by Steve Moon
An audio recorder deployed in a large river valley in eastern Iowa – digital photo by author
The Power of Transcription
Quantifying data in an audio file and preserving it in a transcript is as natural in my sound file analyses as making entries in a journal. A transcript is the most direct method that I have to document the Sasquatch behavior that I discover in soundscape recordings. I build transcripts slowly as I listen to a sound file or group of files, from the beginning to end of each analysis. Listening may be continuous or partial. Partial analysis involves using the free software package Audacity to visually scan a sound file as a spectrogram image, searching for signatures of relevant sounds. I’ve gotten good at visually identifying knocks and vocalizations, and other relevant sounds after many years of practice. But continuous listening reveals far more detail and nuance in the soundscape, providing a more complete accounting of the activity captured in an audio recording. Also performed in Audacity, continuous listening is very time consuming, but I do it routinely for my more critical file analyses. Here I give an example of the use of both continuous analysis, which involves listening to a sound file without interruption, and partial analysis, which involves listening to selected portions of a file. Both are demonstrated below to be an appropriate approach to sound file analysis, depending on the recording environment, and on the investigator’s time restraints.
The process of transcript writing has greatly facilitated my research practice. By describing carefully each event as it is encountered, I gain a deeper understanding of the sounds that I have captured. A transcript which results from continuous listening provides a comprehensive account of the events in a soundscape, whereas a partial review will include only a few of the highlights. The overarching purpose of either method is to enhance the investigator’s ability to recognize and understand Sasquatch behavior as it is captured in a soundscape.
My transcripts organize both data and ideas, representing the soundscape and observed behavior in a hybrid fashion that is specific to my peculiar way of thinking, and which taps directly into my creative processes. Interspersed with sound entries are occasional sentences or paragraphs that provide a back-story, or historical perspective; a non-audio related personal impression of what is going on in the landscape. In this way my ideas continually evolve, changing organically as though I am putting pencil to paper.
Broadening My Analytical Reach
As my research agenda and overall technical approach to audio field recording have evolved over time, so has my use of transcripts in the critically important role of documentation. In the partial transcript below I illustrate a new system that I have recently devised of classifying sounds on a loudness scale, and by geographic orientation. I long ago became frustrated with continuously relying on terms like “faint”, “quiet” or “loud” in transcripts to describe sounds. Quantifying these sounds through the use of a classification scheme seemed a no-brainer, and a step in the right direction. While general descriptions are valuable and perform a necessary function, they don’t reflect the true reality of a soundscape. Quantifying two major attributes of sound, loudness and geographic orientation, which two-channel recordings provide access to, creates a more accurate record. Descriptive terms reflect the investigator’s experiences and direct knowledge of events, and remain a vital part of any transcription. Pairing these with a system of classification makes the transcript a more robust survey instrument, and a much more accurate record of events.
My loudness rating scheme utilizes a scale of 1 to 10 for each sound, with ten being the loudest. 1 can barely be heard, with or without processing. In transcription, loudness is designated with an “L” preceding the loudness value, as in “L5”. The perceived geographic orientation of a sound in the soundscape is designated with an “L” for left, a “C” for center and an “R” for right. The geographic designation is placed directly after the loudness rating in transcription, as in “L5R”. More nuanced descriptions of sounds are noted in each entry line, as in “center left L2L”. The loudness and geographic designators are located at the end of entries for ease of visualization, as in “00.02.30 faint knock center L2C”.
All loudness values are determined subjectively at the time of initial listening, and during subsequent listening sessions. It could be argued that these values are arbitrary, but they preserve the investigator’s personal interpretations, which in my case are based on hundreds of hours of previous experience analyzing sound files. As an investigator gains experience with sound file analysis, ability to quantify loudness quickly improves, to the point where placing individual sounds on a ten-point loudness scale is reliably accurate. The same may be said of quantifying geographic orientation.
Classification of sound data facilitates a more complete analysis of sounds within a soundscape, enhancing the transcript’s usefulness as an analytical tool. Interpretation and visualization of patterns in Sasquatch behavior are discoverable at a glance when there is a set of designators at the end of each sound entry.
Beyond these nuts and bolts ideas about transcripts and the classification of sound is the underlying inherent power of the transcript. Approaching file analysis from a new perspective, whether radically different or just a tweak, can suddenly take your research into new and productive territory, opening new doors onto discovery and enlightenment. Creating a transcript provides many doors through which to explore, as the following transcript fragment and discussion hopefully illustrate.
