0

I am looking for a way to code or find a program that can record the chirps of crickets either live or through a prerecorded audio file (large ~24 hours) for a lab experiment.

I'm not too sure how to approach this as I'm a web developer, but I have experience with JS and python, along with libraries. My initial idea was to use Matplotlib to produce an audio visualizer, and then count each time a certain range of db is reached which matches the db of a cricket chirp, but I have no idea how to approach it.

I have successfully visualized the chirps on a online spectrum analyzer (Spectrum Visualizer of Audio Chirps), and can see it clearly, however I don't know how I can use code to count each "chirp" and record it along with the date and time for each chirp in a table of values / dataset of some sort.

Any guidance or help would be greatly appreciated!

Jayson Tian
  • 1

1 Answers1

0

Really naive solution: For every unit of time (column of pixels in the spectrogram), you can calculate the sum of all the values in a column. For example, if all the pixels in a column are black, the sum for that column will be 0, if some of them are colored, the sum will be >0. Then loop over the sums: if at a certain step you go from 0 to > 0, and then after a while you go back to 0, you've hit a peak. If the output is too "noisy", you can use a small threshold value instead of 0. Tweak it until results seem OK. This works well when the input background noise is pretty much the same all the way through, but if it's skewed up and down during the whole recording, you need something more complicated (for example, the threshold constantly changes with the average of the last n sums)

Or you can Google "peak detection algorithm" and implement one of them. Or you can Google "peak detection library" in your favourite language and use one of them.

janesconference
  • 6,333
  • 8
  • 55
  • 73