Addendum to Field Log 001: The Recordings, Timestamped

Addendum to Field Log 001: The Recordings, Timestamped

Field Log 001 was published with one thing deliberately left open: the exact times of the six audio recordings, and which of them belonged to which waypoint. The recordings are the most important part of all of this — everything else in a field log, the photographs, the videos, the screenshots, the weather, exists to add context to them. The audio files have now been uploaded to the computer, and their properties read. This addendum closes most of what was open, and it is published as an addendum rather than folded silently into the first post, because the first post is already published and its open statement was true when it was made.

What a file timestamp on the recorder actually is

The properties of each file give a timestamp. The first question is what that timestamp means: the start of the recording, or its end.

It is the end — the moment the recording stopped and the file was finalized. Two independent facts establish this:

1. The overlap test. Read as start times, the last two files overlap: 261025_0241.wav would run from 16:03:58 for 4:21, ending at 16:08:19, while 261025_0242.wav would already begin at 16:06:38 — impossible on a single recorder. Read as end times, the same two files sit 32 seconds apart, sequentially, with no overlap anywhere in the day.
2. The 2:19 anchor. Field Log 001 records, from the day itself, that the first of the later recordings was made from the park at about 2:19 p.m. Read as an end time, 261025_0239.wav (ends 14:25:38, duration 5:52) begins at 14:19:46 — the 2:19 recording. Read as a start time, it would begin at 14:25:38, which matches nothing.

So every start time below is reconstructed: end time minus duration. This is stated openly because it is the kind of methodological detail these logs exist to keep.

The reconstructed timeline — October 25, 2026

All six files: 48,000 Hz, 2 channels, 24-bit — exactly the session settings Field Log 001 recorded. Their file sizes corroborate the durations independently: at 48 kHz / 24-bit / stereo, one second of audio is 0.275 MB, and every file's size matches its duration at that rate.

What the timestamps corroborate

• The dawn session. Both early recordings fall inside the dawn photographic run at W1 (first photograph 07:19, last 07:51), with a 4-minute-46-second gap between them. The dawn session was sound as well as image, as Field Log 001 states.
• No audio during the second photographic run. There are no recordings between 07:44 and 14:19. The 8:30–8:38 a.m. run was photographs only.
• The park recording. 261025_0239.wav is the 2:19 p.m. recording at W3, the Pavillon Isaac-Vandandaigue (the park) — the longest recording of the day, 5:52. The NoiseCapture screenshot at 14:26:47 was taken 69 seconds after it ended: record, stop, measure.
• The afternoon's shape. The 15:26–15:31 screenshots immediately precede 261025_0240.wav (starts 15:33:21) — context captured first, then the recording, which is the method working as intended.
• A loose end from Field Log 001 is closed. The day's last photograph, at 15:58:24, was unexplained — it fell outside every timed window in the first log. It was taken 73 seconds before 261025_0241.wav begins: a photograph of, or at, that recording session.

The Interference in The Park Event

Upon listening, the Field Logs from Sunday had interference. There was extreme distortion in several of the recordings, and none in others. 261025_0240.wav — now confirmed as the library recording, Waypoint W4 — is clean, and so are the dawn recordings at W1. The other three afternoon recordings, 261025_0239.wav, 261025_0241.wav and 261025_0242.wav, are distorted throughout. A 10-second sample from 261025_0242.wav was measured on October 26: the interference is a pure tone at exactly 1000 Hz with its harmonics at 2000, 3000, 4000 and 5000 Hz, and essentially no energy anywhere else in the spectrum; it swells and recedes in waves, and it is stronger in the right channel. The frequencies were not random. They were electronic noise — interference, surely.

Preliminary diagnostic testing showed it was not a Tascam DR-05X hardware failure, and not a corruption of some sort through the data transfer: the distortion is there when listening to the files on the recorder's own speakers. It was recorded that way, in the field.

Atmospheric Tone is just the recorded tone, though. I make no claim on the clarity of the recording itself. It's just an artifact. The artifact tells a story and it's the story that matters, not the artifact. It only gives evidence, but there also is no perspective or argument. It is anonymous, disinterested, and impersonal.

