Characteristic Test for EVP

Migrated from the Collective

This practice has been moved from the ATransC Collective. The Collective was intended as a community effort to develop Best Practices. After years of failing to attract help from the paranormalist community, it seems reasonable to give up and unilaterally compose these practices.

These practices are recommendations provided under the Creative Commons Attribution-Noncommercial-Share Alike 3.0 Unported License


Tom Butler


There are a number of characteristics commonly associated with Electronic Voice Phenomena (EVP). If a possible EVP does not exhibit at least some of these characteristics, it may be prudent to set it aside until more evidence is available. This is not to say that a previously unknown characteristic may not be found in a “genuine” EVP, but the majority of examples clearly show a number of these characteristics. As such, it is recommended that experimenters and researchers become familiar with this list, and seriously consider using it as a means of avoiding false positives.


By definition, EVP are unexpected voices that are collected on digital and analog recording media, that are not explained by currently known physical principles. They appear to be ubiquitous, in that experimenters around the world are able to collect them with just about anything that will record human voice frequencies and under just about any recording circumstance. Their nature tends to vary, relative to the experimenter, recording environment, and technique.

The majority of EVP examples are considered Class C, meaning that they are difficult to hear and understand, and it is likely that not all of the words will be correctly deciphered. Nevertheless, Class C examples can sometimes be shown to be phenomenal utterances and often provide useful information. Even experienced experimenters are liable to mistake some environmental sounds, technological artifacts and editing errors as EVP. For instance, the unconscious intake of breath before speaking might sound like the word “help.” During field recording, an unnoticed person might be speaking in another part of the building and the resulting recorded words might be mistaken as a phenomenal utterance.

There is a Best Practice titled Control Recorder for EVP, which suggests using two audio records during experiments in order to reduce false positives. Some experimenters also protect the primary recorder with a portable radio frequency shield, such as two or more isolated and nested metal containers; however, such precautions can be clumsy, and may be difficult for the average person who is just trying to record a few EVP. Given that it has been experimentally established that EVP can be recorded in conditions isolated from ambient sounds, light or radio frequency contamination, it is reasonable to expect the average person to be able to record EVP in uncontrolled conditions.[1] If a person is familiar with the more common characteristics of the voices, and is willing to discard examples that do not fall within the “norm,” it is reasonable to conclude that the resulting EVP are likely to be genuine.

Typical Characteristics of Transform EVP

A characteristic test is not an absolute proof of EVP, but if stringently applied, it should reduce false positives to a reasonable minimum. Typical Characteristics of Transform EVP

