Friday, April 3, 2015

EVP and the Voice of Reason

This is an article I wrote for The Bent Spoon Magazine a couple of years ago:

Once upon a time, there was a wannabe ghost hunter.  She watched TV shows featuring paranormal investigators going into haunted locations and capturing real ghost voices on their recorders.  Finding this incredibly cool, she visited websites where ghost hunters from all over uploaded creepy recordings of spirit voices.  She bought a digital recorder like the ones she saw on TV and did her own EVP (Electronic Voice Phenomenon) experiments.  She lived in a house where a previous owner died on the dining room floor.  Lights went on and off by themselves, faint disembodied voices and footsteps were heard and unexplained shadows were glimpsed out of the corner of the eye.  So obviously, it had to be haunted.  She wanted to prove to others that the ghosts were actually there, and she also wanted to hear what they had to say.  Why were they there?  Were they "stuck" from unfinished business?  Were they attached to the house or something in it?  So, just like the investigators on TV, she held her inexpensive recorder and asked questions.  On playback, she was excited to hear responses.  It was hard to make out the words, but as ghost hunting experts will explain, sometimes the spirits just don't have enough "energy" to speak clearly.  One night, she got a reply which sounded more like a snarl.  It scared her, and after stinking up the house with burning sage, she stopped doing sessions in her own home.

Yep, that was me several years ago. Back before I took the time to learn about recorders, recording techniques, what environmental factors can affect recorders, and what physiological and psychological factors affect how a person can misinterpret sounds.  Luckily, I can laugh at myself now.  But what isn't funny is the fact that there are paranormal investigators going into people's homes or businesses and, because they are making the same mistakes I once made, presenting frightened clients with false positives and calling them ghost voices.  As I mentioned in my article "The Evocative EVP" ( while more ghost hunting groups are finally acknowledging that there are natural explanations for orb photos, many of these same people are still clinging to their EVPs with a death grip.  I believe this might be because listening is more subjective;  you can easily see how orbs are recreated, but replicating false positive EVPs may be more complicated due to various factors.  There have been reliable scientific studies showing that people hear things that are not there.   One study, discussed in Mary Roach's book Spook,  illustrates this and is relevant to EVP review.   Subjects were asked to transcribed a poorly recorded lecture.  Many were able to hear words and even complete phrases.  However, in reality, the recording was nothing but white noise.  Ambient sounds can easily be misinterpreted as voice, especially with priming and when they are within certain frequencies and rhythms causing the brain to automatically switch to speech mode.  Personally, I've participated in many audio reviews where people swore they heard a meaningful response when all I heard was something akin to "Glarmpht".  So even if something sounds like a voice or a phrase, it doesn't mean that it is.  And even if it is, you still have are left with the task of proving that it belongs to a ghost.

Let's address some basic recording protocols that many paranormal teams fail to implement.  If the idea is to capture proof of ghosts by recording their voices, one would think every measure to prevent false positives would be considered and used.  But unfortunately, this often not the case.  In some cases, it is due to inexperience and lack of knowledge, and in other cases, it is plain confirmation bias.  I've personally observed investigators continue to use poor recording techniques, even after explaining to them how they create false positives.  I'm not entirely sure if it is because of a egotistical defiance ("Shaddup, I know what I'm doing!") or they subconsciously sabotage their efforts to continue to capture "evidence" and satisfy their wish fulfillment.   When I wanted to become a paranormal investigator, I started to network with people who have some experience in audio engineering. I quickly learned why some teams get a bunch of EVPs every investigation, but others capture only one or two possible EVPs a year.  I'm sorry to say it wasn't because ghosts like some people more than others.  It is simply because some teams use common sense to reduce false positives.

Conducting an EVP session while holding a recorder and walking around is one of the best ways to create false positives, yet it is commonly seen on ghost hunting shows.  Setting a recorder a few feet away and sitting as still and quietly as possible during a session makes much more sense.  The fewer investigators participating, the smaller the chance of recording noises made by living people.   Investigators should be aware of and note their surroundings and ambient noise, since it can be mistaken for speech. This goes beyond merely tagging every sound.  One tip I got from Midwest Preternatural Research is that before each EVP session, conduct a controlled silence session to listen and record the ambient sound for later reference.  Investigators should have a notebook at all times.  If an anomalous sound is heard, document the time to cross reference other audio, video and notes of other teammates.  Many teams invest in video cameras in hope of capturing a full-bodied apparition holding a sign saying, "Hi, I'm a ghost".  However, video is better put to use to record investigators while they are doing an EVP session for cross reference.  It is amazing how shifting positions can sound like a whisper, a bored sigh can sound like a "hey", etc.  Surveillance cameras should be recording in other areas, again to cross reference "responses" or anomalous sounds.   The problem is, most teams don't have enough cameras to cover an entire building and its immediate surroundings.

