Capturing Electronic Voice Phenomena

COMMENT??? Got A Question??? A Click on the Wizard opens a direct E-Mail Link to the Research Center. Technical questions or just want make a report, just CLICK the Wizard!

For purposes of this discussion I will assume the researcher is simply going to monitor a single room of a typical old house for EVP. This room has minimal furnishings and power is present in the house. The researcher arrives in the afternoon and has never seen the place before. He is here because the owner has claimed the place may be haunted. No other evidence has been brought forth at this time. I am going to touch on each aspect of the investigation and how they are conducted, concentrating on the EVP portion.

But first, the investigator should interview the witness and obtain any background that may aid in placement of sensing equipment later. Are there cold spots? Have apparitions or any thing unusual ever been seen? Any audible noises heard? If so, in what areas of the room? Was anything going on in or around the house that might account for them? Obvious questions intended to simply validate the claim of haunting. For purpose of this discussion, we will assume that the obvious has been ruled out. No TV on in another room, no cars outside, none of the things that might be mistaken for an EVP. No sense continuing an investigation if the cause is something easily explained.

So now we are ready to instrument the room and see what we can pick up. Place your recorder and all other equipment in another room, as far removed as possible from the target area. This room should be in an area of the house that is powered from a separate circuit than the one which supplies the target room. Next, locate the service entrance for the house and turn off as much power as possible. Power lines do radiate electromagnetic fields which could account for a false positive. Leave on only what is needed to operate your equipment. If there is a furnace or air conditioner in the house it should be shut down because changes in temperature can cause extraneous noise as expansion and contraction takes place. This should be one of the first things done to allow temperatures to stabilize. Also, in the case of forced air systems, air movement itself might be mistaken for EVP if it creates any noise or causes something such as the drapes to blow across a window sill. In short, it is imperative to first make the house as quiet as possible to minimize the possibility of a false indication. As is becoming evident, almost every action taken is intended to control every conceivable cause of extraneous noise. Once this is done, you're ready to actually set up to monitor for EVP.

In an average room, about 14 by 14, I would use four mikes as pickups. Two of these would be dynamic mikes; that is ones which use a coil within a magnetic field. The others would be electret type condenser microphones. The reason for the two types is because we are not really certain of the mechanics of EVP. Some feel it is simply an audible sound that is picked up, while others speculate that it may actually be a form of energy imparted to the microphone. By using two different types, a broader spectrum of possibilities is covered. Future research may clarify this, but for now I would recommend covering all possibilities. The coverage of each of these mikes should be a cardioid type. That provides sufficient directional characteristics without narrowing the pattern to an extent that portions of the target area are not covered.

Regarding location of the mikes in the room, I would place a pair, one dynamic and one electret in two adjoining corners angled toward the center of the room. If, during the course of interviewing the witness, he indicated a certain part of the room had greater levels of activity, I would position one set of mikes pointing toward that area. However actually placing the mikes within that spot may not be advisable since it may disturb whatever is responsible for the activity. Better to be close, but not intrusive.

One more thing to consider. If there is physical activity reported, it may be a good idea to also place a video camera in one corner of the room. The use of a wide angle lens would allow visual observation to take place while the room was monitored for EVP. The would accomplish two purpose.; First it could provide collaborating evidence should EVP be detected in conjunction with other disturbances, and second it would affirm that no one entered the room and attempted to hoax activity. Video should definitely be used if at any time during the investigation a researcher plans on entering the target area.

Once the mikes are positioned and all equipment located there in place, the mike cables are connected to the mixer located in the observation area. Four channels should be used, and the mixer adjusted to provide a two channel mixdown. The set of mikes located in the left corner should be sent to channel 1 (left) and the set on the right to channel 2 (right) inputs of the recorder. The reason I recommend a true stereo mix is to allow any EVP to later be analyzed for time discrepancies. This would permit one to determine the location within the room where the EVP originated. Another benefit is that with two separate channels improved noise reduction techniques may be applied to bring out any faint signals which may be captured.

Regarding noise reduction, I generally recommend bandpass filters on the mike lines ahead of the mixer. These will narrow the pass band to that considered voice frequencies. By doing this, out of band frequencies containing only unwanted noise will be suppressed, resulting in a clearer EVP should one be detected. However, it must be noted that this does limit frequency range. You may want to have the filters switchable so you can try it both narrow bandwidth and full bandwidth. The type of EVP you get would determine which is preferable, there is no way to know in advance which way to go. With a little luck you will be set to the one you want when the EVP occurs.

