How we work

Start with the buying advice, then follow the links to check the evidence and the conditions that apply.

1. Begin with a real buying question

We define the model, region and specification variant first. Then we ask what would change the purchase: a missing function, an ongoing dependency, a practical compromise or a claim that needs checking. The number of positives and negatives is not a score.

2. Keep the evidence types separate

EvidenceWhat it can establishWhat it cannot establish
Manufacturer documentationPublished specifications, supported functions and the vendor’s guidance.Independent image quality, long-term reliability or successful operation on an untested unit.
Independent testAn observation or measurement under the tester’s stated conditions.The same outcome on every firmware, variant or installation.
Our own testA recorded observation with the device, software, setup and date disclosed.Anything outside the described test. Our current camera records come from published sources.
Owner reportA reported experience and a reason to investigate.A representative failure rate without a defined sample.

On model records, the first letter in a source marker identifies the evidence type. M is a manufacturer, I an independent test, O an owner report, R a retailer and G a Gear On Record test. The remaining code identifies the entry in the source list.

3. Connect the claim to the source

Links beside the claims and in the dependency tables identify the supporting evidence; the surrounding text states its conditions and limits. A product page is not a performance test. A forum discussion is not proof of a widespread defect. If documents disagree, we preserve the disagreement and explain what the buyer should verify.

4. Show what is missing

A source check does not become a hands-on review by being detailed. Each record states its scope and evidence gaps. We do not generate star scores, invented test results or percentages from public complaints. The camera images on our model records are editorial illustrations based on manufacturer views.

What an outage means for a camera

Internet loss means the outside connection is unavailable while the local router, network and camera remain powered. Local-network loss and power loss are separate scenarios. A camera may record to a card while an app function still depends on a connection. The checklist below describes a future device test.

  • Record the hardware version, camera software, app version, storage and whether the account is signed in.
  • Check local recording, live view, earlier and new playback, detection and notifications separately.
  • Repeat after a camera restart while the internet remains disconnected.
  • Check recovery, recording gaps and timestamps after reconnection.
  • Use Works, Fails or Conditional only for an identified observation or explicit source. Use Not tested for missing observations.

Current status: the launch camera pages describe manufacturer guidance and unverified scenarios. They are not completed outage experiments. A future physical test must publish its setup and observation record before it changes that status.

Our first planned device check is a C210 cold restart with the internet still down. Once we have a unit, we will note its hardware, firmware, app and card. We will run the check twice on a working local network with the internet disconnected and report recording recovery, local playback and timestamp accuracy separately.

Updates & corrections

The source-check date shows when we last read the linked evidence. It is not a promise that the retail product, firmware or price is unchanged. Material corrections should identify what changed. Send a page link, the disputed statement and supporting evidence through Contact & corrections.

The private question on our model records asks whether the page helped you decide. We review reader feedback during monthly editorial checks and investigate factual-error reports promptly. Reasons and specific missing questions guide improvements; a vote does not verify a claim, establish a product rating or represent all readers. Comments are not published automatically. How feedback is handled.

The shop links on Gear On Record are currently unpaid. We choose shops by model match, delivery market and clarity about what the seller lists. Check the seller’s page for the current price, stock, shipping and returns.

If we add links that pay us a commission, we will label them “Ad” and explain the arrangement before the links. The evidence and the reader’s buying needs determine our conclusions and the order of shops.

How we choose what to publish next

We prioritise a specific buying question with enough identifiable evidence to answer it. A model is added to the library when there is a readable record; being mentioned in a guide does not imply that a full analysis exists.

Method published: 16 November 2026. Feedback process added: 17 November 2026. Responsible publisher: Lemmo Murro.

The minimum for a publishable model record

  • An exact documentation region and hardware scope; unknown retail versions stay unknown.
  • A short conditional decision and a dependency table linked to dated evidence.
  • Each important factual claim checked against the actual linked passage in a separate verification pass. A broad FAQ is not a model-specific test.
  • Independent hands-on findings attributed to their original tester and sample. No minimum source count replaces relevance.
  • Open questions and a change note. Manufacturer questions are marked as sent only after they were actually sent.
  • Source dates change only after another evidence check; ordinary editing changes dateModified, not lastReviewed.