Disclosure: Some links on this site earn us a commission. It never changes what we recommend. Read the full disclosure.

The Density Report // Home Air

Your filter light counts hours, not air.

The listing tells you what the maker wants you to know. We go through what the people who paid report months later, find what keeps coming back, and name the source behind every conclusion.

Independent product research across home air, cleaning, auto and AI gear.

How far apart the maker’s own numbers sit

In every category we cover, one product carries a best-case and a worst-case published figure. This is the distance between them.

Auto 24×
Home Air 6×
Cleaning 6×
AI Gear 2×
Auto: a card rated for 5,000 hours of recording is a fourteen-year part at an hour of driving a day, and a seven-month part recording while parked. Home Air: a 2,000-hour filter counter is 83 days running continuously and over sixteen months at four hours a day. Both figures are the maker’s own.
Vendor documents first Specification sheets, manuals, warranties, pricing pages, terms.
Then the people who paid Owner reviews and support threads, read in volume, not sampled.
Every number traceable No estimate is printed. If it is not published, the page says so.
The Density Report

The same number, three different meanings

Every figure below is published by the maker. The only thing that changes is how the product gets used — and that alone moves the answer by six to twenty-four times.

Every figure above is published by the maker. Where a range is shown, both ends are the maker’s own — we do not print an estimate where the documentation gives none, and we do not rate or score any product on this page.

Featured this week

Four pages that show how we work: a number that means something different from what it looks like — and one that works the number out for you.

If you came here to size a unit rather than read a spec sheet, start with the published sizing table; the same table runs as a calculator if you would rather not read across a grid. The three methods behind it disagree by a factor of four on one room, and the pints on the label are a rate rather than a tank.

Latest

Three kinds of page, and each one says which it is: a reference answer, a set of buying criteria, or a full pass through the published record.

Research reviewHome Air

Sizing a Dehumidifier for a Crawl Space

The three published capacity cells, four disagreements between the two sizing methods, and one input no named source publishes.

Nov 2026
Research reviewHome Air

How long does a humidifier last? The number is in the warranty

No humidifier maker publishes an average lifespan. What two warranties do publish is a term: two years, reached either outright or by registering inside fourteen days. What the term excludes, what voids it, and the three records that decide whether it survives.

Oct 2026
Research reviewHome Air

Humidifier not working: three states, not one fault

A humidifier that will not work is usually one of three states: it refuses to start, it runs without making mist, or water ends up outside it. The check that separates them, why cleaning is on the list of causes, and the half of leaks that are not leaks.

Oct 2026
Research reviewHome Air

Air purifier smells: four places the smell can live

A purifier has four places a smell can come from and only two are fixed by cleaning. How to tell which state yours is in, what the CADR number cannot tell you, and why ozone is the wrong answer.

Oct 2026
Research reviewHome Air

How to clean an air purifier filter

Only the outer pre-filter can be washed. Here is which layer you are holding, the published 2 to 4 week interval, and what water does to the layer that cannot take it.

Oct 2026
Research reviewHome Air

Dehumidifier maintenance

Four colour bands with published boundaries, three functions behind one five-second gesture, and a full-tank warning that silences itself in sleep mode.

Oct 2026
Research reviewHome Air

White dust from a humidifier

White dust is the minerals in your water, flung into the air by ultrasonic humidifiers. Here is which units make it, what water stops it, and how to suppress it.

Oct 2026
Research reviewHome Air

How to clean a dehumidifier

One lid, two filters, opposite rules: the mesh washes, the particle filter is replaced. And the manual's three intervals all land in the same four-to-six-week window.

Oct 2026
Research reviewHome Air

Three manuals, three vinegar recipes for the same humidifier

Three humidifier manuals give three different vinegar concentrations for the same job. Your manual wins. Here are the recipes side by side, what they share, and what vinegar never fixes.

Oct 2026
Research reviewHome Air

Mold in a humidifier: the side you cannot reach

Mould in a humidifier keeps coming back from the side you cannot dismantle. Which parts come off, which ones do not, and what the maker says to do about the sealed path.

Oct 2026
Research reviewHome Air

How much does it cost to run a dehumidifier

The federal standard sets litres per kilowatt-hour, not watts. Invert it and the three legal size bands give $76, $123 and $99 a year — the middle one costs most.

Oct 2026
Research reviewHome Air

Dehumidifier not working

Six causes for "does not operate", five of them not faults. One defrost state that looks like a dead compressor. Five error codes, one of which keeps the unit running.

Oct 2026
Research reviewHome Air

Best air purifier for cigarette smoke: what actually works

A HEPA clears the haze; only activated carbon touches the smell. Here is what actually decides whether a purifier helps with smoke.

Oct 2026
Research reviewHome Air

Where should a dehumidifier be placed?

20 cm all round, 40 cm at the outlet, a 5–35 °C window, and a level floor that holds it full. The manual also names one room it will not allow — twice.

Oct 2026

Browse Home Air

One area for now — dehumidifiers and air treatment — with its own buying criteria and its own set of traps in the spec sheet.

How we work

Three rules we hold to on every page.

Real owners first

The starting point is what people who paid their own money say after months of use — not a press release, and not a launch-day video.

Patterns, not scores

One angry review is noise. The same failure reported by hundreds of owners over two years is a design fault, and it goes in the guide. We do not print star ratings we cannot stand behind.

We show our working

Every conclusion names where it came from, so you can check the reasoning instead of trusting the tone.

What this is: a research and curation project, run by one person. We do not sell anything, and we are not affiliated with any brand we write about. A page here is built by reading a product's own documentation and then going through what its owners report — and telling you which of those each conclusion rests on.