05 · Basic · Intermediate
Espresso Technique — Dosing, Distribution, Tamping, Machine
Puck preparation decides most of the result
Written by Chanho Hong
What this chapter covers
What dosing, distribution, levelling and tamping each control; the physical cause of channelling; and what pressure, temperature and pre-infusion actually change — with the pressure-and-crema relationship set out against the measured research.
Espresso is an extreme extraction in which 9 bar of pressure passes through a thin, tightly packed bed. For that reason, how evenly you prepare the puck governs most of the result. This chapter covers the three preparation steps (dosing, distribution, tamping) and the use of machine features.
Dosing — loading the ground coffee
Dosing is the act of placing the ground coffee coming out of the grinder into the basket. It has three aims: ① collect it cleanly, ② keep the density even, ③ deliver the same amount every time.
If the coffee piles up on one side at the dosing stage, or if the amount differs each time, the later distribution and tamping will not fully correct it. Uneven density is the seed of channelling. Good extraction begins with good dosing.
Load the same dose every time by weight (scale reading to 0.1 g), and reduce scatter and clumping by managing grinder retention and static (RDT: a drop of water on the beans). Catch the coffee so that it builds up evenly in the centre of the basket.
① The first suspect in shot variation. When time and flavour are erratic, look at dose variation first before suspecting the tamp. If the target is 18.0 g but the actual dose swings between 17.4 and 18.6 g, that alone makes extraction time jump considerably. Fixing the dose with a scale removes one variable.
② Managing static and scatter. In dry winter air and with light roasts, static and clumping make the grounds scatter and clot, and the density becomes uneven. Controlling static with RDT (a drop of water on the beans) makes dosing clean and reduces the seeds of channelling.
③ Correcting for grinder retention. When a lot of coffee is retained in the grinder, the freshness and the amount differ between the first shot and the next. Managing retention through single dosing or regular purging keeps the dose and freshness of every shot uniform.
Distribution · levelling — building a uniform puck
Distribution is the act of spreading the particles evenly inside the basket before tamping, while Levelling is the act of making the already distributed surface flat. The two are different — distribution deals with three-dimensional density, levelling with the two-dimensional surface.
Uneven distribution invites channelling — water funnels towards the less dense areas, so that region is over-extracted (bitter and astringent) while the rest is under-extracted (thin and sour), producing a mixed defect. A flat surface alone (levelling) is useless if the density beneath is uneven.
Use WDT (breaking up clumps with fine needles) to even out the density down through the bed, tidy the surface with a distributor or by tapping, then level. Order: dosing → (WDT) distribution → levelling → tamping.
"Channelling only happens when the grind is too fine" — false. Besides grind, there are several causes: poor distribution, a tilted tamp, a damaged filter, insufficient bloom and more. "Tamp evenly and there will be no channelling" — partly false. If the density beneath and the particle size distribution are poor, tamping alone cannot prevent it.
① Diagnosing channelling. If the flow runs faster and darker from one side only of the spout, that is channelling. In most cases it disappears once you even out the density beneath with WDT. Pressing the surface neatly (levelling) alone will not stop it recurring if the bed underneath is clumped, so fix distribution (3D) first.
② Reducing variation with tools. When each member of staff distributes by hand differently, use a distributor (spin tool) to standardise the surface angle and WDT to even out the bed, reducing the human variable. Once even distribution is standardised, shot quality becomes stable.
③ Isolating the cause. If channelling persists even with WDT and levelling, suspect not the distribution but the particle size distribution (excess fines), damage to the filter basket, or the tamping angle, in that order. Channelling has several causes, so narrow it down by ruling them out one at a time.
Tamping — compacting the puck
Tamping is the act of pressing and compacting coffee that has already been evenly distributed and levelled. The aim is to reduce the voids between particles so that water passes through the puck against a uniform resistance.
Being level matters more than the force of the press. A tilted tamp lowers the density on one side and creates a channel. Rather than 30 pounds versus 20 pounds, the key is level and consistent, every time.
Using a tamper matched to the basket diameter, keep it vertical and level and compact in a single press. Work from the arm and body weight, not the wrist. A calibrated tamper stops at a set pressure and reduces variation.
① A lopsided shot. If the flow runs off to one side, suspect the tamping angle. Check with a mirror or a spirit level that the tamper is not tilting, and aim the pressure not at "hard" but at "the same every time, and level".
② Removing pressure variation. When the force each person applies differs and the times fluctuate, fix the pressure variable with a calibrated tamper (which stops at a set pressure). The target is consistency, not force.
③ A strategy of minimising variables. Make all three preparation steps (dosing, distribution, tamping) consistent and the only variables left are grind and ratio, so flavour tuning is reduced to two dials. Good preparation is precisely a matter of securing freedom to tune.
