Cutting video based on the spectrogram of its audio

I have a lot of video footage. I want to cut interesting parts into a smaller video. I have found the best way to find interesting parts is looking at the spectrogram of the audio and find regions with interesting patterns in it (eg. the talking part, the music part, silence etc.). Currently I'm using Audacity, I load the audio of the video in there, scroll through the spectrogram, select the interesting part. But Audacity does not have a feature to copy the selection start and length times to clipboard. So I have to look at the selection toolbar and manually copy the timestamps into a text file, then I use a script that turns that into an ffmpeg command line to cut the video. Needless to say this is time consuming and is very error prone. There must be a better way to do this. So I'm looking for either a video editor that allows me to load the footage and then switch to spectrogram view to pick the interesting parts. This would be nice because sometimes it would be nice to see the image too to make a better cut. Or an audio editor with spectrogram view that allows me to copy selection timestamps to the clipboard in a format that ffmpeg understands. That would faster and less error prone, than manually copying them.