A Transcript from an Ongoing Investigation
I typically deploy an audio recorder early in the evening, or even during the middle of the afternoon. This is often a matter of convenience, but also insures that I don’t interfere with ongoing activity later in the day and night. My research preference these days is for passive observation rather than active interaction, although I’ve done plenty of both. My audio recorders are powered by rechargeable power banks and will record for forty-eight hours continuously. The ZOOM F-3 field recorders that I use are set up to record 1 hour and 33 minute long files consecutively. Analysis of sound files begins at a time or file of my choosing. Unless circumstances dictate that I do otherwise, I usually begin my sound file analysis during what I consider to be prime time for a given location. In the case of the location being investigated here, any time after about 8:00 PM and before 3 or 4:00 AM, are considered by me to be prime time. I started this analysis with file _005, which begins at 9:21 PM.
231210 back deck looking north
_005 9.21 PM
00.02.30 faint knock center L2C
00.03.00 loud knock right L8R
00.04.57 knock left L3L
00.06.39 knock right L5R
00.07.37 as above L5R
In the above transcript fragment, if my assumptions are correct, the following trends are evident: For 8 minutes, between 9:21 PM and 9:29 PM, sounds emanate from at least three positions in the recorded soundscape; far left, far center, and near right. The recorder is looking north, so knocks emanate from a position fairly close to the recorder to the east (R), and from a distance in locations to the north (C) and west (L). Knocks are temporally spaced in the following sequence: knock 1 - 1.5 min. - knock 2 - 2 min. - knock 3 - 1.5 min. - knock 4, 1 min. - knock 5. Knocks are on average 1.5 min. apart.
Digging Deeper
I have investigated this rural farm property off and on for the past ten years. My audio recordings and witness interviews have revealed that Sasquatch visit here on a very regular basis, at all times of the year. From direct experiential knowledge as an investigator I’m able to draw a few conclusions from the transcript fragment above. I know that a dense hedgerow trends due north from the recorder position, and is the probable source location for the sound in the center stage (C). I know that the knock from the left (L) probably emanates from a ten acre strip of set aside and wetland about two hundred feet to the west, or from behind the recorder position to the southwest (L), along the same hedgerow that is the probable source of the knock from the north (C). The knocks that emanate from the right are more frequent and louder (L8R, L5R and L5R) than those from the left (L3L) or center (L2C).
Looking northwest into the set aside on the west side of the area under investigation – digital photo by author
An actor (assumed to be a Sasquatch) on the right (R) is in much closer proximity to the recorder than the actors on the left (L) or in the center (C), and seems to be leading communication with the other actors in the group. The sounds on the right could be from at or near the east edge of the farmyard, or from across the short side of an eighty acre field, inside of another set aside. The L5 and L8 rated knocks from the right could be from within 100 feet of the recorder, or from more than 500 feet from the recorder, or both. It’s possible that L 8 emanates from the farmyard, while L5 and L5 emanate from further away, from the east set aside. In this case there could possibly be two actors on the right, making the group total of as many as four individuals.
When there is a lot going on in a soundscape recording, such as in files 004 and 005 from this investigation, I tend to listen to critical sounds multiple times. Repeated listening rewards an investigator with a lot of “Ah ha!” moments. Listening to the knock at _005 00.06.39 (L5R) for a third or fourth time delivered just such a moment. I suddenly realized that there is a second knock on the left 1.5 seconds after the L5 rated knock from the right. The second, quieter knock (L2) is most certainly an acknowledgement, or response to the first knock. I have recorded many instances of this type of call and response pattern in Sasquatch communication. It is as though the second Sasquatch is providing reassurance to the first Sasquatch, saying “I got your back!” or simply “Copy.” With a sound value of L2, the second knock isn’t difficult to hear, but because I was concentrating on other things I missed it.
Other revelations were gained as I listened repeatedly to the L5R knock at _005 00.06.39. I felt that the knock on the right almost certainly comes from the yard, and not 500 feet away, and the quiet response knock on the left most certainly comes from the west edge of the farm yard, or from just inside the set aside. The last two notions are purely subjective. However, I strongly feel that I have gained a deeper understanding of the events taking place in the soundscape as a result of repeated listening. When doing a deep dive into a sound file the file analysis is never really complete. There is always more to discover and to learn.