In the unlikely event that there are cameras covering every angle at all times, does that mean an EVP is really a ghost?  Nope.  One issue is radio frequencies.  Cheaper recorders have less shielding from RFs.  I have mentioned this elsewhere before, only to have testy ghost hunters tell me it is "impossible" for recorders to pick up radio frequencies.  Since I'm not an expert in recorders or radio, I've talked with radio technicians, audio engineers, and a forensic scientist, who assured me that yes, recorders can indeed pick up radio frequencies.  Some people have hypothesized that some EVPs are actually ELF (Extremely Low Frequencies) picked up by a recorder. ELFs are frequencies ranging between 3 and 300 Hz and natural sources include lightning strikes and disturbances to the earth's electromagnetic fields.  Alternating currents in electric power grids fall within this range and electromagnetic interference from high voltage lines are around 50 - 60 Hz.  These EMIs are unintentional sources of ELFs.  (Other sources of EMIs include computer servers, medical systems, robotics found in industrial use, and faulty wiring.  EMIs can be picked up by recorders.) One anonymous reader of my blog sent me an email stating he or she could rule ELFs out because they built a Faraday cage for their recorder.  A week or so after getting this message, I tuned into a web radio show where in the chat room, listeners were arguing whether Faraday cages can block natural electromagnetic fields and ELFs.  So I was prompted to contact a physics professor at a local university for an answer.  He sent me this reply: "The "skin depth" decreases with the frequency of the wave and increases at lower frequencies. Therefore, it is virtually impossible to completely block very low frequencies as you need infinitely thick material. You need to find a balance between how much of the low frequency wave you want to block and the thickness of the Faraday cage you can afford."   So a basic home-made Faraday cage may block some radio waves, but not ELFs. 

On a side note, I recently heard a ghost hunter on web show state that microwave ovens can be used as Faraday cages. After conducting an informal experiment involving cell phones, various microwave ovens and some stink-eye looks from a home improvement store employee, I have concluded that this is false.  While the metal screen in the door may block AM and even FM radio waves, the cell phone signals are sneaking through.  According to the electromagnetic spectrum chart, the AM radio frequency band is between 535 kHz - 1700 KHz, the FM radio frequencies run between 88-108 MHz, cell phones are higher up the spectrum at 824-849 MHz.   The metal grid in the glass door is ineffective from blocking these smaller wavelengths.


Now let's consider paranormal investigators who claim they can identify an authentic EVP from a human voice by using spectral analysis software.  Doing a brief search, I find many of them are using free downloaded software for their analysis.  Because of the poor track record of photographic analysis from ghost hunters in general, I wondered if there was a chance there could be some misinterpretation and misidentification in some of this analysis as well.  I have seen audio experts on TV who analyze audio for court cases.  I was certain there had to be some formal education and training involved.  So I searched for such an expert and found Dr. Stevan Pausak.  His occupation is listed as a Forensic Scientist - Applied Physics and Physical Chemistry. His areas of work include: audio examinations, video examinations, electric and electronic examinations, evaluation of the forensic reports and interpretation of their findings and court testimony in the above areas.  (For a full list of his education, training, and publications please visit:  I sent him email, along with links to various paranormal teams' websites and YouTube videos on spectral analysis.  He sent this reply:  "The spectral analysis is a mathematical tool used by physicists, astronomers, chemists and other scientists and engineers in their basic studies and practical applications. This analysis is applied to mechanical phenomena as sounds and vibration and electromagnnetic waves as light, microwaves,radio waves, x-rays, gamma-rays and so on. The mathematical aparatus, called FFT (Fast Fourier Transform) is at the heart of the modern spectral analysis.  To fully understand and apply the spectral analysis one needs special courses in mathematics and physics provided at the university level.  The MRI, Ultra Sound, CAT scan and some other medical imaging devices as well as radars, sonars, spectrographs (for different parts of electromagnetic spectrum), sound spectrographs and so on, all have the use of spectral analysis. It is not necessary to understand spectrum and spectral analysis in order to operate these equipment or even to interpret the results. Thus the technicians that operate them in general do not have an understanding of a spectrum and spectral analysis. Even MDs that interpret the MRI and other medical imaging results do not have understanding of the spectral analysis.  The so called 'audio engineers' are, in general, technicians with some understanding of the sound spectrum but not the spectral analysis. There is a lot of room for misunderstanding and misinterpretation of the sounds and its sources."

