Hello everyone,
I’m Nishant. We have released a tool called Spectrolipi (https://spectrolipi.com) — for visualizing sound, annotating spectrograms, and exporting ML‑ready acoustic datasets. Please try it and share feedback.
Key features
- Spectrogram playback with adjustable colormap, gain, and zoom
- Annotation boxes: create, edit, import, export
- Quick tools: SCC (Spectrogram Cross‑Correlation) and Repeat annotation
- Species selection with predefined or custom lists
- Export ML‑ready clips with CSV or COCO‑style annotations
- Manage file level Metadata
Local processing so raw audio stays on your machine
Link to the tool: https://spectrolipi.com
Feedback and issues: https://github.com/nishantnnb/spectrolipi/issues
FAQ: https://github.com/nishantnnb/spectrolipi/raw/main/Guide%20and%20FAQ/Spectrolipi%20-%20FAQ.docx
12 April 2026 6:27pm
Hi, this is a very interesting tool. We are currently working with camera trap datasets in Mindoro and starting to explore the acoustic layer within those recordings, particularly for nocturnal species like the Mindoro Boobook.
One of the challenges we are anticipating is the validation step after automated detection, especially when dealing with long recordings and relatively sparse vocal activity. A lightweight interface for reviewing spectrograms and confirming detections feels like it would be very useful in that part of the workflow.
Curious how well Spectrolipi handles longer audio files or batch workflows, especially when the goal is to move from raw detections to a curated dataset. Looking forward to exploring this further.
Kristoffer Noriel Ordoñez