Ghost Hunting Equipment: What to Buy First

The Survey has been issuing equipment since 1974 and getting most of it back broken. What follows is the order in which it would spend money if it were starting again, which is not the order anybody selling equipment would suggest.

1. A notebook, and a pen that works in the cold

Four pounds. The single most useful item the Survey has ever issued, by a distance that is not close, and the one most likely to be missing from a group that owns two thousand pounds of instruments.

Everything else on this list produces a number. A number that is not written down beside a time, a position and what you were doing is not a measurement, it is an anecdote about a light. Six months later the notebook is the only part of the night that still exists.

Get one with a hard cover and squared paper, because you will draw plans in it. Take two pens. Ballpoints fail in the cold and pencils do not.

2. A carbon monoxide detector

A domestic audible alarm is a few pounds. A read-out instrument that shows parts per million is better and still cheap. Either is a serious instrument and neither is optional.

3. A red-filtered torch

Your dark adaptation takes about twenty minutes to build and about one second of white light to destroy. A group with white torches spends the entire night in a state of permanent partial blindness and then reports seeing things at the edge of vision, which is what partially blind people do.

Red light preserves rod sensitivity well enough to keep most of it. A head torch with a red mode, or a plain torch with a red filter, is a few pounds and changes the quality of everything you see all night.

4. A thermometer with a proper air probe

Not an infrared gun. An infrared thermometer measures the temperature of a surface, from whatever spot size its optics give it at that distance, weighted by that surface's emissivity — which for a mirror, a polished floor, a window or a stainless steel handle is wrong by a long way. It cannot measure air at all.

A cold spot in the air is a claim about air. To test it you want a thermocouple or a thermistor probe with a small thermal mass, given time to equilibrate. It is slower and much less photogenic and it is the only one of the two that answers the question. There is a whole note on this because it is the single most common instrument error in the hobby.

5. A recorder that records uncompressed

Not a phone voice memo, which will hand you a compressed file. A lossy codec is a perceptual model: it works by deciding which parts of the signal you cannot hear and deleting them, and the first thing it deletes is quiet broadband detail thirty decibels down — which is the entire subject of the exercise.

You want WAV, 44.1 or 48 kHz, 16-bit or better, and you want the automatic gain control off if the device will let you turn it off. A recorder with AGC will pump its own noise floor up and down around every event and give you an evening of artefacts to mark.

6. Then, and only then, an EMF meter

The one everyone buys first. There is a whole note on what these measure and the short version is: buy a three-axis instrument, read its frequency band before you read its marketing, and expect a number rather than a mood.

The five-LED instruments are not frauds and they do work. They are single-axis, coarse, and tuned to your mains wiring, and in an old building they will spend the night telling you about the lighting circuit.

7. Cameras, last

A camera is the item people most want and the item least likely to produce anything defensible, because a camera in a dark building produces artefacts by the hundred and every artefact looks like evidence to somebody who wants it to. Sensor noise, lens flare, dust, moisture, insects, breath, IR hot spots, motion blur, and the compression the camera applies to its own footage all generate exactly the shapes that get posted.

Buy the tripod before the camera. Half of what is claimed from handheld footage disappears when the frame stops moving, and a tripod is also the only way two frames of the same corner can honestly be compared.

What we would not buy at all

• Anything advertised as detecting spirits, energy, or presences. No such instrument exists. What exists is instruments that measure ordinary quantities and marketing that says otherwise.
• Random-word generators sold as spirit boxes. A box with a vocabulary is a box that chose the vocabulary. If you want the effect honestly, our own tells you exactly what it is doing and prints the seed for every word it produces.
• Dowsing rods, at any price above zero. Two lengths of coat hanger will demonstrate the ideomotor effect identically and the demonstration is worth doing.
• Anything sold as a "ghost box app" with a numeric field reading. A web page cannot read your phone's magnetometer; the number is invented. This is worth checking before you trust anything else that app tells you.

The order matters more than the list

Everything above the EMF meter costs under £150 in total and improves every measurement you will ever take. Everything below it costs more and improves nothing until the things above it are in place.

The instruments on this station are all free and they run in a browser, so you can find out whether you actually enjoy the procedural half of this before you spend anything at all. That is genuinely what they are for. Start with the bench.