A distortion event is just that. It was what I am calling The Interference in The Park Event. In a sense, I started out going on a field expedition and ended up doing an investigation into a mysterious noise, a distortion that had a source I still don't quite know.

So I tested. You have to test things, otherwise you don't know — you can't know. Fieldwork, the kind that I do, is experimental, so you need to do testing: test first, the test-driven approach adapted to my interdisciplinary art-research practice. I did the two-recorder tests at home; they were inconclusive, or at least everything seemed fine. Then I went back, later yesterday afternoon, to the same park bench where I had recorded Sunday with the same recorder. I recorded one session with my phone on airplane mode, then another with the data on and all the apps running. Both were distorted, exactly the same way. And this has happened before, at this park — I remember now.

The bench is right by the library. There is also a stage there, where they often have concerts and things during the spring and summer months — maybe there is actually a source of interference in that. It's also next to town hall, so maybe there is a source there. And there is a cell tower maybe 300–400 meters from there, maybe a bit further, but not that far. The source is still unknown. What the tests have established is where it lives: not in the recorder, not in the transfer, not in the phone — at the site.

You never know where it's going to take you. It takes you down these rabbit-holes. I just follow the lead, I follow the recordings. Where the recordings go, I follow. It was like a forensic investigation of a mysterious source of noise. And I realized that I've always been fascinated by noise my whole life, and have been studying it for 30 years — in my music, in painting, in text, in everything, in my fundamental perception of the world as someone who does noise synthesis. Noise has a texture, a spectrum, a distribution.

This is Self-Reflexive Documentation, the Meta-Archive, Post-Archive, the Deep-Archives Project. Media archaeology of a distortion artifact. I literally found myself doing a forensic investigation. It was amazing.

And I may still make Atmospheric Tone videos with these recordings, distortion and all. I don't think it's my job to clean everything up all the time. Audio engineering, sure — if you're in a recording studio recording an album, you want everything to be super clean. But field recordings are not meant to be cleaned up too much. It might be hard to listen to with all the distortion, but in the end the important and critical part is to be true to what was in the field. And then do testing, and the accounting of what happened.

The last two files, placed

261025_0241.wav and 261025_0242.wav are placed at W3, the park. The basis is stated plainly, because the certainty here is of a different kind than the timestamps: I can't remember exactly, and I can't tell from the recordings themselves, since there is so much interference in them — but I'm pretty sure that's what it is. The corroboration runs the same way: the interference in these two files is the park's interference, the site-bound source established above, and the day's last photograph, at 15:58:24, was taken at that session, 73 seconds before 261025_0241.wav begins. Recollection, signature, and photograph agree.

UCS names, as far as they can now go

• 261025_0239.wav can be renamed now: AMBPark_PavillonIsaacVandandaigue_AG_FL001.wav
• 261025_0240.wav can be renamed now: AMBRoom_BibliothequeArmandCardinal_AG_FL001.wav
• 261025_0237.wav and 261025_0238.wav are settled as W1, Quai du parc Saint-Henri; their CatIDs — WATRFlow, WATRWave, or AMBBird for the whole dawn ambience — will be chosen by listening to what each recording actually holds.
• 261025_0241.wav and 261025_0242.wav, placed at W3, take the same CatID, with the session told apart in the UserData field: AMBPark_PavillonIsaacVandandaigue_AG_FL001_1600Session.wav and AMBPark_PavillonIsaacVandandaigue_AG_FL001_1600Session2.wav.

The methodology sharpens a little with each pass: capture the context, state what is not yet known, and let the files themselves — their timestamps, their sizes, their contents — corroborate or correct the account. The recordings come first; everything else is context for them. One note: it may seem strange that I wrote these first few Field Logs without really having listened to the actual audio files, but it's only because I am still setting everything up to run as smoothly as possible going forward. The audio content this time was secondary to the actual setup. Future excursions should not leave so many open questions, except for anomalies, like in Sunday's recordings, which I still need to probe. I also downloaded some new apps, Sensor Logger and the Physics Toolbox Sensor Suite which will allow me to add a great deal of sensor data from my mobile phone. Again, I'm still deciding on the exact methodology I will be using in coming field expeditions.

— A.G.