Créer un compteSe connecter

Recherches récentes

Aucune recherche récente

Options de recherche

Disponible uniquement lorsque vous êtes connecté.
mastouille.fr est l'un des nombreux serveurs Mastodon indépendants que vous pouvez utiliser pour participer au fédiverse.
Mastouille est une instance Mastodon durable, ouverte, et hébergée en France.

Administré par :

Statistiques du serveur :

429
comptes actifs

mastouille.fr: À proposAnnuaire des profilsPolitique de confidentialité

Mastodon: À proposTélécharger l’applicationRaccourcis clavierVoir le code sourcev4.3.4

#birdNET

0 message0 participant0 message aujourd’hui
Public
Ronald

Dafür, dass ich hier quasi im Wald wohne habe ich viel zu wenig Ahnung von den Tieren um mich herum. Mein Vater kennt sich da besser aus, und als der noch hier gewohnt hat konnte er sich ziemlich dafür begeistern dass wir hier einen #Pirol haben.
Rein akustisch würde ich den nicht erkennen, und gesehen habe ich ihn auch nie. Scheint aber noch da zu sein. 🤗
#Wald #Tiere #Fauna #Vogel #BirdNET

Screenshot aus BirdNET.
Public *
Jon Sullivan

I'm off to bed but my computer cranks on. It's using #BirdNet to ID all the bird songs recorded by our garden's #AudioMoth from the first half of the week, while an R script is hooked into AWS to transcribe the 772 geotagged audio notes I made today on my monthly 12 km run up from the city through Victoria Park and around the native forest on Sugarloaf peak.

Some parts of our crazy Big Tech timeline I'm OK with. 😄

A screenshot from my Mac showing a window for the BirdNet app, which has started processing 233 15 minute long consecutive recordings from the AudioMoth recorder in our garden. It's identifying all of the bird songs in the recordings. The other window is R, showing progress transcribing my wildcounts audio notes from my ecological surveys this week. It's on its 7th survey, which has 772 audio notes from my hill run this afternoon. The smaller window in the bottom right shows progress in copying the 3.5 days of AudioMoth recordings (13.19 GB) across to our NAS for storage.
#wildcounts#EcologicalMonitoring#nz
Suite du fil
Public
Jon Sullivan

I was puzzling over the weak seasonal pattern in the tauhou in my BirdNet graphs, since they're more common in the winter in our garden. I carefully checked my R code. Sure enough, I found an error on one line, where I'd added the BirdNet seconds onto the start time from the filename. Turns out adding seconds onto time is a bit slipperier in R than I'd thought.

Here are the greatly improved graphs (up to yesterday's recordings).

A graph showing the trends in bird singing per day for four native birds that visit our garden in Christchurch, New Zealand. From the top, there's the tauhou (silvereye), korimako (bellbird), tīrairaka (South Island fantail), and riroriro (grey warbler). Tauhou are common in our garden and peak in winter months. Korimako swing up and down with years and seasons, most common in winter months, but show no clear long-term trend. Tīrairaka appear to be more common in winter months and vary a lot. Riroriro were more common in the early months than later months. Both tīrairaka and riroriro are detected much less frequently than the others and it's hard to see their data on the graph as I'm using the same scale for each species. The y-axis is the average (mean) daily number of detections per 20 minute interval (where one detection is whether or not a bird was detected in 15 seconds).
A graph of the relative proportions of the bird songs each day made by different species in my garden in Christchurch, New Zealand. I've filtered out the rarities and obvious mis-identifications. The graph goes from 20 June 2021 to 5 January 2025. By far the most common bird is the orange on the graph, which is tauhou/silvereyes, which have a prominent winter peak. The next most common is house sparrows, followed in winter months by korimako(bellbirds) visiting our sugar water feeders, and then blackbirds.
#birds#EcologicalMonitoring#BirdNet
A répondu dans un fil de discussion
Public
Jon Sullivan

@lightweight @yojimbo By the way, here’s what the unit looks like in our garden. It’s in a sistema case with a small hole by the microphone covered with a sound transparent water proof mesh sticker. It takes three AA batteries. I use some older rechargeable batteries which in the summer (longer days) I swap out and recharge every 4 days.

A photo of a small plastic case strapped to a fence in a garden. Inside is a green rectangular piece of tech that is an AudioMoth that records sound. There is a small hole cut in the plastic case covered in a sound transparent water proof sticker. Under this is where the microphone is.
#AudioMoth#BirdNet
Public
Jon Sullivan

Since June 2021 I've had an #AudioMoth constantly recording bird song in our garden from 2 hours before sunrise to 2 hours after sunset. Over the past months I've been using #BirdNet to identify all the bird species.

Today I've managed to spit out my first graphs. I'm excited to finally see results so wanted to share them. There's all sorts of things happening.

Next I'll be comparing these automated patterns with all the counts I make manually.

A graph of the relative proportions of the bird songs each day made by different species in my garden in Christchurch, New Zealand. I've filtered out the rarities and obvious mis-identifications. The graph goes from 20 June 2021 to 5 January 2025. There's quite a lot of variation day-to-day but surprisingly weak season patterns and no big long-term trends (that I've noticed so far).
A graph showing the trends in bird singing per day for four native birds that visit our garden in Christchurch, New Zealand. From the top, there's the tauhou (silvereye), korimako (bellbird), tīrairaka (South Island fantail), and riroriro (grey warbler). Tauhou and constantly common in our garden. Korimako swing up and down with years and seasons but show no clear long-term trend. Tīrairaka appear to be more common in winter months and vary a lot. Riroriro were more common in the early months than later months. The y-axis is the average (mean) daily number of detections per 20 minute interval (where one detection is whether or not a bird was detected in 15 seconds).
#birds#EcologicalMonitoring#nz
ExplorerFlux en direct
À propos