  1. EVP are distinctive: EVP have a distinctive character of cadence, pitch, frequency, volume and use of background sound. The voices have a distinctive sound to them that is difficult to describe. For instance, EVP messages often have an unusual speed of enunciation; the words seem to be spoken more quickly than normal human speech.
  2. Frequency range: EVP are formed in available background sound. As such, if there is a high frequency component in the background sound, say caused by whistling wind, it is possible that the EVP will be of similar frequency range. If there are both higher frequency and lower frequency components in the background sound, it is possible to find EVP formed in both regions of the sound. In some instances, two different voices might overlap.
  3. A need for background sound sources: Research has shown that the voice in EVP is formed out of ambient sound energy.[2] Because of these characteristics, it is standard practice to assure the availability of ambient sound for voice formation, even while isolating the recording device or process from uncontrolled ambient sounds, such as crowd noise.
  4. Missing frequencies: Spectral analysis of EVP samples has shown that the fundamental frequencies of voice associated with the human voice box are sometimes missing in EVP. He describes the typical EVP as a “thickening” of the background noise to form the voice.[3]
  5. Precursor sounds: Sounds are often heard prior to an occurrence of EVP. Although these vary in nature, they tend to be within tenths of a second of a phrase and are a “popping” or “clicking” noise reminiscent of the “squelch” sound caused when the automatic gain control engages as the “push to talk” button is depressed on a Citizen’s Band radio.[4]
  6. EVP show evidence of being limited by available energy: Utterances tend to have about the same amount of audio power in their associated sound wave from one EVP sample to another. That is, a short EVP will tend to be louder than a long EVP. A very long phrase might be composed of two or more average length phrases separated by minor pauses. Also, an utterance may trail off at the end, as if the energy is being depleted before the message is delivered. Again, this is as if the communicator is attempting to manage available power as “packets” of energy. The evidence is very strong that EVP are energy-limited phenomena.[5]
  7. EVP are complete words or phrases: Message are typically one to two seconds in duration and are not truncated at the beginning or end. If EVP where radio interference, they would often begin in the middle of a word. EVP messages are usually complete thoughts, as well.[5]
  8. The voices in EVP are often recognizable: It is common for an EVP to contain the recognizable voice of the discarnate person thought to be speaking. It is also common for that entity to say something that was typical of what he or she would have said while in the physical. Their personality clearly remains intact even though the person no longer has a physical body.
  9. EVP is found wherever the practitioner listens: This suggests that the source of audio noise is not a factor for EVP, so long as the audio energy is suitable for voice formation. In practice, the majority of techniques for recording EVP involve sound conditioning, rather than unique forms of psi detection. For instance, upscaling infrasound so that it can be heard by human ears or downscaling ultrasound, really constitute techniques of sound conditioning, and the resulting EVP is not evidence that the utterance was formed beyond human hearing, but that it was formed when the audio energy was made available to the recording process.

Typical Characteristics of all forms of EVP

  1. EVP Are in the language of the practitioner: Alexander MacRae has conducted experiments in a place that has no English language radio or television stations, yet resulting EVP were in English, which is his primary language.[5] It is typical for the EVP, no matter where they are recorded, to be in a language that the experimenter understands. There have been exceptions to this which were apparently intended as a demonstration, but as a rule, EVP will be spoken in a language understood by the experimenter or an interested observer.

This brings up an interesting point of speculation about psi-based communication. Mental mediums often report that they receive communication from nonphysical entities as images which they must interpret. These images are not just mental pictures. They are packets of information that are sufficiently complete for the receiver to fully understand their meaning. Robert Monroe [7] referred to this form of information as “Thought Balls.”