Speaking of noise, some investigators place white noise generators in the area claiming that this aids the EVP in coming through. I do not recommend that method. First the additional noise simply masks any low level audio you might record. Secondly, any noise added to existing hiss or noise being generated internally by your equipment could mix or heterodyne with the random white noise. This can result in spurious signals at the sum and difference points between the noise at any given instant. The possibility exists that such spurious signals may be mistaken for EVP. This is why I attempt to remove as much extraneous noise as possible leaving only the desired signal.

Technique is another factor to be considered. Some simply set the equipment up and start recording. Others want to go in and attempt to converse with whatever is there. On any investigation I do, I require everyone to be accounted for at all times. That means everybody stays in the observation room while monitoring is taking place. It eliminates the creation of false EVPs by some investigator talking near the area under study. I also leave a second recorder running in the observation area simply to record the comments and thoughts of the investigators as monitoring goes on.

Sometimes though, it is good to try to converse and that means that someone will enter the room under observation while monitoring is going on. If that is the case, that person goes alone and they enter the room moving into view of the camera. At all times while they are there they remain visible to the other investigators via camera. They may speak as they feel led, talking clearly and loud enough to easily be heard. Under no conditions are they permitted to whisper, and they must also keep physical movement to a minimum. Recording continues the entire time they are in the room. I also log the time they enter, and the time they leave the room for future correlation with any EVP captured.

That about covers how the field investigation is conducted. The work comes later, when the tapes must be listened to, sometimes multiple times, to see what, if anything was captured. Recordings are best evaluated using a good pair of headphones, not simply ear plugs. You should get comfortable, in a quiet area, and play the tapes back. Take a break every 15 minutes, since the constant white noise present on the tape can tend to cause your attention to drift, resulting in a missed EVP.

Signal processing is an area of controversy. Common sense dictates that the least amount a signal is altered makes the signal more credible. But sometimes filtering can bring out a voice from the background and make it intelligible. So should you use processing? Personally I do use some pass band filtering at times. But always the original recording is maintained and only a copy altered. Another consideration is the method of processing used. I use a graphic equalizer for filtering. I do not use digital processing since that method actually alters the signal's properties by changing wave shape. The graphic equalizer simply removes unwanted frequency bands which contain only unwanted noise. You can selectively remove these bands to lower the noise level without drastic changes to the recorded signal.

If you should get an EVP and wish to isolate it for later use, such as posting on the web, be sure to not segment it too closely. Generally I would recommend that you include at least 3 seconds of the tape before, and another three seconds after, as part of the clip. Two and a half seconds to play a two second EVP does not allow the listener time to properly attune his ears to the background.

Finally a little about the equipment itself. The mikes I covered previously so I won't repeat that here. The bandpass filter should cover the range from 200 Hz to about 3 kHz. This is the range of normal human voice, male, female and children. If you want to be even more selective, you may use narrower pass bands. Voice is comprised of two components, vocal and fricatives. The fricatives are similar regardless of whether it is a child, a male, or a female voice. They are the sounds made by the tongue and teeth, "S", "T", and "K" are some examples. The pass band should be between 1.5 and 3 kHz. The vocal component differs somewhat. For a male voice, the recommended pass band is between 200 and 400 Hz, a female pass band would be 300 to 600 Hz. Children generally are between 350 to 700 Hz. The bandpass filter can be customized to whichever you are trying for, or simply use the wider 200 to 3 kHz range for all. The gain in performance of the wider band is noticeable, but not to a great degree.

The mixer is simply a standard audio mixer board such as used in small public address applications. The scenario described here uses a four channel version, which should do in most applications. One important capability it requires is to be able to mix to a stereo output. (pan pots) Some smaller ones do not have this feature.

The recorder can be any two channel (stereo) unit with good noise specifications. It should be housed in a steel cabinet as opposed to plastic since metal provides better shielding against noise and outside interference. such as power lines. I recommend a standard analog recorder, not digital. Digital systems require A to D conversion which in some cases can result in intermodulation distortion and aliasing.

Lastly, an area of controversy. Tapes may be new or reused, but only if they are bulk erased, not simply erased by the recorder. However ALL tapes, even new, should be bulk erased before use. There are residual fields left from the manufacturing process. Put a new unrecorded tape in a good quality stereo and turn the bass and the volume up. Hit play and listen for a "whomp, whomp" sound as the tape runs. Those are residual magnetic fields. They will cause false EVP when you combine them wit the steady hiss you get in the field. Your recorder erase head will not eliminate all of them. For the same reason you can't completely erase a tape for reuse. The erase head and the record/play heads do not track perfectly. That is why, when you reuse a tape you can still hear a faint voice left from the previous recording, even though you recorded over it. The bulk eraser, properly used, will eliminate both problems since it erases the entire tape, not just the narrow bands where the program material is recorded.

Hope this has been informative and hope you get many good EVPs in your quest!