Espresso Masterclass #9 — Reading the Shot: Channelling, Tiger Stripes & Spent-Puck Analysis
By Wing Yuen
Every shot you pull is broadcasting a detailed report on its own internal state. Post nine is about learning to read it.
Across the first eight posts of this series we have built the espresso puck from the ground up — the physics that governs it, the numbers that describe it, the grind that forms it, the preparation that packs it, and the pressure and pre-infusion that drive water through it. All of that is work done before the shot begins. Once the pump starts, the bed is sealed inside a steel basket at nine atmospheres and you cannot touch it, adjust it, or see into it.
Except that you can. The extraction itself is an output signal, and the spent puck is a physical record. This post is about forensics: reading the stream while it runs, reading the cake afterwards, and pairing both with refractometry so that your diagnosis stops being a hunch and starts being a conclusion. It is the last analytical skill in the series before we assemble everything into a working practice in post ten.
The Masterclass at a Glance
Course: Espresso Alchemy — a data-driven masterclass for elite baristas
Format: Private, maximum 4 participants
Duration: One full day, 10:30–16:30
Location: Tasse Coffee Roastery, Takadanobaba, Tokyo
Languages: English / 日本語 / 中文
Price: ¥80,000 per person
Certification: Certificate of completion + digital resource folder
Chapter 1
The Naked Portafilter Is a Measuring Instrument, Not a Party Trick
Strip the spouts off a portafilter and you remove the last thing standing between you and the truth. A spouted basket merges every stream into one tidy pour, which is exactly why it is comfortable — it hides the geometry of the extraction. A bottomless basket shows you the underside of the puck in real time: where liquid first appears, how quickly the wetting front spreads across the mesh, whether the emerging droplets coalesce into a single centred rope or fight each other from opposite corners. Treat that view as an instrument reading rather than entertainment. Every shot gives you roughly twenty-five seconds of continuous, free, high-resolution data about the internal state of a puck you can no longer touch, and the barista who learns to read it stops guessing at what went wrong inside the basket.
What you’ll take away
- Why a spouted basket structurally conceals the faults you most need to see
- The three observation windows that matter: first drops, convergence, and the final third
- How to set up your line of sight and lighting so the stream is actually readable
Behind the shot
Position yourself so you are looking at the stream from the side and slightly below, against a light background. Most baristas watch from directly above, which is the one angle that flattens the cone and hides sideways jets entirely.
Chapter 2
The Anatomy of a Healthy Extraction — Establishing Your Baseline
You cannot recognise a fault until you know precisely what correct looks like. A well-prepared puck wets evenly, so first liquid appears across the whole basket face within a second or two of itself, then draws inward into a symmetrical cone that converges into a single rope roughly a centimetre or two below the basket. That rope should stay centred, stay continuous, and darken smoothly rather than flickering between shades. The convergence point is the single most informative feature of the whole shot: it is the physical summation of the flow field across the entire bed. When permeability is uniform, all the streams meet in one place. When it is not, the convergence point wanders, splits, or never forms at all. Pull twenty deliberately good shots and burn that image into memory before you go hunting for defects.
What you’ll take away
- The full timeline of a healthy shot from first wetting to shutdown
- Why the convergence point is the highest-information feature to watch
- How to build a personal visual baseline you can diagnose against
Behind the shot
Uniform first-wetting is a property of the puck you built thirty seconds earlier, not of the machine. If the basket face lights up unevenly, the diagnosis belongs to your distribution step, not your brew pressure.
Chapter 3
Channelling: How a Preferential Path Forms and Then Feeds Itself
Flow through a coffee bed follows Darcy’s law: the rate is set by the pressure gradient and the permeability of the medium, with measured espresso permeabilities landing in the range of roughly 10−13 to 10−14 m². The critical word is local. Permeability is not one number for the puck; it is a field, and water solves for the cheapest route through that field. A pocket left slightly less dense by a clump, a fissure opened by an off-axis tamp, a rim where the bed never sealed against the basket wall — each is a low-resistance path. Once flow concentrates there, the mechanism turns vicious: accelerated flow erodes particles and flushes fines out of the channel, permeability rises further, and still more of the water diverts into it. Channelling is not a single event but a runaway feedback loop, which is why it is so much easier to prevent than to correct mid-shot.
What you’ll take away
- Darcy’s law read correctly — permeability as a field, not a constant
- The self-reinforcing erosion loop that turns a flaw into a channel
- Why over-extracted and under-extracted coffee arrive in the same cup simultaneously
Behind the shot
A channel does not make coffee taste one thing. It delivers scorched, hollow liquid from the channel wall alongside sour, barely-touched liquid from the starved regions. The cup tastes muddled and thin at once — a signature no grind adjustment will fix.
Chapter 4
Tiger Stripes and Blonding — Reading Colour Without Being Fooled
This is where most self-taught diagnosis goes wrong, because two very different phenomena both make the stream lighter. Blonding is the whole stream lightening together, gradually, in the back half of the shot as soluble material is depleted — it is normal, expected, and your primary cue for when to stop. Tiger striping is the marbled dark-and-light pattern in the mid-phase, and it is genuinely ambiguous: gentle striping across a converged, symmetrical cone is unremarkable, while sharply defined pale streaks that appear early and stay fixed in one sector mark a region blonding ahead of its neighbours — a channel exhausting its local coffee. The discriminating question is never “is it light?” but “is the lightening uniform across the stream, and is it happening at the expected time?”