Using machine features — pressure · temperature · pre-infusion
Once preparation is done, the machine finishes off the variables. A modern espresso machine can adjust pressure, temperature and pre-infusion, so different cups can be drawn from the same puck.
| Variable | Standard / range | What it changes |
|---|---|---|
| Pressure | usually 8–10 bar (WBC 8.5–9.5 bar) | Governs flow rate (speed). Too low and extraction falls short; too high and channelling and over-extraction become a risk |
| Temperature | WBC 90.5–96℃ | The balance of acidity and bitterness. Too high gives bitterness and burnt notes; too low gives sourness and underdevelopment |
| Pre-infusion / Pre-brew | wetting first at low pressure | Saturates and swells the puck slowly, giving less unevenness, channelling ↓, extraction ↑ |
When the coffee is wetted slowly at low pressure before the main pressure is applied, the puck swells evenly and fine differences in density are buffered. The result is less channelling and a more even rise in extraction — the same principle as the bloom in filter brewing.
A machine that varies the pressure during extraction (profiling) uses curves such as low pressure at the start (a gentle wetting) → main pressure → a decline at the end, to fine-tune acidity, body and cleanliness. Note, however, that if puck preparation is poor, no profile can rescue it.
Pressure and crema — the increase is real
Raise the pressure and the volume of crema really does increase. In a study comparing arabica espresso at 7, 9 and 11 bar, foam volume rose monotonically from 5.1 mL to 6.9 mL. In the same experiment, however, foam persistence barely changed, at 24.7–30.0 minutes, and sensory quality was best at 9 bar, while 11 bar in fact split the preference group. In summary — pressure ↑ → crema volume ↑, but stability stays the same and flavour does not rise with it.
The gas in crema is CO₂ trapped in the beans during roasting. The higher the pressure, the more CO₂ stays dissolved in the water during extraction, and at the moment it leaves the basket and the pressure drops abruptly to atmospheric, the supersaturated CO₂ bursts out as fine bubbles. On top of this, the strong shear created by high pressure breaks the bubbles up more finely. What holds the foam together is not the gas but the surface-active compounds of the melanoidin and protein families together with high-molecular-weight polysaccharides. That is why volume (pressure, CO₂) and stability (surface-active compounds) are governed by different factors, and why raising the pressure does not extend persistence.
One common explanation is wrong — the account that "high pressure emulsifies more lipids, so crema increases" has the evidence against it. Measurements show a clear negative correlation between fat content and foaming capacity. Lipids sit closer to the side of making less foam but longer-lasting foam.
Raising the pressure in order to get crema is a poor-value choice. For the same purpose, managing freshness is far more powerful — after grinding, more than 70% of the CO₂ is released immediately from 500 µm particles, so grind immediately before extraction. Set the pressure by flow rate and extraction uniformity, not by crema. Take 8.5–9.5 bar as the standard, and if channelling is frequent, look towards lowering it.
In addition, the volume of crema is not an indicator of flavour. There is research showing that the appearance of the crema changes the drinker's expectations, and that those expectations then pull the actual sensory evaluation along with them. Crema is what sells the shot before it is tasted, not evidence that the extraction went well.
"The higher the pressure, the richer and better the shot" — false. The volume of crema increases, but the flavour does not follow. Go beyond 9 bar and the water digs into the puck, raising the risk of channelling and over-extraction; in fact 11 bar was not preferred in sensory evaluation. The target for pressure is not "hard" but correct and consistent. More crema = a well-drawn shot likewise does not hold.
① A light roast tasting sour and sharp. Raise the temperature to the upper end (94–96℃) and give a longer pre-infusion to bring extraction up evenly. This trims the underdeveloped acidity and moves it towards sweetness and body.
② A dark roast tasting burnt and bitter. Lower the temperature and shorten pre-infusion to suppress over-extraction and burnt notes. Dark roasts with heavy lipid migration to the surface are particularly sensitive to temperature.
③ Handling channelling with light beans. Dense light roasts absorb water poorly and channel often. Wetting the puck slowly with pre-infusion (low-pressure pre-wetting) so that it swells evenly reduces channelling and raises extraction evenly. Machine variables are fine-adjustment tools to be used according to the degree of roast and the density of the beans.
Questions this chapter answers
- Does raising the pressure produce more crema?
- Yes. In a study comparing 7, 9 and 11 bar, foam volume rose from 5.1 mL to 6.9 mL. But persistence barely changed, and sensory quality was best at 9 bar. If more crema is the goal, freshness is a far stronger lever than pressure — the difference CO₂ content makes is about 4.8×, against pressure's 1.35×.
- Are distribution and levelling the same thing?
- No. Distribution is getting the grounds spread evenly through the basket, horizontally and vertically. Levelling is making the surface flat. A flat surface over a clumped interior still channels, so levelling on its own is not enough.
- Why does pre-infusion work?
- Wetting the coffee slowly at low pressure before full pressure arrives lets the puck swell evenly and cushions small differences in density. Channelling — water finding one path — drops as a result and extraction rises more evenly. It is the same principle as the bloom in filter brewing.
Figures and regulations are verified against a source hierarchy: Tier 1 in-house knowledge base → Tier 2 SCA·WBC and peer-reviewed literature → Tier 3 web. Anything not adequately supported is marked [to verify] in the text.