EVP Recording Guide: How to Capture and Analyze Electronic Voice Phenomena

null null

Electronic Voice Phenomena, or EVP, is the name given to a sound in a recording that resembles a voice and was not made by a voice in the room. Whether a spirit, an unknown intelligence, a piece of equipment, a building or the listener is responsible is a separate question from how the sound is captured, and the two get mixed together constantly.

Whether you are a seasoned investigator or a curious beginner with a phone, the order of operations matters more than the gear. This guide covers the capture method, the equipment and what each one actually measures, how evidence is classified, and the ordinary explanations that have to be ruled out before anything else is claimed.

A disclosure before anything else. The recorder recommended below, NOCTEM, is published by Cha0smagick Labs, and so is Eerie Roads. We make them, and the same rigour is applied to them as to any other tool: the equipment section says what each device actually measures, and the section on ordinary explanations says what to rule out before anyone uses the word voice.

The History of EVP

Interest in EVP dates back to the early 20th century, when Thomas Edison speculated that a device might be built to communicate with the dead. In the 1950s, Swedish painter Friedrich Jürgenson accidentally recorded what he believed were spirit voices while recording bird songs. His work inspired Konstantin Raudive, who conducted thousands of recordings and documented over 100,000 alleged spirit voices.

Today, EVP is a standard practice in paranormal investigation, supported by modern digital recording technology that makes capture and analysis more accessible than ever.

How to Record EVP: Best Practices

Preparation

  • Use a dedicated session recorder that keeps the signal unprocessed, such as the one in NOCTEM. Generic voice memo apps apply automatic gain and noise processing, and that processing is tuned for speech: it will remove or smooth exactly the quiet, irregular sounds a recording claim depends on. The recorder is the least important part of the setup. The room, the silence and the control recording matter more.
  • Minimize background noise. Turn off HVAC systems, refrigerators, and other ambient noise sources. Every background sound is a potential false positive.
  • Set your recording levels. Real-time level monitoring is worth having, because a recording that clips is worthless and a recording quiet enough to leave out the interesting part is also worthless. Set the gain by ear and by meter, and write down where you set it.

During the Investigation

  • Use the EVP "question and answer" technique: Ask clear questions and leave 10-15 seconds of silence after each for potential responses.
  • Record continuously. Don't stop and start, some of the best EVP captures happen when you least expect them.
  • Note environmental conditions. NOCTEM logs location, time, and environmental sensor data alongside every recording.

Post-Investigation Analysis

  • Listen with headphones in a quiet environment. Many EVPs are subtle and easily missed on speakers.
  • Use spectral analysis. Visualizing the audio frequency spectrum can reveal patterns that ears miss.
  • Get a second opinion. Have another investigator listen to potential captures. If multiple people hear the same thing independently, it's a stronger piece of evidence.

Pro Tip: NOCTEM's EVP recorder is optimized specifically for paranormal investigation. Unlike generic audio recorders, it captures high-sensitivity audio while maintaining professional-grade clarity, and stores everything locally, no cloud uploads, complete privacy.

Classifying EVP Evidence

Paranormal researchers classify EVP into three categories:

  • Class A: Clear, distinct voices that multiple listeners agree on without suggestion. Rare and considered strong evidence.
  • Class B: Moderate quality. The voice is discernible but requires some interpretation. Most common in serious investigations.
  • Class C: Poor quality. The voice is faint or obscured by noise. Often dismissed as pareidolia or interference.

Common EVP Challenges

Audio pareidolia is the biggest challenge in EVP research. The human brain is wired to find patterns in noise, and this includes hearing voices in random audio artifacts. Experienced investigators mitigate this through rigorous documentation, multiple reviews, and classification systems.

Environmental contamination is the second major challenge. Radio interference, mechanical sounds, and even distant conversations can be misinterpreted as EVP. This is why high-quality equipment and controlled environments are essential.

Start Recording EVP with Professional Tools

NOCTEM's professional EVP recorder gives you high-sensitivity audio capture, real-time monitoring, and secure local storage. No subscriptions. One payment. Lifetime access.

🎙️ Get NOCTEM on Google Play

Where EVP Recording Guide: How to Capture and Analyze Electronic Voice Phenomena sits in the library