What you’ll take away
- The precise distinction between blonding, tiger striping, and channel blonding
- Using timing and symmetry — not brightness — as your decision criteria
- How to time your cut-off from colour without chasing phantom faults
Behind the shot
Localised pale streaking that is already visible in the first eight seconds is almost always a fault. The same appearance arriving at twenty seconds is usually just the shot ending. Same picture, opposite verdicts — timing is what separates them.
Chapter 5
Spritzing, Side Spray and the Edge Channel
Spray is channelling made loud. When a thin jet escapes the bed at high velocity instead of joining the converging cone, it fires sideways, hits the basket rim or the drip tray, and atomises. The direction is diagnostic. Spray from one consistent sector points to a distribution or tamp-level asymmetry in that sector. Spray from around the perimeter usually means the bed never sealed against the basket wall — classically from under-dosing, from a tamper narrower than the basket that leaves a compacted ring of loose grounds at the edge, or from knocking the portafilter side with the tamper and fracturing the puck margin. The insidious version is invisible: work on side channelling from poor tamper fit has been associated with extraction losses on the order of half a percentage point of yield, a difference you will not see in the stream at all and will only catch on a refractometer.
What you’ll take away
- Reading spray direction as a map back to the specific prep error
- Edge channelling, dose level, and tamper-to-basket fit as one linked problem
- Why some channelling is measurable but not visible — and what to do about it
Behind the shot
Before you blame technique, measure your basket and your tamper. A tamper even half a millimetre undersized leaves an uncompressed annulus at the wall that no amount of careful distribution will rescue.
Chapter 6
Spent-Puck Forensics — What the Cake Tells You Afterwards
The puck keeps a record of the shot, and knocking it straight into the bin throws away your most direct evidence. Turn it out onto a pale surface and read it deliberately. A healthy spent puck is an even, coherent cake, uniformly damp, releasing cleanly with a flat or gently domed face. A pit or crater marks where a jet from the dispersion screen or a low-density pocket let water punch through. A crack running across the face indicates the bed fractured under pressure — often from excessive tamping force on an unevenly distributed bed, or from a puck disturbed after tamping. A soft, soupy patch against a firm remainder is the clearest signature there is: that region carried far more than its share of the flow. Cross-reference what the cake shows against where you saw spray, and the location will usually agree.
What you’ll take away
- A systematic reading protocol for craters, cracks, soft patches, and rim damage
- Distinguishing screw-hole and screen artefacts from genuine channelling craters
- Triangulating puck evidence against your live stream observations
Behind the shot
Photograph the puck face, and note where the fault sat on the clock. Over a week of shots, a defect recurring in the same position stops being technique and starts being equipment — a worn screen, a warped basket, or a group head that is not level.
Chapter 7
Closing the Loop — Visual Diagnosis Verified by Numbers
Visual reading is fast, free and available on every shot, but it is qualitative and it has a floor: below a certain severity, channelling is real, costly, and completely invisible. Refractometry is slow and quantitative and has no such floor. Used together they are far more than the sum of their parts. The stream and the puck tell you where the fault is and what caused it; TDS and extraction yield tell you how much it cost and whether your correction actually worked. The discipline is simple and unforgiving: change one variable, pull the shot, record what you saw, measure what you got, and only then decide. That loop — observe, measure, adjust, re-measure — is what converts a barista who is good at espresso into one who can explain, repeat and teach every result they produce.
What you’ll take away
- Pairing qualitative observation with quantitative measurement in one workflow
- Detecting sub-visual channelling through unexplained yield deficits
- A one-variable-at-a-time protocol for confirming a fix rather than assuming it
Behind the shot
If your shot looks clean and tastes fine but the extraction yield sits stubbornly two points below where the same coffee performs elsewhere, suspect invisible edge channelling before you suspect the beans.
Who is this series for?
For the barista who has stopped believing that a bad shot is bad luck. If you have ever watched a portafilter spray across the drip tray and simply ground finer without knowing why, or thrown away a cracked puck without reading it, this post is aimed squarely at you. It suits working baristas responsible for consistency across a shift, cafe owners training staff to diagnose rather than guess, competitors who need repeatability under pressure, and serious home enthusiasts with good equipment who want their results to be explainable.
No physics background is assumed. What is assumed is a willingness to observe carefully, change one thing at a time, and let the measurement rather than the habit decide.
What’s next in the series
- #10 — Building a Data-Driven Espresso Practice at Home
Reading a shot properly is a skill built by repetition alongside someone who can tell you what you are looking at. In Espresso Alchemy, our one-day data-driven masterclass in Tokyo, we deliberately induce channelling, spray and fractured pucks on the bench, then diagnose and correct each one with a refractometer in hand — so you leave able to identify a fault, name its cause, and prove your fix with a number.
Espresso Masterclass is a 10-part educational series from Tasse Coffee Roastery exploring the science behind exceptional espresso, and the lead-in to our Espresso Alchemy data-driven masterclass in Takadanobaba, Tokyo. Post 9 of 10.