A filter is not a general-purpose improver
The advice to filter your water is nearly always given without saying what for, which leaves the reader to assume that filtration is a single quality that water can have more or less of. It is not. Each technology targets particular things and passes everything else through unchanged, and the four properties that matter to tea — hardness, alkalinity, chlorine and total dissolved content — are affected quite differently by each one. Someone whose water is chlorinated and someone whose water is high in bicarbonate have different problems and need different equipment, and either of them buying the other’s device will conclude that filtering does nothing and that the whole subject is enthusiast noise. That conclusion is understandable and wrong, and it is reached often enough to have become a genuine belief in its own right. The productive order is to find out what your water is like first and choose the device second, which is the reverse of how these purchases usually happen — the jug is bought because it is in the supermarket, and the water is never looked up. Reading a supplier’s analysis costs nothing and takes a minute, and it turns a guess into a decision.
Activated carbon: chlorine, chloramine and taste
This is what almost every jug filter and most tap-mounted units are built around, and what it does well is adsorb chlorine, many organic compounds, and the substances responsible for musty or earthy off-tastes. For a great many households that is the whole problem, because chlorine is unpleasant in tea out of proportion to its concentration: it reacts with phenolic compounds in the leaf to produce notes that are present in neither on their own, which is why chlorinated water can make tea taste of something that is not obviously chlorine. Removing it is cheap and the improvement is immediate and obvious, which is why carbon filtration is the recommendation that most reliably satisfies people. What carbon does not do is remove calcium, magnesium or bicarbonate in any meaningful quantity. Water that was hard before a carbon filter is hard after it, and anyone expecting a jug to fix scale or alkalinity will be disappointed. Chloramine, which many suppliers now use instead of free chlorine because it persists further into the network, is more stubborn than chlorine and needs a larger carbon bed or a longer contact time. A small jug handles it less completely than the marketing implies, and standing the water uncovered — which works on free chlorine — barely touches it.
Ion exchange, and the thing a water softener does not do
Ion-exchange resins swap one ion for another rather than removing anything outright. A domestic water softener typically exchanges calcium and magnesium for sodium, which is why softened water stops depositing scale in the kettle and the boiler, and it is also why softened water is a poor choice for tea. The hardness ions have gone, but the bicarbonate has not: the alkalinity that dulls aroma, darkens liquor and encourages the metallic film on a cup of black tea survives the exchange entirely intact, and sodium has been added on top. The result is water that is soft by the plumbing definition and actively unhelpful by the brewing one, which is a genuinely counter-intuitive outcome and catches out people who have been told for years that soft water is better for tea. Filter jugs marketed for tea and coffee often use a mixed resin that targets carbonate as well as hardness, which is a materially different product from a household softener despite sharing the underlying technology. That distinction is almost never made clear on the packaging, and it is the one worth asking about, because the carbonate-targeting version is the only common domestic device that addresses alkalinity directly.
Reverse osmosis and distillation: removing nearly everything
These take out most of what is dissolved rather than a selected fraction, and the consequence is covered at length in the guide to pure water: what comes out is a blank rather than an improvement. As a domestic tea solution used on its own, reverse osmosis overshoots badly. It removes the alkalinity that was the problem and the mineral content that was doing useful work, in one undifferentiated step, with no control over the balance that remains — and the balance is what the whole subject is about. The tea that results is clean, thin and characteristically hollow, and the usual response is to blame the leaf. Where RO earns its place is as a base for water that is then remineralised deliberately, which is a real and quite small practice, or in regions where the supply is so heavily mineralised that nothing short of near-total removal helps. Both situations exist and neither is the common one. There is also a practical cost worth stating: domestic RO units discard a substantial volume of water for every volume they produce, and they need membrane changes on a schedule that is easy to forget.
Scale inhibitors, which are not filters at all
Polyphosphate cartridges, magnetic and electronic devices, and the various in-line units sold under names like conditioner or descaler are stocked alongside filters and do something quite different from any of them. At best they change how dissolved calcium behaves as it comes out of solution, so that it forms a loose sludge rather than a hard adherent scale on a heating element. The mineral content of the water is unchanged, and so is everything that content does to tea. This is genuinely useful for a kettle, a boiler or a commercial machine, where scale is an equipment problem with a maintenance cost, and it is entirely irrelevant to the taste of the cup. The two purposes get conflated constantly in retail, partly because the marketing language for both is about water being better. If the goal is a kettle that stays clean, these have a case and should be judged on that. If the goal is a better cup, they are not addressing it, and no amount of them will. The claims made for the magnetic and electronic varieties in particular are contested even on their own terms, and TeaHQ takes no position beyond noting that the mechanism proposed has nothing to do with brewing.
The maintenance problem, which is bigger than the choice
An exhausted cartridge is worse than no cartridge. Carbon that has stopped adsorbing continues to pass water perfectly well, and the bed itself accumulates organic matter over time — a filter left in a jug for months in a warm kitchen can contribute more off-flavour than the unfiltered water ever did, and the failure is silent because nothing about the jug changes. Ion-exchange resins likewise stop working and the water simply reverts, usually so gradually that nobody connects the change in the tea to the filter. Manufacturers state a volume or a period, and both are conservative estimates against an assumed household average rather than a measurement of your own usage; a household making a great deal of tea will exhaust a cartridge well before the calendar says so. The practical habit is to change filters early rather than late, run the first litre or two to waste after fitting a new one to clear loose carbon fines, and keep the jug refrigerated and emptied rather than standing warm and half full for days. None of this is expensive, and all of it matters more than which brand of jug was chosen.
Matching the device to the fault
If the tea tastes of the swimming pool, or has a chemical edge that is present in the water on its own, a carbon filter is the answer and it is the cheapest intervention available. If the liquor is dark, dull and flat, with a film on the surface of black tea and aroma that seems muted regardless of the leaf, the problem is alkalinity and carbon will do nothing at all — a carbonate-targeting resin, bottled water chosen from its analysis, or a partial dilution with pure water is the direction. If the tea is thin, sharp and hollow, the water may already be too pure and the answer is not to filter it further but to stop. And if the fault appears only with delicate green and white teas while black tea is fine, that pattern is itself diagnostic: it points at the water rather than at technique, because robust teas hide water faults that delicate ones expose immediately. Diagnosing from the pattern of which teas fail is quicker than diagnosing from the analysis, and the two together usually settle it.