Posts by Cameron Dunn
Best Practices for Data Management
Introduction The introduction of the Machine Learning Detector into Raven Pro has made the analysis and interpretation of large-scale passive acoustic monitoring (PAM) projects easier and more efficient than ever. This article will suggest best practices for data management that will save you time and reduce analysis errors. Discussion When new users explore Raven, they…
Read MoreKeyboard Shortcuts
Note: The effect of some shortcuts depend on what is visible in Raven. A few depend on the last think you clicked on in the user interface. Windows Mac OS Action Area ctrl-a ⌘A Select All Edit Menu ctrl-A ⇧⌘A Save All Selections As File Menu ctrl-a ⌘A Mark all frames Selection Review ctrl-c ⌘C…
Read MoreWaveform Plot Appears Broken
Introduction Raven users sometimes report that the waveform plot appears broken, usually when opening 24-bit or 32-bit recordings. The problem is that the default Raven window preset, “Default 1.3 Power”, was designed for displayed 16-bit audio. There are several strategies for displaying 24-bit and 32-bit waveforms using a more adequate amplitude scale Solution #1 Activate…
Read MoreCorrelation – Counterintuitive Results
Introduction Raven users sometimes report that Raven Pro’s correlation feature reports correlations far different from what they were expecting. This article will explain some of the reasons for this. Explanation #1 By default the correlation feature correlates the waveform view. If you want to correlate the spectrogram view, be sure to select “Spectrogram”. Explanation #2…
Read MoreMachine Learning Detector – Adding New TensorFlow Models
This document describes the requirements of metadata and TensorFlow model format for Raven Pro to integrate a TensorFlow machine learning model.
Read More