I had time to set up a little quasi experiment with the filters. Started with one cupping brew and then split it evenly through two V60s, one w/ T90 and one w/ Abaca. Marked the cups on the underside and tasted blind. This is a Colombian castillo natural microlot with a prominent artificial candy note (think Jolly Rancher or Kool-Aid).
Result: No discernable difference between them. That said, they were absolutely disgusting. Immersion brews are pretty much always a mess for the types of coffees I drink. Reminds me of Campbellās vegetable soup broth. Heavy and soupy mouthfeel with muddled āearthyā flavors. The wild candied note was barely perceptible to the degree that if you didnāt know it was there, you wouldnāt pick it out in this brew.
Anyway, my guess is that if this immersion is hitting around 21% EY, Iām intentionally coming in below that on filter. Likely somewhere in the 18% to 20% range and possibly lower depending on the coffee. Thatās basically why I only listen to Lance and Patrik. They were the only two influencers Iām aware of who were saying āa lot of these coffees taste best at 18% EYā and provided measurements during the industry fever pitch of higher and higher extractions.
Caffe DāBolla: uhā¦ this is some guy running a one-man roasting/cafe operation out of like some spare space in an office tower. He seems like he knows what heās doing but dude, I walk in, nobody is in the place, itās maybe 11 AM, I ask for āespresso for hereā and heās like wellā¦ and hands me the menu and Iām āuh, is this in DOLLARS? Am I reading this correctly?ā
Anyway, I end up getting his cheapest espresso for ELEVEN US DOLLARS (thanks trump) and itās ā¦ ok. Nothing amazing. The space was nice considering how cramped it was, heās got nice drinkware, but dude was just really offputting. Somehow heās been in business for years (he was in a different location last time I was in town apparently). Oh and he doesnāt open until 10AM, C-.
Cupla Coffee: oof. Shot was BAD. Basically undrinkable, extremely sour, no crema at all. Oatmeal was terrible (thin layer of edible oatmeal at the very bottom covered by four inches of slop and milk foam). Thereās some big teener vollyball tournament in town and they were swamped so I guess Iāll cut them a little slackā¦ D+
FYI : If you try cold blooms, consider grinding finer.
I brewed another cold bloom at 122 F (50 C), this time with a washed Mexico chiapas. Itās about as sour and hollow as the others Iāve tried previously and Iām using higher brew temps and maximum agitation from pouring and swirling. My grind for this coffee is probbaly between 750-800 um. Would likely need to hit 600 um range and possibly lower since agitation capability will be limited by clogging. Thatās a good use-case for T90 which is the most resistant to clogging (after Sibarist), but Iāve yet to make a great cup of this coffee with T90 using regular techniques. YMMV
*Ultralight washed coffees might not be the best candidate for cold blooms. Light-medium naturals and/or heavily-processed coffees that are easier to extract might be better in the sense that if your normal brewing technique comes out too extracted, switching to cold bloom might drop it into the ideal extraction zone.
I went ahead and ran it back with the much finer grind and T90. Iām guessing itās in the 500-600 um range. The cup isnāt underextracted but itās also not great and maybe slightly overextracted. Crystal clear (almost watery) compared to a normal brew but it just tastes like regular coffee without the interesting stuff. Anyway Iām closing the book on this method for now. In addition to grinding finer than usual, maybe try using a slightly lower ratio. If anyone gets a good result with it please let us know.
cliffnotes are theyāre saying you can get the same amount of TDS with 10% less coffee if youā¦ pour different? like, the idea of changing the grind size isnāt on the table here? and to get the same TDS from a smaller mass of coffee, donāt you by definition have to overextract?
Hereās the best part: itās not actually different. JG wrote about this five years ago in the Ad Astra blog then formally published it with additional theory and evidence in his book on Jan 1, 2021 (which they cite). This paper uses a lot of the same jargon and theory that appears in Ch 6 of the book (e.g., laminar flow, water jet theory). The real difference I suppose is that this is a proper laboratory study using some serious equipment and facing peer review, but kind of wild seeing popular press pitching this as a new discovery when weāve been using it for a half decade.
And all of the stuff about reducing the amount of coffee used by 10% is complete cockamamie. Notice how there are no actual sensory tests included in the study, just anecdotes about how they tasted the coffees and this was all very interesting and whatever. The same type of claim has been made about the zero bypass brewers.
Using a gooseneck kettle, to maximize extraction** you should pour from the highest distance* above the coffee bed that allows an intact water stream. If your stream breaks (youāll know because it splashes and crackles on top of the slurry), this leads to a major reduction in agitation and therefore extraction. Pouring from too low of a height (i.e., laminar flow) also results in lower agitation.
*The maximum height you can pour from before breaking the stream depends on the flow rate, which you can control manually to some degree but is also dictated by the design of the kettle.
**Whether you should be trying to do this is a different issue entirely that depends on a lot of other factors.
I spent the past week or so drinking coffee from a K machine using a refillable pod and some pink bourbon I had ground in advance, and the result was not awful. It was almost as good as the specialty joint about 10 mins from where I was staying.
Update: Iām now using a third party stainless steel pod with little paper filters for my travel setup. You can use them without filters, but two things happen: (1) more oils and sludge, (2) the āpuckā is annoying to remove. These particular filters arenāt great, and I can tell because thereās still a tiny bit of sludge at the bottom of every cup, but itās better and more convenient for me than no filter. The grind size that worked best was just my normal pour over setting (850 um range).
I suppose you have a pretty limited selection of paper filters for something like this? You could probably 3D print something to shape a bigger filter to fit but that seems like more trouble than itās worth unless this was going to be a regular thing.
Yeah I didnāt see too many filters available and a Negotiator feels a bit overkill. The bottom of these things are irregular b/c thereās a cutout intruded inward to fit some bit on the machine. In other words itās not just a large thimble.
Btw if youāve never taken a real K-cup pod apart to examine, theyāre more interesting than youād imagine. It has a fairly nice built-in filter made of a structured paper formed to a slightly conical shape and away from the walls of the pod.
oh interesting, my inlaws have a reusable k-cup (theyāre only interested in Community Coffee Half Caf with Chicory, not sure why my brother in law bought them this since they make a full pot every morning, total fucking waste) but itās made of some sort of plastic mesh, and itās a fucking mess to clean it. I never really looked at it that closely.
I have to buy a bag of that shit every time the come visit, they act like Iām a fucking weirdo because I wonāt drink it (I mean, I am a weirdo but not because of that, I wouldnāt drink it even if I were a normie, my extremely normal wife who puts tons of milk in her coffee wonāt even drink it).
The most interesting thing about Maxwell House is that I was able to get some prominent notes to pop on a V60 pour that donāt show up in their drip machine. The bad news is that those notes are chewing tobacco and tire rubber.
I have two of these little metal pods so I guess thereās no reason why I couldnāt use one of them as the negotiator. Although Iām not sure which paper Iād try to cram into such a small space. The ridged Kalita style would probably work best.
