Michael Esposito/Per Svensson


Sound localization relates to the listener being able to identify the origin of a specific sound signal by direction and distance. This passive acoustic location is based on a sound or vibration created is the object being detected. Active acoustic location creates a sound or vibration producing an echo, which can be measured for location. This is also known as echo location. One of the characteristics of Electronic Voice Phenomena is the lack of sound localization. In most recordings, human voice, sounds from animals, machines and nature have a sense of sound localization. Sound localization as we know it is a well studied biological effect. The human auditory system uses many cues for sound source localization such as spectral information, correlation information, timing analysis and analysis between time and level changes between both ears. This lateral information is analyzed by our ears in a very specific manner. Time changes are mainly observed in frequencies below 800 Hz. As sound from the right side reaches the right ear sooner than the left ear the interaural time difference is measured in phase delays for low frequencies and group delays at higher frequencies. Level changes are mainly observed in frequencies above 1600 Hz. As sound from the right side reaches the right ear sooner than the left ear the interaural level difference is measured at a higher level. Both time and level differences come into play for sounds between 800 Hz and 1600 Hz.

Human voices are roughly in the range of 80 Hz to 1100 Hz. The voice frequency or audio frequency used for the transmission of speech is between 300 Hz and 3400 Hz. Below 80 Hz on the auditory level, it is virtually impossible to use either time or level difference to determine a sound's lateral source. The directional evaluation between the phase difference between the left and right ears becomes too small to tell any difference. Intelligible speech, where we should be able to speak words that can be heard, starts at about 200Hz. Human speech is a known spectrum and with the help of the perceived sound distance can be roughly estimated.

In determining sound localization in the median plane; front, above, back and below, the shape of the outer ear plays a keen role. The outer ear, including the external ear canal, creates direction-selective filters that create filter resonances. These outer ear transfer functions create and implant resonance based direction-specific patterns to determine localization.

One of the main auditory system clues we use to determine sound localization is reflection. The listener is subjected to two types of sound. The direct sound is a sound that a listener hears without being reflected off of some type of surface. A reflected sound is a sound that has reflected off of at least one object. Sound localization can be gained by the ratio of the direct sound to the reflected sound or echo. If the reflective object is too close to the direct sound and the listener, the sound created will be a reverberation versus an echoic reflection. In the case of reverberation, the phase difference overlaps with the direct sound and the difference will be too small to gain any useful localization information. An echo on the other hand, creates an independent sound, completely separate from the direct sound. Our auditory system will only analyze the direct sound, which arrives first, for sound localization. This is known as the law of the first wave front. This echo cancellation physically occurs in the Dorsal Nucleus of the Lateral Lemniscuses, allowing sound localization even in echoic environments.

In EVP, there is no direct sound source. Atmosphere is manipulated creating interference and air pressure too small to register a direct sound. Is it possible that it does create a reflected or echoic sound that can be picked up by recording devices? If an entity can only push wavelength towards an object and all we register is the reflected sound, then echo cancellation would cause the sound we hear upon playback to appear to have no sound localization. In order to determine direction, a strong attack in several critical bands of our hearing range must be experienced. This strong attack from a direct sound must prevail over reflections, which arrive later. Directional cues from the direct sound become unstable from the reflected sounds, which add no loudness inside the critical bands. So, no new directional information is triggered by the auditory system. One of the reasons the voices we capture seem to come from everywhere at once may be because what we are listening to is their echo and without the direct sound response with which to measure it, we cannot determine any sound localization. These psycho acoustical phenomena would be akin to an anti-audio masking where the absence of one sound affects the perception of another sound.In Ghosts of Ogilvie Station, Per Svensson and I look at just these phenomena. In the Ogilvie Station, Chicago, originally built in 1911, there are many ghosts. There is a great deal of sound localization also. Each of the 16 tracks has an individual speaker that allows a female voice to repeatedly call out the track number. By walking in front of, behind, in between and among the different speakers at varying paces, I was able to generate an almost unearthly sense of binaural simulated sound localization. In different areas of the station, I was able to capture EVP. This was well away from the speakers calling out the track identities. On the recording, the EVPs sound flat and as if coming from no direction at all, while the almost Vocoder like female voices of the loud speakers sound not only very specific in their direction but crash into each other and compete for the listener's attention often creating new sounds and even a sort of musical sound. To further illustrate the purity of difference between the sounds with the passive acoustic location and the EVP seemingly without, Per uses an Oscilloscope, OSCILLOGRA?PH UO963 NORDMENDE, to measure the frequency relationships and provide an audio output through its tone generator. The result is a very clear exercise in how we experience the audible and inaudible world around us.

Michael Esposito

In his own observations of the exercise, Per Svensson offers the following:
To define something of a different kind. The not yet defined.

...The acoustic material or the field recordings from Ogilvie Station I received from Michael Esposito is of a nature that describes both absort of Sound Architecture and/or the sound structures of a different material in a large room.The sound material even describes various Distances - as you can experience with the transportation of sounds in this specifically unshaped room via echo where Undefined Voices randomly occurs.

In a sound space like this, the sounds seem to bounce on some material and return, even move backwards and split into clusters. The sound material appeares to be recorded with a moving recording device that registers the 3 dimensional planes, or even 11 dimensional planes, if you like, in this space or the building.

This at first is a sound sensation and a feeling of leaving or moving inside the actual space.
After a while, an image occurs. A transparent sound image. A see-through sound.
A split image formed into one. Different times at the same time.

The audible senses are placed in a specific but undefined space that is presented to a poly dimensional location. Images appear as the ear and the eye are connected to these imaginary, but not yet defining landscapes. Differences may occur whether the sender and receiver are still or moving. As with the Doppler effect, the frequency of a sound changes to your ears as it is the sound that is moving towards you and away from you, but with the same actual frequency sent out. Therefore this undefined spacial sound could define a change or a movement in the actual undefined space. The hearing or registration of the Ogilvie material is to me defined as moving sound registration. Vice versa the doppler effect? As if you travel and listen and see at the same time, these dimensions defines this sound material.

On this flexi recording I have both recorded and played the original sound recording from the Ogilvie Station with my Yamaha MT8X portable cassette recorder, and at the same time run the sound through my Oscillograph Nordmende from 1958 and returned the modulated sound back again to my MT8X and recorded the frequency modulation of the material on a different channel. In this way, the frequency modulation is directly linked to the basic sound material of Ogilvie Station.

The Oscillograph UO963, Normende (fr. 1958)

With this device I am able to see frequencies of sound electronically. When the Ogilvie material was run through the Oscillograph imaginary, undefined images appeared from the sound material. (see photo). An undefined shape occured on the screen as I was modulating the frequencies of the sound on my Oscillograph.

The modulation of this recorded material was made in various Hz frequenzy levels from Z-B-10-35-150-500Hz and in KHz frequency levels 1,5-5-15-50-150-600Khz. In the actual sound modulation into the synthesis of the sound there is no sound location, but pure sound frequencies.

Sound is used for registration of certain spacial forms to create and form images and memories that form a description of a place, a voice, a location, or certain senses...or in its most interesting form - to define something of a different kind, the unknown, ...something not yet heard or not even experienced before. The spiral rotation of sound. Sound Images. Fractals.

Per Svensson

Echo Voices/The Ghosts of Ogilvie Station is part of an ongoing collection of essays and observations into the nature and experience of Electronic Voice Phenomena. The collective, entitled "The Phantom Monographs," will have future releases which will be available over time and may be assembled in a final completed monograph.