by Karen Frazier
A simple and inexpensive Radio Shack hack ghost box (Radio Shack 12-469)

Can the dead communicate with us through a field of white noise? That's the theory behind the ghost box, a piece of equipment that seeing more common use with many paranormal researchers.

The concept behind the ghost box is quite simple: a normal transistor radio is modified so it rapidly sweeps the stations of either the AM or FM band instead of remaining on one station. This creates a field of white noise. Investigators hypothesize spirits may be able to use this for communication.

History of the Ghost Box

In 2000, a paranormal researcher named Frank Sumption began experimenting with a computer program called "EVP Maker,"  which was developed by Stefan Bion. The software essentially randomized the elements of speech in such a way it created background noise spirits might be able to use to form speech. Sumption experienced some success with the software, but set out to create a more portable form of noise randomization. The result was Frank's Box, a white noise generator with variable AM tuning. After experiencing some success with the design, Sumption continued fine-tuning his device, producing and testing more than 25 versions of the original Frank's Box. With the success of the Frank's Box, paranormal researchers sought to create their own versions from inexpensive AM/FM radios.

The Shack Hack and Other Ghost Boxes

The Jensen SAB-55
Paranormal researchers found it was quite easy to make a simple ghost box that scanned either the AM or FM band of a transistor radio by simply cutting a wire. Common radios used were the Radio Shack 12-469 (which requires the cutting of a pin on a circuit board), or the Jensen SAB-55 (which requires cutting a purple wire). Other homemade versions exist, as well, but these are the two I was easily able to purchase and make on my own. Both have yielded results, and for someone as non-technical as me, were super easy to modify. If you're new to ghost box research and would like to cheaply and easily begin testing, I can recommend making either of these versions, as the instructions are readily available on the Web.

Commercially Available Versions of the Ghost Box

The RT-EVP 2 available from GhostStop
There are also several commercially available versions of ghost boxes (also called spirit boxes) that don't require any hacking skills on your part. GhostStop sells a couple. SSPR has recently purchased both the B-PSB7 and the RT-EVP 2. We haven't yet begun experimenting with them, but given our history with the Jensen Hack and Shack Hack versions, we're hopeful these models will work as well as, if not better than, the homemade devices we currently have.

My Experience with Ghost Boxes

While I can't say without equivocation that ghost boxes work, I have had some rather interesting results.
  • On a private residence investigation, we were doing a ghost box session in the vicinity of a sword that belonged to the homeowner's grandfather. The box kept repeating the word, "Opa." When we asked the homeowner, it turned out they had called the grandfather "opa." At the same time, the ghost box kept saying the grandfather's name, which was an unusual one. The opa and name were repeated in the same voice each time.
  • When investigating a music recording studio, I was using the ghost box in the control room. The room was shielded from outside radio waves. As a result, the ghost box only made static noise and didn't tune in to any radio stations. Investigators were able to have a complete conversation over the ghost box with a guy named Steve from Cincinnati, who spoke clearly through the ghost box and used a consistent voice, accent, and language throughout the conversation.
  • In an investigation at Wellington, I was using a ghost box on a trail overlook while other investigators had set up a laser grid in the snow shed. The ghost box began talking about the lights and asking to have them turned off. When I asked, "Do you mean the lights in the snow shed?" the ghost box answered, "Yes." When I asked if they would like me to ask them to turn off the grid, the ghost box answered, "Please." When I came back from asking them to turn off the lights, the ghost box said, "Thank you." Each piece of the conversation was delivered in the same female voice.
Limitations and Cautions 

While the above events were compelling, it's important to understand the limitations of the ghost box. Because it generates white noise (and sometimes random snippets of words from radio stations as it sweeps past them), it's easy to fill in the blanks. The brain tends to recognize patterns, and it often attempts to create order from chaos. There's a particular quirk of human psychology known as pareidolia, in which the brain perceives significant information from random stimulus. This is what causes people to misinterpret a foot shuffle as a whisper in an EVP recording session or to see the face of Jesus in a pancake.

It's possible many of the words heard on a ghost box are simple pareidolia. To guard against this, SSPR records all ghost box sessions and carefully analyzes the audio after the fact to determine whether what we thought we heard in the moment was actually what the ghost box was saying. When I perform this analysis, I listen for consonants and syllables. If the sound from the ghost box doesn't exactly match the actual spoken word (for instance, if the ghost box says buh instead of but), I throw it out. When I do have hits on words or phrases, I pass it on to another person, who also checks to see what they hear. I don't communicate with the second set of ears what I heard, because that can contaminate their perceptions. 

I also throw out random snippets that are a single syllable or even a single word, which can often arise just from a voice on the radio as the box sweeps past a station. For example, if I hear the word, "love," I'll very likely throw it out. It is a commonly used word and probably was just spoken on a station. If, on the other hand, I hear the phrase, "She loved me," (something we actually heard during a ghost box session), it is more likely to be from something other than words on the radio, because the box is sweeping the band too quickly to allow all of those words from one person.

Context is also important in this analysis. If what you hear makes sense in the context of where you are, and the conversation is consistent and not random (as was our conversation with Steve from Cincinnati), this is more likely to be significant than if the conversation doesn't make sense or is mostly random words and gibberish.

Finally, repetition may be an issue. When I hear a repeated sound, I time it with the scanning of the bands. If the sound (even if it sounds like a word) occurs at the same spot in the band every time, it's clearly signal interference and I throw it out. If, however, it repeats a word or phrase at random times (as was happening when the ghost box was saying the word "opa"), and it fits within the context of the situation, I will be more likely to consider it possible evidence.

The Bottom Line
Like every other piece of equipment, the ghost box is experimental in the field of paranormal research. No controlled scientific studies prove any version of the ghost box communicates with spirits of the dead. When used appropriately, however, it may provide another bit of evidence that can help us understand what is occurring in a specific location.