Frequently Asked Questions

An ordinary seed feeder with a camera pointed at the perch, a battery behind it and an app that sends you a clip when a bird lands - usually with the species identified. The full explanation of how they work is here.

Most offer one and some require it for species identification specifically, which is the feature the category is sold on. A few state plainly that no subscription is needed - they are collected in the no-subscription roundup.

Most makers do not publish the price on the listing you buy from, which is the core problem. Rather than invent figures we give you the method: hardware plus the monthly plan times 36 is the three-year cost, and that is the number to compare. The full guide is here.

On published specifications the Birdbuddy Pro Solar, because it is the only model in the category publishing a sensor dimension - the figure that most affects a backlit photograph. The full roundup ranks eight models.

Almost always because it is 2.4GHz-only and your router publishes both bands under a single network name. Temporarily switching off the 5GHz radio, pairing the camera, then switching it back on is the standard fix. The Wi-Fi guide has the detail.

Nearly all. The one exception we have found is the Birdfy 4K Full-Metal, which publishes dual-band support and joins a combined-band network the way a phone does.

It may carry that many labels - large training sets are ordinary in image classification. But the IOC list recognizes 11,227 species worldwide, the whole of North America has 1,161, and your feeder will see a few dozen. The audit explains why the number does not discriminate between products.

Good on common feeder birds, less reliable on similar species and anything unusual. No maker publishes an accuracy figure, so anyone quoting one is inventing it. Cross-check anything surprising with Cornell's free Merlin app.

Less than the number suggests. A small lens cropped to a perch a few inches away, compressed for upload over 2.4GHz, does not deliver four times the useful detail. Sensor size and dynamic range matter more - and almost nobody publishes them.

It depends on how busy your feeder is, because power is consumed per wake-record-upload cycle rather than over time. A weak Wi-Fi signal is the most common cause of unexpectedly fast drain, since the camera burns power retrying.

In open sun, rarely. In shade, regularly - and possibly more often than the non-solar version of the same camera, which sometimes carries the larger battery. The solar roundup has a five-minute test for which yard you have.

They will try, and on a camera feeder what gets chewed is the housing keeping water off the electronics. A metal body, a weight-triggered port or - most effectively - a baffled pole are the answers. The squirrel-proof roundup is candid that siting beats hardware.

No. The camera is silent, static and does not flash in daylight. What people are seeing is the normal delay before any new feeder is found and trusted - two weeks of nothing is entirely normal. More here.

Days to weeks, and longer in a garden with no history of feeding. Moving it nearer cover and scattering a little seed underneath both help; changing several things at once does not.

In Wi-Fi range first, within about ten feet of cover second, with the sun behind the camera rather than behind the bird third. The setup guide works through the compromise.

Black oil sunflower as the default - it is taken by the widest range of feeder birds. Safflower deters grackles and squirrels; nyjer brings goldfinches. Avoid cheap mixes that are mostly filler.

Yes - these are outdoor devices. Cold reduces battery performance and thin plastic gets brittle over repeated freeze-thaw cycles, so a metal body is an advantage in hard-winter regions.

Not as a first purchase. A seed feeder works every day of the year and starts within weeks. A bath shows you species a feeder never will, and a nest box shows extraordinary footage for about six weeks a year - both are better second cameras than first ones.

One with no recurring charge, whatever its specifications. A present that asks the recipient for a card number to work is not a present. The gift list ranks on exactly that basis.

No, and we say so on every page. We compile published manufacturer specifications, print the ones makers decline to publish, do the arithmetic and cite every figure. How we pick sets out the method and its limits.