_005 00.06.39 A knock from the right (L5R) acknowledged with a knock on the left (L2L) 1.5 seconds later
unprocessed sound file - 231210_005 00.06.39
00.06.39 knock right L5R
00.06.41 knock left L2L
Visualizing Activity within a Landscape
The loudness of a sound at its source varies widely, as does the effect of distance on that sound, as is evident in the continuation of the transcript fragment below. In the case of L8R, L5R and L5R above, there are probably two or three point sources for these, somewhere in a distance / loudness / geographic orientation continuum in the landscape, to the right or east of the audio recorder. Or perhaps just one point source. Given the unpredictable variation inherent in data captured with a two-channel audio recorder, sound sources are impossible to pin down accurately in a landscape, so out of necessity are best visualized as circles encompassing areas.
_005 continued from above
00.10.19 unknown metallic object making noise close possibly on deck L7L
00.10.43 as above but brief L6L
00.18.02 possible distant gunfire L3C
00.21.07 possible quiet knock L2R
00.27.09 quiet knock close center left L4L
00.30.00 pause review due to abundance of anthropophony
In sound file analysis and transcription, the investigator who captures the audio being analyzed should be familiar with the source areas of captured sounds. Photos of a study area, archived in the same folder as the sound files and transcript, are the best and easiest method for jogging memory about of the layout of a place. I photograph landscapes with my cell phone, which captures metadata about the geographic location, and the date and time that each photo was taken. I also photo document any artifacts that I find, such as the “asterisk” stick structure below, found in the far northwest corner of the study area in a dense hedgerow of trees that runs north and south for a half mile in either direction. Stick structures have been built back in that corner consistently, and have been numerous. Structures have been found over the entire property, and large objects have been moved. It’s an active place!
A stick structure in the form of an asterisk, found in the far northwest corner of the study area in a dense hedgerow of trees – digital photo by author
Northwest corner of the study area, looking west through hedgerow trees to the set aside – digital photo by author.
It was a typical winter day when I investigated this local rural farm in December of 2023. A gas furnace in the old farmhouse at the center of this property was making a lot of noise as it vented gasses to the outside air. I had forgotten about the noisy furnace when I deployed my audio recorder on the back deck of the house. The furnace vent, and a plethora of other human-made noise, or “anthropophony”, made listening to sound files from this overnight investigation an extremely challenging, make that painful, process. File processing was of little help. I paused my total listening review of file _005, and decided to load up the file that preceded it, thinking that I would listen carefully to the last thirty minutes or so of that file to see if there is a prequel to the above activity.
I loaded file _004 and opened it in a single window view (1 hour and 33 minutes) in Audacity. I opened the window as a “spectrogram”, and in the “spectrogram settings” set the “Max Frequency (Hz):” to 5,200 Hz. This allowed me to scan for sounds that fall well above the typical 1,200 Hz window that most investigators, including myself, use as a standard spectrogram view for file analysis. It pays to look at the higher frequencies, because whistles routinely fall above 2,000 Hz. It didn’t take long for me to discover more activity, and it was of an unexpected nature. Several anomalies became apparent, and my investigation proceeded in a new direction. I switched from doing a total listening analysis to doing a partial analysis.
File _004 in its entirety, viewed in Audacity with a maximum frequency of 5,200 Hz
In this window it can be seen that the offending house furnace cycles on for fifteen minutes, four times during the 1 hour and 33 minute long sound file. Furnace quiet periods are just under 11 minutes in length. My attention was first drawn to the signature at the 33 minute mark. Expanding that portion of the sound file led to the discovery of an unusual sequence of sounds.
_004 7.48 PM
00.33.12 three metallic knocks spaced .4 seconds apart center right L5R
00.33.15 whistle L9C
An unusual whistle occurs 1.8 seconds after a set of three metallic knocks that are evenly spaced at about .4 seconds apart. Because these sounds are isolated in the sound file as a single group of events I have to assume that they are related. If my ideas regarding the sources and geographic orientation of these sounds are correct, the knocks and whistle are not produced by a single Sasquatch, but are instead an example of communication between two Sasquatch. The timing and context of these sounds indicate to me that this is an example of Sasquatch intergroup communication. I am unable to identify the whistle, but have heard and recorded three similar, although much shorter, examples of the same style of whistle during investigations in the past two years. The above interpretation is based on my own field investigations, and on data that I have previously captured and analyzed in audio recordings over a period of several years. The online app “Merlin”, maintained by the Cornell University Ornithology Lab, was unable to identify the whistle as being produced by a known bird.
_004 00.33.12 - three metallic knocks spaced .4 seconds apart (L5R) followed by a whistle (L9C)
unprocessed sound file - 231210_004 00.33.12
Continuing my partial review of file _004, I discovered what sounds like a tree or tree branch lying on the ground noisily being manipulated on the right. This occurs 1 minute prior to the metallic knocks and whistle at 00.33.12, above.
