Skip to content
We may earn a commission when you buy through links on this site. It never changes what the planner recommends. How we make money
HMC
Region
My list
Menu
Planner Cameras Doorbells Locks Subscriptions Answers How we calculate

How we calculate

The interactive Planner uses an illustrative placement sketch, followed by a device list and evidence-based recording comparison. Legacy example pages use a separate specification-based calculation described below. Neither is a site survey or a hardware test.

Current interactive Planner

The starting view is 90° wide and 25 ft deep. Automatic placement tries directions in 30° steps; manual editing lets you change positions, directions and the illustrated field of view. Day and night use the same illustrative range. A verified device field of view can be applied during assignment, but the drawing does not predict identification distance, motion triggering, illumination or night performance.

System cost and recording evidence · version system-cost-3

Upfront cost is the selected hardware subtotal plus the card allowance multiplied by reviewed card slots, plus your other upfront budget. Three-year cost adds 36 times the whole-system monthly bill; five-year cost adds 60 times that bill. Bundles are priced once and items marked already owned have zero new purchase cost. The selected list is the scope: unlisted existing cameras are not charged automatically.

A zero recording fee is used only for the exact reviewed local-storage configuration in its reviewed market. Cloud, hub and HomeKit Secure Video requests, unknown models and missing regional evidence require confirmation. For reviewed US Tapo and Wyze configurations, an optional cloud estimate chooses the lowest price among eligible reviewed offers per provider, respecting exact configuration, market, device cap and billing period. Owned cameras count toward service coverage; each shared plan is charged once. Account and home scope need user confirmation. Annual costs use exact annual invoices multiplied by 3 or 5, not rounded monthly equivalents. Tapo base annual pricing is excluded because its per-device billing unit is unclear. This is not a guarantee of the cheapest checkout offer or feature equivalence between tiers. Other plans use a custom whole-system monthly budget with capacity unchecked. Prices are dated catalog offers in USD or CAD, never currency conversions. Unpriced hardware and required dependencies remain visible and the result is labelled incomplete.

Budget entries are user estimates. Taxes, shipping, electricity, replacements and future price changes are excluded unless entered. A and B use the same layout targets and current catalog data. Shares preserve selections and budgets; prices may change on reopening, so they are not historical quotations. Recording claims have an exact model/variant, market, review date, official source and conditions. Unknown and conflicting evidence remain explicit. Recording reviews cover 12 US and 3 Canadian configurations. Canadian cloud prices are not yet verified. Cloud estimates retain the reviewed local storage setup; they add cloud history without removing card allowances. Continuous cloud recording and premium AI eligibility are not inferred from a subscription price.

Inspect the reviewed configurations and reproduce them in Planner.

Owned equipment and upgrade comparisons

Inventory records describe equipment at other locations until explicitly moved into the selected installation. Reuse requires a current catalog binding, an exact variant and an available quantity. A move deducts the chosen quantity from inventory and adds it as an owned line; matching typed names never establishes verified support.

The upgrade tool saves the visible selection as A. Expansion keeps all lines and adds a new purchase. Replacement removes only the specified quantity from an owned individual camera or doorbell line in B, then adds the chosen new model and variant. Other purchases, hubs and inventory remain. Kits cannot be split by this tool. Existing A is overwritten only through the labelled action. New purchases never inherit owned status.

Owned camera purchase prices are sunk costs and excluded from future spending. B's storage-card, setup and custom monthly allowances are cleared, and the estimate remains incomplete until budgets are reviewed and applied again. A reviewed cloud billing period is kept but its account/home confirmation is reset. Service fees for equipment continuing elsewhere must be included in the user's whole-system budget. No resale, trade-in, cancellation or performance gain is assumed. Required hub checks use exact catalog bindings, not brand affinity. Saved positions and layout targets are unchanged; a replaced device may need reassignment and expansion may need more positions. These are cost scenarios, not proof of equivalent coverage.

Open the keep-versus-replace examples to review the inputs in the Planner. Plan links include owned equipment and both alternatives; default device-list shares omit owned equipment.

Outage evidence and inference

The failure matrix refers to exact recording configurations in their reviewed market. Internet loss keeps power and the local network; router loss keeps camera power but removes the network; household power loss removes mains power and router operation with no backup assumed. Battery cameras are assumed to have charge remaining. A and B use the same selected failure.

Official documentation and dependency inferences are labelled separately and cited. For example, an outlet-powered camera with no backup cannot operate during the stated power failure; a charged battery plus a documented offline recording route can support a conditional inference, not a measured runtime. Wyze's scheduled timelapse guidance does not establish ordinary event/continuous outage recording. Evidence does not transfer across variants or regions automatically. Hub/NVR routes, cellular failover, retained cloud clips, card-reader access, restart recovery and storage corruption are outside this matrix.