Thirty entries spread across 12 subjects, which is what happens when entity lore is treated as a way of thinking rather than a topic.
  1. EVP vs Spirit Box: Complete Guide to Paranormal Audio...mind science
  2. NOCTEM App Review 2026: The Best Ghost Hunting App...entity lore
  3. Ghost Hunting Methods and Equipment Guideentity lore
  4. Synchronicity Hunting: A Beginner's Guide to Tracking Meaningful Coincidence...entity lore
  5. Angels, Spirits, and You: A Framework for Contact (2026)entity lore
  6. Entity Lore and Invocationentity lore
  7. LILITHentity lore: A complete working guide to LILITH: what it is, the entity lore and invocation structure behind it, and the full protocol. Includes step-by-step...
  8. Electronic Voice Phenomena (EVP): The Science of Spirit Voicesastral parasites
  9. 5 Best Ghost Hunting Apps for Android in 2026...secret knowledge
  10. Paranormal Investigation Apps vs Professional Equipment (2026): What Actually...mind science
  11. EVP & Spirit Box Session Setup: A Complete Guide...astral dreams
  12. Ghost Hunting Apps for Android Compared: Find the Best...mind science
  13. What Actually Happens in an EVP Session: No Jump...astral parasites
  14. NOCTEM Paranormal Suite: Complete Feature Tour (2026)mind science
  15. Eerie Roads: Real Field Results After 50 EVP Sessions...servitor craft
  16. Types of Paranormal Phenomena: A Complete Guidemind science
  17. Your Smartphone Is a Paranormal Investigation Lab: The Ultimate...mind science
  18. What 100 EVP Sessions Taught Me About Grief (2026)astral parasites
  19. What Is a Spirit Box? How It Works +...astral parasites
  20. Ghost Hunting at Home: Your First Investigation Guide (2026)home
  21. Halloween EVP Night: How to Run a Public Session...astral parasites
  22. Dream Journaling Methods & Best Apps (2026)astral dreams
  23. How to Read Your Natal Chart: Complete Beginner's Guide...chaos basics
  24. Cybermancy: Complete Guide to Digital Sorcery (2026)technomancy
  25. 10 Best Tarot Apps for Android in 2026: Honest...tarot
  26. Design, Charge, Forget: The Complete Chaos Sigil Workflow (2026)sigils
  27. Digital Spellcasting: How to Build a Tech-Enhanced Ritual Technomancy...technomancy
  28. How to Make a Digital Sigil: Complete Guide to...sigils
  29. Ouija vs Digital Spirit Box: Fear Compared (2026)technomancy
  30. Best Moon Phase Journal Apps (2026)lunar

What the equipment actually measures

The first useful thing to understand about EVP recording is that a recorder records audio. Everything that follows is about how faithfully the audio reaches the file, and how much of the room reaches it as well.

SetupWhat it measuresIts honest limitation
Phone voice memoWhatever the phone microphone picks up, usually wide and compressedAutomatic gain control and noise suppression are tuned for speech and will remove or smooth exactly the quiet, irregular sounds a recording claim depends on. A phone can capture a loud voice perfectly and a faint anomaly badly
Phone plus wired lavalierThe same microphone, closer to the source, with fewer room reflectionsStill automatic processing on most apps. Wired matters more than expensive
External microphone into a phoneA flat, unprocessed signal from a capsule you choseNeeds an adapter and a level that does not clip. Cheap adapters add noise of their own
Handheld recorder, built-in micUncompressed audio with a real preamp and usually a limiter you can switch offThe built-in capsule is narrow and the machine is designed for field notes, not for a quiet room
Handheld recorder with an external micThe same machine with a capsule chosen for the jobThe best of the lot, and the cheapest of the good options
Digital SLR or mirrorlessVideo with audio, long takes, timecodeAudio is a secondary function and is often processed. Video frames carry their own interpretative risk, which is dealt with below
Static or sweeping radioAudio with a noise floor that never sits still, because the radio is being tunedThe noise floor is doing the work. A voice-shaped sound in a radio sweep is much weaker evidence than the same sound in a quiet room, because the radio itself manufactures voice-shaped artefacts all the time
Session recorder or EVP appUnprocessed audio, with markers, a session log and a way to tag events as they happenIt is a recorder with a workflow around it. It does not make the microphone better, and it does not decide what you are hearing

Two practical conclusions come out of that table. The first is that the microphone and the room matter more than the brand of the machine, and that the single most damaging default in phone recording is automatic noise processing, which exists to make speech sound clean and will happily reinterpret a breath. The second is that any setup with a synthetic noise floor, which includes every spirit box and every radio sweep, has to be compared with the same setup running with nobody in the room, before the anomaly means anything.

Before you call it a voice: the ordinary explanations

Most alleged EVP has an ordinary explanation, and the interesting part of the practice is the work of finding it. These are the ones that account for the large majority of disputed recordings, and each has a diagnostic that takes a minute.

ExplanationWhat it looks likeHow to rule it out
Mains hum and its harmonicsA steady tone, or a stack of tones, at 50 Hz in most of the world and 60 Hz in the Americas, and the odd harmonics of itLook at a spectrum view. If the peaks sit at 50 or 60 Hz and their multiples, you have an electrical problem, not a voice
Fluorescent lights, dimmers, switch-mode power suppliesIntermittent clicks, buzzes and sibilance that come and go with the lightingRecord with the lights off, then with them on, and compare. The difference is the lighting
Building vibration, traffic, a fridge, a pipeLow-frequency thumps, often rhythmicPut the microphone on the floor and listen to the contact rather than the air. A structure-borne sound is usually audible in the recording as a dull thud rather than a sound
Handling noise and clothingCrackle and rustle that the operator generatedRecord a minute of silence at the start and the end of every session. Any noise in those two minutes is a sound you have to discount
Another device in the roomSpeech, notification tones, a video playing somewhere elseRecord with every device but the recorder switched on, one at a time
Radio or phone interferenceWords, or something that sounds like words, in a synthetic or smeared formCompare against the same equipment in a different room. Voice-shaped artefacts travel with the sweep, not with the ghost
Paracusia and expectationThe listener hearing what the context suggestsReview blind, with the dates and locations removed, and let someone else label the files. This is the most common explanation in the literature and the one no instrument detects