_004 00.32.17 - tree or branch is broken or otherwise messed with (L9R)
unprocessed sound file 231210_004 00.32.17
00.32.17 - tree or branch is broken or otherwise messed with L9R
At 00.20.46 a whistle similar to 00.33.12 occurs, ending 1 second before a very quiet knock in the center of the soundstage. This whistle shares many characteristics with the previous whistle, but takes a different form, further leading me to believe that these are not made by a bird. The accompanying knock is also compelling evidence that it is not a bird. I feel that these sounds are, as with the knocks and whistle above, communication between two Sasquatch. Again, this interpretation is based on my field investigations, and data that I have previously captured and analyzed in audio recordings. Note that the cacophonous furnace is not blowing at this time.
_004 00.20.46 a whistle similar to 00.33.12 (L9C), ending 1 second before a very quiet knock in the center (L2C)
unprocessed sound file 231210_004 00.20.46
00.20.46 whistle similar to 00.33.12 L9C
00.20.48 quiet knock center L2C
Nothing else of interest was noted while performing a partial analysis of file _004. Combining the events that I observed in both files, I am able to construct the following larger transcript fragment, describing an overall sequence of events that occurred during the hour and a half that falls between 8:14 PM and 9:48 PM.
231210 back deck looking north
_004 7.48 PM
00.20.46 loud whistle center L9C
00.20.48 quiet knock center L2C
00.32.17 tree or branch is broken or otherwise messed with L9R
00.33.12 three metallic knocks spaced .4 seconds apart center right L5R
00.33.15 loud whistle center L9C
_005 9.21 PM
00.02.30 faint knock center L2C
00.03.00 loud knock right L8R
00.04.57 knock left L3L
00.06.39 knock right L5R
00.06.41 knock left L2L
00.07.37 as above L5R
00.10.19 unknown metallic object making noise close possibly on deck L7L
00.10.43 as above but brief L6L
00.18.02 possible distant gunfire L3C
00.21.07 possible quiet knock L2R
00.27.09 quiet knock close center left L4L
Glancing at the newly assembled partial transcript, I see that for 12 minutes, between 8:14 PM and 8:21 PM, there are two loud and unidentifiable whistles that occur in the center of the soundscape (L9C and L9C). These are both accompanied by knocks (L2C and L5R). One minute prior to the second whistle, a tree gets messed with at close range, to the right or east (L9R). The next hour appears to be relatively calm, although a real time listening survey could prove otherwise. An hour after the whistle and knock sequences, a series of knocks occur over a period of 25 minutes. The tree being messed with at 8:20 PM approximately corresponds in geographic orientation to the loud knock at 9:24 PM. A Sasquatch may have been hanging around for an hour on the right, or easterly side of the soundscape, or repeatedly moved to within close proximity of the audio recorder from that direction. It’s also possible that two or three different Sasquatch approached the audio recorder from the east at different times.
The two whistles seem to emanate from the center of the soundstage, at fairly close range (L9C and L9C). Two quiet knocks are captured in the center of the soundscape (L2C and L2C), at 8:16 PM and 9:24 PM. The first of these two knocks occurs during the period when the loud whistles occur, within one second after the first whistle ends, also in the center of the soundscape. This indicates the simultaneous presence of two Sasquatch in the central area of the soundscape.
The possibility exists that there was a much larger group of Sasquatch present than just three. An important attribute of Sasquatch behavior that is on display in these sound files is constant back and forth communication. The Sasquatch that are present in the landscape seem to be there with purpose, and their activities are well coordinated. It was a group project. Questions arise such as: Where did they travel to the study area from? Why did they appear on this particular night? Do they make return visits often? Do they represent members of the same family or clan? What is their purpose here? What attracts them to this place?
Studying transcripts from future investigations of this property will almost certainly reveal as yet unrecognized patterns in Sasquatch behavior, and perhaps even provide an opportunity to unravel a few of the mysteries of Sasquatch culture. All of the events represented in this partial transcript include specific loudness and geographic orientation designations, which made events easier to visualize and begin to understand. The classification system is an easy way to quantify and catalog the loudness and orientation of each sound in a recorded soundscape, providing a much more accurate summary of what occurred in the landscape. Had I used terms like “quiet” and “loud” to describe sounds without accompanying classification, I would have been unable to create an accurate summary of the events that occurred in the study area on the evening of December 10 of 2023.
__________________