These pages
are intended to present work in progress for discussion and will be appended
whenever possible. Mistakes have been, and will doubtless continue to
be made. These will be included, and corrections added where necessary.
Some knowledge of Avebury, acoustics and archaeology is assumed, so it will not be explained here. I am not an archaeologist. I have some training in acoustics and I happen to live very close to Avebury, but I'm not trying to prove a theory as I don't have one. I've simply noticed that repeat echoes exist inside the Avebury Henge, and that this might possibly be an intended element of the design. By displaying my findings here, I hope to provoke some discussion and ideally, learn more about Avebury.
These echoes may well be just an accidental effect, but there is reason to suspect otherwise. William Stukeley, as early as 1720, noted that the stones of the Avebury circle were arranged with their smoothest sides facing the inside - as were the stones of the two inner circles. At Stonehenge the stones have been dressed, but again, they are smoother on the inner side. Why should this be? There may be some other reason lost to history, but the fact remains that the flatter, smoother, inner surfaces of the stones do reflect sound extremely well and can produce clear repeat echoes. The echoes of human voices are not very loud but sharp sharp percussive sounds and short whistles echo well.
Listening
to Avebury's acoustics is made more difficult by the busy and very noisy
road that passes through the middle of the Henge. Most of the stones are
missing; houses have been built inside the henge that reflect sound; there
is a busy pub and a good deal of military aviation noise. It's often very
windy. Trees reflect sound extremely well, especially if growing in clumps,
and the two groups of mature Beeches at the eastern and southern entrances
to Avebury reflect sound as efficiently as a cliff face.
Work is already underway on building complex computer models of Avebury's
acoustics, with removable buildings & trees; but it is still possible
to get worthwhile results simply by walking around the Henge at a quiet
time and listening.
Recording is best done early on Sunday mornings, or late at night. This
can be difficult and time consuming, so I aim to work my way round the
Henge in short sections, presenting the results on these pages as I progress.
EQUIPMENT:
The 16 bit digital recordings are made on a Sony TCD-D8 portable DAT recorder, at 44.1 kHz, using modified omni-directional 'tie-clip' mics mounted inside my ears. The resulting binaural recordings can produce fully three-dimensional surround sound when played on stereo headphones.
My sound source is a pair of Claves, or 'click sticks' made of ebony. When knocked together they produce an extremely loud and short pulse of sound that is ideal for triggering echo effects. Because the Claves are so loud, they appear distorted on the recordings - recording level has been set for the echo return, not the source. The DAT recordings are transferred digitally to a Soundscape DAW for editing; the length of echo delays can then be timed extremely accurately. Soundscape's graphic display doesn't work in milliseconds though -it uses a timecode format of hours, minutes, seconds, frames and samples, so it requires some calculation to convert the delay times to ms. Running at 25 fps, 16 bit, 44.1 kHz means that each frame contains 1764 samples.
By a combination
of measurement on the ground (using an odometer) and timing of the echoes,
it is often possible to pinpoint exactly where particular echoes are coming
from.*
From my first tests it is apparent that it is the stones themselves that
are reflecting sound - but in some parts of the Henge that are without
stones, sound does seems to be reflecting from the earth bank. Each recording
has been documented by taking photos, and by marking position on a large-scale
map - one copied from Isobel Smith's book 'Windmill Hill and Avebury'
(1965).
* As shown later, this turned out to be a rather over-optimistic claim. The odometer I used was not accurate enough for these calculations - in this instance, the map was far more accurate.