The last row is the one that separates an investigation from an entertainment, and it is cheap: renumber your files, remove the context, and have somebody who was not there sort them. A recording that only makes sense with the location attached has not been shown to be anything yet.

None of this is a debunking of the people who do this seriously. It is the reason the serious ones insist on it. A claim that survives a spectrum analysis, a control recording, a blind review and a check of the ordinary sources of noise is worth something; a claim that has not been through them is not, and the difference is entirely in the work rather than in the belief.

Discussion & Comments

Tried this practice? Tell us what happened below. Reader results are the most useful feedback we get — they show other practitioners what to expect and tell us which guides to write next.

Frequently Asked Questions

What is EVP in paranormal investigation? ▼
What's the best EVP recorder app? ▼
How do you analyze EVP recordings? ▼
Why do paranormal investigators use EVP? ▼

A recording workflow that protects the evidence

An EVP recording is not just an audio file with an intriguing name. It is a chain of decisions: who was in the room, what was said, how the microphone was positioned, whether the recording was stopped or edited, and what happened afterward. A useful archive preserves that chain so a future listener can distinguish an observation from a conclusion.

Begin with consent. Everyone who can be recorded should know the purpose, the approximate duration, who will listen, and how the file will be stored. If a participant wants anonymity, avoid putting their full name in the filename. A simple log can use a session code, date, location type, device, microphone, and consent status. Do not publish a raw voice recording by default; a public upload can expose a private conversation even when the interpretation is harmless.

Capture the raw file

  1. Use a known-good recorder and leave enough battery and storage for the full session.
  2. Record a ten-second room-tone sample before speech begins. Note unusual sounds such as refrigerators, traffic, clocks, or other people.
  3. Place the microphone consistently and do not move it during the session. If you change position, make a new marked recording.
  4. Speak a date, time, session code, and consent status into the file before asking questions.
  5. Stop the recording deliberately. A clipped or overwritten file is easier to identify than an ambiguous “best take.”

Keep the original untouched. Work on a copy, and record every trim, gain change, noise reduction, and playback speed. If a clip is shared, include the original session code and a note saying whether the file was enhanced. A clean-sounding clip is not automatically more authentic; processing can make a low-level sound easier to hear and also change its character.

Listen in layers

First listen without seeking or naming a message. Note the time, speaker, and exact sound you heard. Then play the same segment again with headphones and a lower volume. Finally, compare it with the room-tone sample. The Eerie Roads app page can help organize a location-based practice, but it cannot authenticate a recording or prove that a sound came from a person who is not present.

Use a vocabulary that is honest: “I heard a form that resembled speech,” not “the voice said my name.” Keep alternative explanations beside your interpretation: traffic, a pipe, a device, a cough, or a familiar word fragment. If a clip is emotionally intense, pause. Repeatedly replaying a frightening sound can reinforce fear without producing better evidence.

Share responsibly

Before publishing, remove names, locations, contact details, and unrelated conversation. Explain the recording method, processing steps, and uncertainty. A transcript should be marked as a transcript, and a translation should be reviewed by a person who understands the language. The EVP and spirit-box comparison guide can help you describe what was actually used.

Frequently asked questions

What equipment is required for an EVP recording?

A reliable recorder and a consistent microphone position matter more than an expensive brand. Document the device, room conditions, and any processing.

Can noise removal prove that a voice is real?

No. It can make a sound easier to hear, but it can also introduce artifacts. Preserve and disclose the original and the processing steps.

Should I publish the full recording?

Only with consent and after removing information that could identify a participant or expose private speech. A short excerpt with context is often safer.

When should I stop listening?

Stop if listening increases fear, sleep disruption, conflict, or compulsive checking. A recording is not worth sacrificing your wellbeing or someone else’s privacy.

Liked this article? Share it with fellow paranormal investigators!

Put this into practice

The catalogue is the practical part: seventy-two entries with a rank, a form and an office, and it is browsable offline.

Arcana Goetia: Ritual & Sigils, $3.99. The seventy-two entries, the sigil generator and the invocation guides, all offline. Android app, one-time purchase, no subscription and no account.

Get it for $3.99 on Google Play Complete Treatise on Magical Servitors, $9.99 What is inside Complete Treatise on Magical Servitors — $9.99

© 2026 Cha0smagick Labs. All rights reserved. | Home | Blog | NOCTEM App