  1. EVP are not ambient sound or broadcast programming: EVP are not ambient sound or broadcast programming: Studies have been conducted to determine if EVP are stray radio signals, ambient but unnoticed voices or other sounds. EVP were collected in an electrical, audio and radio frequency shielded room.[8] In one study, a radio and a recorder were placed in a padded chamber which was then buried under ground. The recorder did not record radio programming but did record EVP, which were transformed from the noise produced by the radio.[1]
  2. Party line: Some EVP sound as if they are comments intended for someone other than the experimenter. This is much like momentarily listening in on a party line telephone call. It is not uncommon in both field and controlled recording situations to record comments that seem as if unseen people are discussing the experimenter’s actions in much the same way that you might discuss the activity of someone that you were watching.
  3. EVP are appropriate to the circumstances: There are numerous examples of EVP that are clearly direct responses to questions recorded just prior to the EVP phrase or to the circumstances. An example of an EVP being appropriate to a circumstance is an EVP recorded by Lisa Butler. The Butlers were asking a woman about the upstairs lighting and sound room for the Frank Sinatra Theater at the Cal-Neva Casino at Lake Tahoe, CA. They had heard that the heavy door to the room often shut for no apparent reason, scaring the crews setting up lights and sound systems for shows. The woman told them that she would never ever go up there. Lisa’s recorder was on while she thanked the woman for her assistance. On the recording, Lisa can be heard saying, “Thank you very much.” Underneath her voice, is a clearly heard paranormal voice saying, “Please don’t come”. However politely said, it seems obvious someone did not want to be disturbed.
  4. Precognitive responses: Answers to questions may be recorded prior to a question being asked, so that the answer, as a phenomenal message, is on the soundtrack followed by the practitioner asking the question. More research is required before making informed speculation about this observed characteristic, but the indication is that, while time may be meaningful to us, our time may well be irrelevant to a nonphysical entity. Alternatively, the entity may be sensing what the practitioner is about to ask as a mind-to-mind exchange following the “avatar model” as discussed in the Comparing Personality-Body Models.[9]
  5. Vocalized questions elicit more EVP: There is evidence that the communicating entities are able to read our thoughts, as illustrated by the occasional EVP which clearly responds to a comment just seconds before the comment is expressed.[5]
  6. The “Newness” effect: The experimenter’s excitement in trying a new detection device or recording technique may be the source of improved EVP collection. As the new approach becomes “normal operating procedure,” the improvements generally fade back to a more “normal” Quality and Quantity (QQ) of EVP collection. This suggests that it is important for the experimenter to maintain peaked interest during experiments. This is also one of the reasons it is speculated that the experimenter is an integral part of the recording circuit. The experimenter is apparently supplying the necessary psi energy to enable a nonphysical to physical transfer of energy.
  7. Effective devices unique to the practitioner: Exceptionally effective EVP and ITC collecting systems have been developed; however, these typically work well for the developer, but do not necessarily work as well for other experimenters. This paradox supports the belief that the experimenter is part of the recording circuit. It has also reinforced the concept that the communicating entity may be specific to the experimenter.
  8. EVP can be thoughts of living people: Two experiments appear to show that at least some EVP might be initiated by living people who were sleeping or perhaps only distracted at the time. In these prearranged experiments between a practitioner and a sleeping person, questions were clearly answered by a communicating entity, and the answers are appropriate for the sleeping person. This fact of EVP suggests the possibility that EVP can become an important tool for consciousness research. For instance, is it possible that a patient in a coma might initiate an EVP when requested?[10]
  9. Understanding EVP may be like learning a new language: As discussed in the EVP Online Listening trials report, people with little or no experience listening to EVP will typically correctly report words in Class A transform EVP on average of 20% to 25% of the time. In contract, an experienced practitioner should correctly understand close to 100% of Class A utterances.[11] This number drops to 0% to 5% for studies of radio-sweep (Spirit Box, Frank’s Box, Ghost Box); 0% if single-syllable utterances are omitted.[12], [13]


  1. Weisensale, Bill (1981), “Shielding a Recorder from Radio Frequency Interference for EVP,” Spirit Voices, Issue 3, 1981. Republished: Association TransCommunication website. “Eliminating Radio Frequency Contamination for EVP,”
  2. Gullà, Daniele (2004), Computer–Based Analysis of Supposed Paranormal Voice: The Question of Anomalies Detected and Speaker Identification”
  3. Presi, Paolo, Italian ITC researcher with Il Laboratorio, Bologna, Italy, biopsicocibernetica Closed.
  4. Butler, Lisa (2002), “Precursor Sounds in Physical Phenomena,” Association TransCommunication,
  5. MacRae, Alexander (2000),The Mystery of the Voices, Self published CD, Portree Skye, Scotland. for details about the Alpha Device,
  6. Blank
  7. Monroe Institute, 62 Roberts Mountain Road, Faber, Virginia 22938,
  8. MacRae, Alexander (2003), Report of an Anomalous Speech Products Experiment Inside a Double Screened Room, as printed in the Journal of the Society for Psychical Research,
  9. Butler, Tom (2010), “Comparing Personality-Body Models,” Association TransCommunication,
  10. Butler, Lisa (2002), “French Sleep Experiment,” Summer 2002 AA-EVP NewsJournal,
  11. Butler, Tom (2008), “EVP Online Listening Trials,” Association TransCommunication,
  12. Butler, Tom (2009), “Radio-sweep: a Case Study,” Association TransCommunication,
  13. Leary, Mark (2013), “A Research Study into the Interpretation of EVP,” Association TransCommunication,



Leave a Comment