He generously invited me to call him so he could answer any further questions.  During our conversation, he stated that in his opinion, the average ghost hunter, without specific education and training, is not qualified to properly analyze spectrographs.  More importantly, even if they are able to identify a human voice, all they have is evidence that they recorded a human voice - not proof of a ghost.  Leaping to such a conclusion does not follow valid scientific methodology.  He explained that acoustics play an important part of what and how sounds are recorded as well as what we may (or may not) hear at the time.   Sounds (including voices) can carry through pipes and ducts.  He went on to explain that sound propagates differently through different materials.  It does not propagate through air as well as it does solids, such as soil and rock.  An example of how well sound can travel for miles is railroad tracks:  if you stand by the track, you won't hear a train if it is miles away.  But if you put your ear to the rails, you can.  Another example is water.  It is possible to hear someone whisper from the opposite side of a river that is a kilometer wide - as long as you're close to the water.  So during investigations, it is possible to record a living person's voice that has travelled from a different area.  It may be faint, and while investigators may not hear it at the time, the microphone on a recorder is able to pick up and record it.  This information is supported by an article "Analysing EVP and Paranormal Sound Recordings" by the Association for the Scientific Study of Anomalous Phenomena:

The article also describes how the "Cocktail Effect" can explain how some investigators may hear a sound at the time while others do not. People get used to background noses at various rates. So while some may have already "tuned out" some of the ambient sound, others may not.  In addition, if a faint voice is "buried" in the ambient sound, it may not be noticed by investigators at the time, whereas a microphone can detect it.  On the other hand, because many microphones are directional, they can also miss sounds that investigators did hear at the time.  This explains away one of my own audio clips that puzzled me for a while. During one case, I heard voices from an unknown source.  The other investigator with me in the room also heard them.  The other two investigators at the location were in a different room.  When I called them, they said they hadn't been talking, nor did they hear the unidentified voices.  I had two recorders operating in different areas in the room at the time.  One picked up the voices (faintly), the other did not.  For a while, I mistakenly thought this was an anomaly and possibly paranormal.  But based on the information about acoustics and microphones above, there is a natural explanation for this phenomenon.

Are people recording ghosts from beyond the grave? I want to believe it is possible. But wanting something to be true doesn't make it so.  Until all other possibilities can be methodically ruled out, I'm not convinced that EVPs are really voices of the dead.

Sunday, March 1, 2015

Believing Is Seeing

It is said that seeing is believing. But, beliefs are NOT facts.  When someone says, "I know what I saw", it is inaccurate.  They know what they believe they saw.  Only a portion of the visual process involves the eye, the rest is perceived in the brain in a series of complex processes.  Furthermore, our expectations influence that perception.  

Studies suggest that even language and cultural expectations influence how we see our world.  Various cultures have different words and categories for colors, and actually see them differently from others.   Various experiments with the Himba tribe have supported this.  English, like many of the world's major languages, has 11 basic colors terms.  Himba has five, each covering a wider range of colors.  For example, "zoozu" is a term used for various dark colors that we would call darker blues, greens, browns, purples, reds or black. One experiment involved showing Himba subjects a color wheel of green tiles and having them pick out the one slightly different shade of green. While generally, Westerners had more difficulty selecting it, Himba subjects identified it right away.  Interestingly though, a new color wheel was presented with all green but one blue tile that was obvious to Westerners.  However, the Himba had difficulty identifying it.  (This experiment with the Himba tribe is included in the following video: )
Recently, we saw how people's perception of color varied when a picture of a dress went viral.  Some saw it as white and gold, while others saw it as blue and black.  Scientists who study color explained the discrepancy was due to how our eyes evolved to see in daylight and how our brains tried to match it to the daylight color gradient, filling in the details for us.  (Personally, I saw it as blue and gold.)

