Grind Consistency and Taste

What grind consistency actually means

A grinder does not turn beans into identical fragments. It produces a particle size distribution, meaning a mixture of sizes and the relative amount of each.

Within that mixture, fines are the smallest fragments, while boulders is an informal name for unusually large pieces. Neither term has a universal boundary across every brewer and measurement method. One espresso study defined fines as particles smaller than 100 micrometres, but that research definition is not a universal brewing standard.

Two samples can have the same average size yet behave differently. One might contain mostly medium particles; another might combine abundant dust with conspicuous chunks. Calling both “medium grind” hides a useful distinction.

There is also a difference between consistency within a dose and repeatability between doses. A grinder that produces the same distribution each morning is repeatable, even if that distribution contains a substantial fine fraction.

For the wider relationship between grinder settings and brewing, start with grind size and extraction. Here, the question is what the mixture of particles does to the cup.

Why different particles extract differently

Extraction is the movement of soluble material from roasted coffee into water. Particle size influences both the accessible surface and the distance dissolved substances must travel from inside a fragment.

Smaller particles generally offer more surface relative to their volume and shorter internal travel distances. Larger particles have longer routes from their interior to the surrounding water. Under otherwise comparable conditions, these differences can produce unequal extraction.

That does not mean every fine particle becomes unpleasant while every large fragment remains sour. Extraction depends on water access, contact time, temperature and the surrounding brew, not size alone.

In a flowing coffee bed, the problem becomes more complicated. Small fragments can occupy spaces between larger ones, changing the bed’s permeability, or how easily water passes through it. Altering the distribution therefore changes both the particles being extracted and the conditions under which extraction happens.

This is why “grind finer” can improve one brew but make another less predictable.

Fines and bitterness: a useful warning, not a diagnosis

The simple explanation runs like this: fines extract quickly, so they over-extract and make coffee bitter. The first part has a sound physical basis; the second jumps too quickly from extraction behaviour to flavour.

Controlled sensory research on drip coffee shows that bitterness, sourness, sweetness and other attributes respond to brew strength, extraction yield and roast. There is no reliable one-to-one translation from particle size to a single taste.

A fines-rich grind may contribute to a harsh cup through several routes. It may alter extraction, restrict flow or make an existing brewing problem harder to control. Yet fines are not an ingredient that can be assigned a fixed bitter flavour independently of the recipe.

They are also a normal part of espresso grounds. Removing them changes the coffee bed rather than simply removing something undesirable.

The practical rule is to investigate fines when bitterness arrives alongside a marked change in flow or repeatability. Do not condemn them merely because the cup tastes bitter. The distinction between flavour symptoms and extraction is explored further in over-extraction versus under-extraction.

Boulders and sourness: the other oversimplification

Large fragments can extract more slowly, leaving a brew insufficiently developed under a recipe that gives them too little opportunity to release soluble material. A cup that tastes sharp and thin may therefore justify investigating an excessively coarse grind or an unusually broad distribution.

But sourness is not proof of boulders. Controlled sensory work shows that perceived sourness also depends on strength, extraction and roast. A taste description cannot reveal the particle distribution by itself.

Uneven extraction can further confuse diagnosis. Different parts of a coffee bed may receive different amounts of flowing water, while differently sized particles respond at different rates. A cup can consequently be difficult to balance without fitting neatly into “too coarse” or “too fine”.

Before changing several settings, check whether the result repeats. A single disappointing cup is a weak basis for judging a grinder.

Why the brewer changes the effect

Particle distribution matters across brewing methods, but the mechanism and the useful response differ.

Pour-over and fines migration

Fines migration means small fragments moving within the coffee bed. Accumulation near a filtering boundary is a plausible contributor to restricted drainage, often called clogging or stalling.

The evidence needs careful wording. Espresso modelling supports a mechanism in which moving fines become trapped and increase resistance. That does not establish that every slow pour-over has the same cause, or that a muddy-looking surface proves migration has occurred.

