One of the site's first how-to articles is Perfecting the Naked Extraction
, prompted by the proliferation of bottomless portfilters. Verifying that an extraction looked
even became the first step in problem diagnose (similarly, the how-to article Weiss Distribution Technique
(WDT) became the de facto way of assuring even distribution!).
As part of the review process, every writeup includes photos of a few bottomless pours. Most of them are "barista eye candy", but I thought for this review, I'd take a more critical look to see if it revealed a difference between the no-fuss, automatic La Marzocco Swift Mini and the other end of the grinder spectrum, the Kafatek Monolith Flat, which is a single-dose grinder known for consistent performance.
The prior post shows the video from the two extractions, first from the Swift Mini and then from the Monolith. Apologies in advance for the unevenness of the lighting; I recorded the Monolith extraction later in the day and the sun was setting. I had a bright studio light illuminating the business end of the portafilter, but the background wasn't illuminated, hence the exposure and color tone differences.
For this closer look, I captured a few key frames from the video above and eliminated the color differences, since for this comparison, I'm only interested in the evenness.
Below are the extraction parameters:
- La Marzocco 17 gram basket, dosed accordingly
- 9 seconds of preinfusion at 3 bar, then 9 bar for total of approximately 34 seconds
- Brew ratio of 50% (17 grams in, 34 grams out)
- Coffee: Caffe Lusso's Gran Miscela Carmo Espresso Blend
The Swift Mini was dialed in by adjusting the dose/grind with the top dial for grind setting and bottom dial for dose. After that, it was as simple as lock in the portafilter, press a button, go. Since I wanted to double-check the weight, I used a second portafilter for the grinder and removed the basket retention clips. That way, I could prepare the basket and confirm the Swift Mini's dose weight.
The Monolith was dialed in with a manual movement of the grind chamber barrel after loosening the grind clamp; the coffee dose was pre-weighed and the grinder run until empty. To prepare an evenly distributed coffee bed, I followed dosing with the WDT. As an extra precaution, I followed that by a quick leveling with a coffee distribution tool (adjusted to the correct puck depth), then a final tamp.
Initial beading - how even does the extraction appear?
While nearly all "espresso porn" photos feature the fully formed cone of the extraction, I believe the initial beading is far more indicative of evenness. If the initial beading is slightly
off, puck compression may close off early fissures, but if the beading is significantly uneven, channeling is almost certain to follow.
The two photos below capture the point at which approximately half the basket's holes have shown signs of beading and half have not:
La Marzocco Swift Mini - initial beading
Kafatek Monolith Flat - initial beading
While these are not perfect evenness, they're both very good. The Swift Mini does appear to have less initial beading directly beneath the center hub of the distribution "fan" (see A puck's view of the the Swift Mini
photo earlier in this thread). Despite my careful attention to distribution, the initial beading depicted in the second photo of the Monolith extraction shows an ever-so-tiny sign of unevenness.
That said, we're looking at fleeting moment very early in the extraction. Minor differences like this can (and arguably often do) mean nothing.
Fully formed stream - now how even does the extraction appear?
Given the inconclusive results above, the next two photos move to later in the extractions. This time, I captured the frame where the first formed stream falls. Presumably any initial fissures have closed (or not) by this point and the signs of channeling, if it's to occur later, should present themselves.
La Marzocco Swift Mini - initial stream
Kafatek Monolith Flat - initial stream
Whoa boy, that's really, really
close. It's worth noting that the location of the stream can be misleading; it's just as likely to indicate the level of the grouphead as channeling (i.e., in this case, the stream is closer to the camera and slightly to the right, suggesting the grouphead may be tilted in that direction). I haven't bothered checking the level, but maybe I should?
Past the midway point - now how even does the extraction appear?
Most of the "showoff" photos of espresso extractions are snapped around the midpoint where most signs of unevenness are hidden under a large single cone-shaped stream and the so-called striping is at its peak. These photos were captured around the 10 second mark before the pump was turned off. It's slightly after the favored showoff point, but before the extraction has entered the timeframe where the pour speed increases.
La Marzocco Swift Mini - final moments of coherent stream
Kafatek Monolith Flat - final moments of coherent stream
These last two photos demonstrate my earlier point about the diagnostic value of early versus late observations of espresso extraction pours: Basing comparisons on the last half of the extraction is likely to end in a photo finish; if not, it's likely a case of gross distribution errors that led to significant channeling (i.e., the provenance of beginner baristas).
Presented with the two videos above, I would judge the second one from the Monolith every-so-slightly
more even based on some signs of the first video from the Swift Mini initially favoring the perimeter. This is consistent with my observation that with careful attention to puck preparation, the WDT/Monolith combined with gentle preinfusion from the La Marzocco Strada is unbeatable, if your goal is producing brag-worthy looking espresso pours. Of course, if your goal is producing the best tasting
espresso, as our group blind taste tests demonstrated, it's a lot harder to choose a winner.
I've had the La Marzocco Swift Mini for a little over three months, testing it day-to-day, in a small group taste test, and a large espresso-loving crowd. Before packing it up and forwarding it to the next reviewer, I would like to compare the same two grinders, this time with a light(er) roasted coffee versus my usual medium roasted coffee selections. Look for this in my next and final installment.
PS: If you have coffee suggestions for this last test, please contact me offline