A place to build and store humpback and orca signal catalogues (aka catalogs). This repo was created by Emily Vierling as part of the Salish Sea Humpback Vocalization Catalogue project.
- Specify a directory of audio files, each a FLAC file containing a single example of a signal.
- Pre-process the raw audio to optimize the spectrogram for human listeners.
- Generate a spectrogram of each FLAC file.
- Convert each signal clip to an audio format optimal for browser-base playback.
- Display the spectrograms with signal labels in a HTML5-based user interface that allows playback via the audio element.
Version 1:
- manually generated Wordpress portfolio page, served at this URL -- https://www.orcasound.net/portfolio/humpback-catalogue/
- source files (audio clips and images) stored in this repo (and shared with our OrcaLab collaborators)
- built by Emily Vierling in 2022 (with mentors Val and Scott Veirs, and some help from her sister!)
- More info about Emily's humpback catalogue and version 1 of this repo
Vision for version 2:
- Simple flat HTML site that can be generated locally and viewed/used in your favorite browser
- index.html with single example of each signal
- Individual pages with unique URLs for each signal type?
- Where to store and whether to display "extra" examples of each signal (limit how many?), and/or "dirty" examples (e.g. with low SNR or variable/aberrant versions?)
OR
- Leverage the open source code from the HALLO project's killer whale (3 ecotype) web-based catalog?
- How to deploy to a web domain like humpbackcat.net (and code/architectural implications --> containerization; Github actions and CI/CD?)
- How to also design repo to be cloned for easy offline browser-based use, e.g. while onboard a boat beyond cellular coverage?