just a passionate guy doin my best…

  33. This pressure profiling thread reminds me of a conversation I had with Peter Dupont (Coffee Collective) about brewing espresso using coffee that was roasted for conventional brewing i.e. French press, Aero press, Melitta, etc. We were using the lighter roast for the purpose of making Americanos with the crema skimmed off (best I’ve ever tasted). Suffice to say to execute this properly the flow of the espresso needed to be slowed using a finer grind than I might normally use for a normal 2oz espresso. The theory that Peter had (or at least what I understood of it) was that the lighter roasted coffee required more energy or more contact time with the water to break through the cell walls to extract properly, or at least to get what we were going for, clean bright acidity with clarity.

    I bring this up because I was making these Americanos using a GB5, I never asked but I’m assuming it was outfitted with typical gicleurs from my experience using other GB5 with gicleurs, would have been made way easier with a machine with pressure profiling capabilities. This is only a theory of mine, but it would be my guess that I would be able to use a more normal dose and grind with a pressure profile machine rather than having to rely on grind fineness to slow the extraction. Being able to ramp the pressure up slower would allow the energy from the water to pass through the cell walls for a more complete extraction before being expelled into the demitasse.

    Reading Tim Windelboe’s blog post also makes me think that he had such an experience while extracting his Kenyan. His pressure profile never reaches 9 bars and his initial ramp up is long and slow. His extraction time, dose, and ending volume with his description of the end product (sounds delicious!) suggest that while lacking perhaps a little mouth feel (Tim correct me if I’m wrong) created a really balanced shot, which in my experience Kenyans are not usually known for.

    While I never excelled in Science I did try to pay attention. It seems to me that like temperature, pressure while brewing espresso is another means of applying energy on the coffee bed. Lower pressure = longer saturation time = more extracted from the coffee. Again this is only a theory but I’m guessing the inverse is also true.

    Here though is where I’m really taking a leap of faith and where I have serious questions about pressure profiling espresso machines. For me it comes to the question of what differentiates espresso from conventional coffee? The only significant differences I can think of are concentration and crema or emulsified oils. Emulsified oils created by hot water of varying temperature applied with pressure on a bed of coffee. My question is, at what point in bars does the pressure start to emulsify the oils in the coffee? Because I’m wondering if a slow long pressure ramp up might actually be brewing the coffee like conventional brewed coffee to a point and then later in the ramp up start emulsifying the oils from the coffee resulting in a more complete extraction in a more dense brighter coffee as Tim experienced. This might also explain the thinner mouth feel due to less emulsified oils in the cup.

    Of course I’m just throwing ideas around, but unlike Schomer (whose opinion I respect) I am really excited about the possibilities that pressure profilers might have on espresso, that my long-winded post can attest to. Though I would say putting a Strada or the like in front of a novice Barista might be a mistake.

  34. yeah I get that, and I think another way to think about the slayer is in terms of managing the grind size.

    Maybe a channeled shot is just more obvious with the slow ramp up.

  35. sorry for the off-topic: what do i read? there is a synesso here in vienna? where does it stand, in a bar/cafe and is it possible to take a deeper look at it?

  36. Thanks for posting the PP “guide”, James. Good to know people are actually reading it. The graphic was intended to help guide people who have not used a PP machine before to get a basic understanding of the effects of PP on espresso. It’s also meant to be vague/generalized enough so that we DON”T get too specific on how certain PP curves have direct, consequential affects on the shot. The fact is that, based on our PP usage in the past year, we feel it is too bold to make such a statement. While I can say certain PPs work a certain way on this coffee, it may not necessarily work on another different coffee. This is particularly true for blends, as different component coffees may react differently. The “guide” is meant to open up people’s minds on what’s possible with PP, not to serve as a rulebook. The hope is to have people use PP if they so choose and let them dictate what’s best for their coffees.

    I cannot tell you the chemistry that occurs inside the coffee under different PP, or whether certain PP that bumps up acidity is really a change in extraction or an underextraction. Even with my background in Food Science, I’m more comfortable to refer such a question to a qualified food chemist or chemical engineer. However, if I were to postulate (to humor myself), my guess is that lower pressure in the beginning of a shot extracts the more subtle, volatile components of the coffee, without the less volatile, more dominate ones that come out at 9 bar. This results in brighter, fruitier notes. Lower pressure at the end of a shot brings out more oils but leaves the spent, bitter parts of the coffee in the puck, resulting in more body.

    As for PP as another new feature to sell more machines or to fulfill people’s interest, I’d point out that PP has been there all along. Levers, paddle GS/2, gigleurs, and bypass valves have all been used in the past to manipulate pressure during brewing. Gigleurs of varying sizes were hot commodities earlier this decade as a means to modulate water pressure introduction to the puck. Although you’re not directly controlling the pressure, you’re, in fact, pressure profiling (same goes for preinfusion on a Synesso). In that sense, PP isn’t all that different than using gigleurs or preinfusion, except that it gives you more control over when and how much.

  37. I can understand manufacturer’s reluctance to make explicit claims about the effects of a certain aspect of profiling. However, it would be really good if people started making statements for the rest of us to test. Once we start testing them then we learn something, whether it turns out be to a universal rule, or whether it works for certain coffees.

    You’ve given us two worth testing:

    - Lower pressure at the start of the shot results in increased aromatic complexity
    - Lower pressure at the end of the shot yields increased body and reduced bitterness

    While we still need to define low pressure – be it sub 5 bars at the start or sub 9 bars at the end – I think we have some things that people can experiment with and report back. Even if they are totally wrong, it doesn’t matter because we’ll know more than we did before.