Legacy examples: specification-based engine

The remaining geometry and subscription methods on this page describe the legacy example reports, not the current interactive sketch. Opening an example transfers its actual devices and mounting positions; its older coverage percentages are not carried into the illustrative editor.

1. How far a camera can usefully see

Security designers measure detail in pixels per foot (px/ft) at the target. We use three levels: detect at 8 px/ft (something is there), recognize at 20 px/ft (a person you know, a car's color and shape), and identify at 40 px/ft (a stranger's face). The planner's range for each camera is the recognize distance:

range = (sensor width px ÷ 2 ÷ tan(field of view ÷ 2)) ÷ 20

  • Example: a 2K camera (2560 px wide) with a 140° lens recognizes out to about 23 ft.
  • Lenses wider than 150° are calculated as 150°. Ultra-wide lenses are distorted, and the simple formula would otherwise overstate their range.
  • When a brand publishes only "2K" or "4K" rather than pixel counts, we use the lowest common width for that label (2K = 2304 px, 4K = 3840 px) and mark it as assumed.
  • Range is capped at 60 ft, and at the published motion-detection range when there is one. Battery and solar cameras only record when their motion sensor fires, so without a published figure we cap them at 40 ft, which is typical for PIR sensors.

2. Night vision

  • We use the published night-vision distance when there is one.
  • Floodlight cameras (700 lumens or more) light the scene, so we allow at least 40 ft at night, never more than the daytime range.
  • With no published figure we assume 30 ft for infrared, 30 ft for spotlight color night vision and 25 ft for low-light color sensors, and label it as assumed.
  • Areas you mark as lit at night use the daytime range.

3. Blind spots

Each area in a layout is split into 2 × 2 ft cells. A cell counts as seen when it is inside a camera's lens angle, within its range, and the straight line from the camera to the cell doesn't pass through the house, garage or another building.

A required area is a blind spot when less than 70% of it is seen, or when no camera sees one of its key spots (the front step, the garage door, a gate). At night, an area is flagged when its seen share drops below 60% of the daytime figure or a key spot goes dark.

4. How the planner places cameras

Layouts come with realistic mounting points: eaves, porch lights, outlets and the door frame. For each goal the planner adds one camera at a time. It tries every free mount, direction (15° steps) and eligible model, and keeps the one that helps most, until the required areas are covered and the next camera adds little. It never uses more than 8 cameras.

  • Lowest 3-year cost: the most coverage per dollar of hardware plus 36 months of plan fees.
  • Easiest install: no new wiring, and one brand where possible, so one app.
  • Most coverage: the most area seen by day and night within your budget.

If required areas can't be covered, we loosen settings in this order and tell you: budget +20%, then allow monthly plans, then allow wiring. Cameras you already own stay where they are and cost nothing.

5. Power at each spot

  1. Door frame with doorbell wires: wired doorbell. Otherwise a battery doorbell.
  2. If you can run cable: PoE first, then a floodlight on an existing light fixture, then an outlet.
  3. If you won't run cable: an existing light fixture or outlet.
  4. Otherwise solar, unless the spot is in full shade.
  5. Otherwise battery.

6. Battery estimates

Brands state battery life under a usage assumption, such as "300 days at 230 seconds a day". We convert that to your spot's expected motion events per day, assuming 20 seconds of activity per event. When a brand gives only a number of days, we assume the claim was measured at 10 events a day. We never estimate longer than the brand's own claim. In cold climates we multiply by 0.5: lithium cells lose capacity and won't charge below 0 °C (32 °F).

7. Wi-Fi reach

signal (dBm) = −30 − 22 × log10(distance ft ÷ 3.3) − outside walls × 10 dB (wood) or 15 dB (brick, concrete) − inside walls × 4 dB

We count one inside wall per 12 ft of path through the house. At -65 dBm or better the signal is good, down to -75 dBm it is borderline, and below that it is likely too weak. We suggest a mesh node just inside the wall nearest the weak spots.

8. Cost

  • One-time cost is the list price of every device plus anything it requires: a base station, PoE switch or lock bridge. Devices you own cost $0.
  • Monthly cost is the cheapest plan tier that unlocks what each device needs: recording for Ring, Arlo, Blink and Nest, plus anything extra you ask for. For several devices of one brand we compare per-device and unlimited plans.
  • 3-year total = one-time + 36 × monthly.
  • Canadian prices are stored per product, not converted with an exchange rate. Missing Canadian hardware or subscription prices remain unknown; US amounts are not added to a CAD total.
  • microSD cards, installation labor and taxes aren't included.

9. Data and updates

10. Limits

This is a planning tool, not a site survey. Real yards have trees, fences, slopes, glare and neighbors' lights. Before you drill, tape the camera up or run it on a long cable and check the live view. Mounting height, tilt and weather also change what a camera sees, and we don't model them.