When we believe in ghosts and told a place is haunted, and then further primed with stories of specific apparitions, shadow people, or strange lights, the likelihood increases that we will misperceive what we see to fit our expectations.  Many ghost hunters still "investigate" in the dark.  While it makes for entertaining TV, it also creates so many issues hindering valid research. As I've mentioned in other posts, our eyes are not well adapted to see in the dark.  When our brains try to make sense of what our eyes see poorly in a dark environment, it "fills in" details that aren't really there.  But even in bright conditions, a person's beliefs and expectations will influence what they think they see.  Priming also explains how different people may "see" the same thing at a location.  

In another post, I mentioned watching a TV program which showed an interesting experiment at Loch Ness.  On one side of the shore, people were asked to witness the researchers launch a log into the lake.  They asked the witness to draw it.  Not surprising, they drew a straight log.  But then, the researchers asked people on the other side of the lake, who were not privy to the log launch, to draw what they saw.  Interestingly, the straight log now grew a neck and head, similar to what the Loch Ness monster reportedly looks like.  Were these people lying or dimwitted?  No, they saw a shape in the water across the lake, a group of people watching said shape, and then their expectations about the notorious location filled in the rest.

Priming also enhances pareidolia, or matrixing, in photos and videos.  How many times have we seen paranormal pictures of blobs with a caption of "little girl" or "demonic face" and then a whole lot of viewers post "I see it too!"  But in reality, there is nothing there.  We are hard wired to see faces in random patterns.   It is an evolutionary tool that helped our species survive.  To use Michael Shermer's (The Believing Brain) example, when our ancestors were strolling along a savannah and saw what looked like face within the tall grass, they immediately had to decide whether it meant danger or not.  To proceed without knowing whether it was a random pattern made by the grass or a predator, would be detrimental to survival.  Through experience, our brain evolved to err on side of caution.  

Pareidolia can make you see faces in this picture.  But one face is really there, and wants to meet you for lunch.
Source:  ArtWolfe/Caters News

In her book, Eyewitness Testimony, Elizabeth Loftus explains how our brains do not record an event like video tape.  The process is quite complex.  There are three stages of how we form memories of an event.  The first is how we perceive an event.  As discussed above, we see there are various factors which influence how we can misperceive what we see.  The next stage is retention.  During this stage, discussing, or overhearing someone else's account of the same event can unintentionally manufacture new details and dramatically change the witness's memory.  Finally, there is the retrieval stage where some time after the event, the witness is asked to recall specific details.  Obviously, if there are flaws in the first two stages, the retrieval stage will be inaccurate.  And even if the event was perceived accurately, details can be forgotten, leading to failure of retrieval.  Furthermore, the conditions at the time of retrieval influence the accuracy.  Some factors include what type of questions are asked, how the questions are worded and who is asking them.  While Loftus's book focuses on the reliability of memory in criminal cases, it can be applied to other situations, including paranormal claims.  Like witnesses of crimes, many witnesses of seemingly paranormal experiences report it caught them unexpectedly, feeling fear, and "it happened so fast", all which can increase the likelihood of misperception.

All this does not mean paranormal researchers should dismiss eyewitness accounts out of hand.  As scientific paranormal investigator Ben Radford explained in his book, Scientific Paranormal Investigation, it IS a valid starting point for an investigation.  However, if a researcher is there to conduct a productive investigation, it is important for them not to take the account at face value, because as discussed above, there are many factors complicating how people perceive and interpret experiences. Many times clients call in paranormal investigators to validate their experiences.  I've seen paranormal team websites boast things like "We believe you" or "You're not crazy, we'll help" and similar slogans.  While I agree that it is helpful to reassure a sincere client that they experienced something and aren't necessarily going bonkers, it is NOT helpful to encourage them that it was paranormal, without providing scientific, valid evidence.  When we label ourselves investigators or researchers, we should strive to understand how the human mind processes the natural world before we offer presumptions about the supernatural one.


Denise Grady, Discovery Magazine, "The Vision Thing: Mainly in the Brain"
 June 01, 1993

Rachel Adelson, American Psychological Association, 
"Hues and views: A cross-cultural study reveals how language shapes color perception."
February 2005, Vol 36, No. 2  Print version: page 26

Adam Rodgers, Wired, "The Science of Why No One Agrees on the Color of This Dress"
February 26, 2015

Elizabeth Luftus, Eyewitness Testimony
Cambridge; Harvard University Press, 1996

Benjamin Radford, Scientific Paranormal Investigation
Corrales NM ; Rhombus Publishing Company 2010 

Michael Shermer,  The Believing Brain 
New York:  Times Books  Henry Holt and Company, LLC,  2011