  38. johanna wechselberger runs a vienna coffee school.
    robert (berlin)

  39. > I would like to add that the first generation of espresso machines
    > brewed at a pressure around 1.5B. For the purpose of this conversation
    > I wish to stipulate that espresso can be defined as “a brew obtained
    > by percolation of hot water under pressure though a cake of roasted
    > ground coffee, where the energy of the water pressure is spent within
    > the cake.”

    Scott, I wasn’t implying that coffee made on a PP machine wasn’t “espresso.” And my comment certainly wasn’t a put down of the Strada; instead, you and the others at LM deserve congratulations for bringing a programmable PP machine to market. Was I WAS saying is that everything affects everything else, and that spending a significant portion of the extraction time at pressures much lower than 9 bar seems to result in a lighter mouthfeel. At least that’s the way I experience it.

  40. Yes, we’ll have it back in 1-2 weeks…it was totally ruined in Italy and had to be repaired in Hamburg. It’s in our Vienna School of Coffee. On Fridays and Saturdays we are open for everybody, so you can come and play ;-) If you need more info email me … office@viennaschoolofcoffee.at
    cheers
    joey

  41. P/P Espresso Machines to me is a matter of having the ability to adjust-on-the fly each and every variable for fine tuning, rather than fixing what is not already broken. Its not a means to an end?? As an analogy, think back to the carburetor days when mechanics were tuning cars with just spanner and spacers (metaphorically, aka Lever Machines, mostly mechanical based), and as time went by Engine ECU’s became more and more complicated and flexible in terms of fine-tuning down to the last 100-250 RPM range, whether it be related to Advanced/Retardation of Ignition Timing, Air Fuel Ratios (aka Temperature, Flow or Pressure Profiling in the coffee world), also having enough feedback sensors like Air Flow Meters or MAP sensors (PID’ing in coffee?) collaborating with the pre-programmed ECU Maps to ensure there’s some sort of real-time feedback and thereby automatic adjustments under closed-circuit conditions.

    Food for thought: putting 100 RON fuel into a car nowadays gives it more power or economy because the ECU can sense its higher energy content and adjust the settings to suit. When was the last time we have seen an Espresso Machine being able to tune itself to say make pre-programmed shots to say exactly 92.8 Degrees, 20g Dosage, 5 secs pre-infusion at 3 Bars, 3 secs at 5 Bars, then pour 39ml in exactly 40.5 seconds each and every time for instance. These variables are still controlled manually by Grind/Dosage/Tamp/Basket as an open-loop execution by the Barista. Now that we have Programmable Electronic Dosers as well – when is anyone going to start making a very dependable Swift Grinder with a proper dependable basket and repeatable robotic Tamp Pressure to suit?? The same theory can also apply to roasting and its equipment, but to a lesser extent since you need to rely much more on instinct.

    It might be stretching it as a comparison, but fundamentally a Mini Cooper from the 60′s isn’t really that much different from a current BMW Mini Cooper S, they still have the same spirit and DNA, its just that one is much more sophisticated and carries tune-ability down to the variable/cam shaft timing or whatnot.

    Ultimately, what I am trying to suggest is that a Lever Machine might as well have ‘fluked’ it all just by chance back then, due to its convenient non-linear extraction nature compared to the earlier generation fixed pressure, fixed temp/flow electronic pump Espresso Machines. But even if you invest a lot more money into inventing the up-coming high-tech, infinitely tunable Pressure/Temp/Flow Profiling machines, at the end of the day your final parameters still aren’t going to be that much different from the old-school Lever’s or the trusty Semi-Autos, because THOSE EXTRACTION PARAMETER HAVE ALREADY PROVEN TO BE MORE OR LESS CORRECT OVER THE YEARS, DON’T DEVIATE TOO MUCH FROM IT FOR THE SAKE OF IT..

    So its not really about opening a new can of worms in my opinion, its more about versatility and the ability to fine-tune every single minute point of the extraction in the end, not because you necessarily require it, but because you know you can access it only if its utterly necessary. And just to quote back on some Engineers I heard who commented on the current science behind the latest Espresso Machines – the current batches are still far away from its theoretical limitations in terms of an engineering feat. Its entirely plausible that a Super-Duper-Auto machine with Built-In Grinder and Cup Warmer of sorts with a full loop-back, close circuitry system and enough sensors linking back to the computer to be created and make the same tasting shot (or milk stretching) nearly everytime to the Barista’s pre-determined program, for that particular batch of coffee beans on that day, even taking into consideration the high sensitivity of coffee extractions. And that’ll just operate like an Auto-Pilot: boring.

    But when does it all start becoming too excessive and entering the realm of diminishing marginal returns? For instance, Gwilym’s shot for me, pulled on his manual Lever Machine was drop dead gorgeous and better than anything else I’ve ever experienced, but I don’t see much on-the fly adjustments on his beautiful Arduino except for what he can manually control within certain limitations.

    One are’t even allowed to adjust the Water Temp on the WBC machine during the competition. Isn’t there some sort of conflict within there already? Will relaxing the rules and letting the competitors adjust their Temperature and Pressures on the fly during the competition, end up having someone out there discovering something we haven’t already explored yet in the end?

    HK.

  42. Sorry to be late to the party, but I wanted to add a comment. There are a lot of Synesso and LM paddles already in services that at least have the capability to switch from line to pump and back to line. I would love to hear some data points on those, partly for selfish reasons (I own one), and partly because it might inspire their operators out in the real world to make use of them.

    As I’ve noted before, I almost always see baristas use these paddles as on/off switches, and I don’t think their capabilities have been well explored.

Submit a comment