A useful diagnostic trial is to keep the grind unchanged and use gentler, more repeatable agitation. Treat any improvement as a clue about your brewing system, not as proof that you have measured fines movement.

Espresso and channelling

In espresso, grinding finer does not always increase extraction. A study combining mathematical modelling with experiments found lower extraction at very fine settings than the simple uniform-flow model predicted.

The researchers interpreted this as evidence of uneven flow. Channelling, where water favours some routes through the puck, can leave other regions less effectively contacted.

Do not keep grinding finer solely because a shot tastes insufficiently extracted. Look at distribution and preparation as well as the grinder. Our guide to extraction time and grind size explains why the clock cannot diagnose the whole shot.

Immersion brewing

In immersion, the grounds steep in a shared body of water rather than relying throughout on fresh water flowing through a packed bed. Size still affects extraction, but an espresso-style flow explanation cannot simply be transferred unchanged.

Research on immersion cold brewing found faster extraction with smaller grounds under the tested conditions. The authors also discussed contrasting results from brewing in filter bags, where confinement can restrict movement and mass transfer. Equipment and particle distribution interact.

What burr quality can and cannot tell you

Burrs are the cutting and crushing surfaces inside a burr grinder. Their design and installation help determine the distribution produced, but a “burr grinder” label does not promise identical particles.

Manufacturer guidance supports checking cleanliness and correct seating. Mazzer specifies clean burr support surfaces when fitting replacements; Baratza’s operating instructions likewise describe correct positioning and routine cleaning.

For an unexpectedly inconsistent grind:

1. Unplug the grinder before opening or cleaning it.
2. Follow its manual to remove accessible retained grounds.
3. Check that removable burr components are seated correctly.
4. Ask the manufacturer about suspected damage or wear rather than improvising an alignment procedure.

Burr quality is best judged by controlled brewing results and relevant measurements, not by appearance alone. Flat or conical geometry, burr material and size are incomplete descriptions of a whole grinder. Without comparable evidence, none should be treated as a universal guarantee of better flavour.

Sifting: a useful experiment with trade-offs

Sifting separates grounds according to which sieve openings they pass through. A stacked system can retain large fragments above one screen, collect an intermediate fraction and allow smaller material into a lower chamber.

That makes it useful for seeing what “inconsistent” means in a particular sample. However, removing particles changes the distribution and may change drainage as well as extraction. A sifted brew is not simply the original brew with its defects removed.

Try this comparison:

1. Keep a control. Reserve unsifted grounds from the same coffee and grinder setting.
2. Record the separation. Note the sieve openings and use the same shaking routine whenever repeating the test.
3. Weigh the retained brewing dose. Do not accidentally compare a full unsifted dose with a smaller sifted one.
4. Hold the recipe steady. Keep coffee and water masses, water source, temperature and brewing routine unchanged.
5. Compare flow and flavour separately. Record drainage behaviour, then taste both coffees at similar drinking temperatures.

Repeat before deciding. If sifting helps, the useful finding is that changing this distribution improved this recipe. It does not establish that every fine particle should be removed.

Sifting also adds handling and leaves separated coffee to deal with. It is therefore better treated as an investigative tool than a compulsory daily step.

What the evidence supports

The strongest evidence concerns physical behaviour: particle distributions influence extraction kinetics and resistance to water flow. Mathematical models explain possible mechanisms, while brewing experiments test whether those mechanisms fit measured results.

Evidence for a universally tastier distribution is weaker. In the 2024 espresso fines study, adding fines chiefly changed permeability within the tested conditions; their extra surface area had only a marginal additional effect on extraction efficiency. Its sensory assessment was explicitly not a systematic double-blind test.

That finding does not contradict the principle that smaller particles can extract faster. It shows that, in that flowing system, the indirect effect through brewing time and permeability dominated.

The sensible aim is therefore a repeatable distribution that suits the brewer and produces the flavour you prefer, not perfect uniformity at any cost. Judge grind consistency and taste together, while holding the rest of the recipe steady.