I have a problem with my neighbors dog. The dog barks for several hours every day. I need to provide authorities with an audio recording of the barking. The recording is the easy part thanks to Audacity. However, for maximum impact, I would also like to include a bark “count”. In other words, how many times the dog barked over the course of the recording. I originally counted the barks manually which was extremely time consuming since the sound file was 90 minutes long, but was amazed at the results. I counted 1937 barks in 90 min. That is powerful evidence!
Although the sounds (barks) are at random intervals, looking at the waveform and spectrogram of the audio clip, each bark is fairly well defined.
My Question: Is audacity capable of counting these “bark pulses” programmatically?
Thank you.
I need to provide authorities with an audio recording of the barking.
Did the authorities actually say you needed to provide an audio recording? Public nuance triggers tend be from the logged number of calls and number of different phone numbers, not the number of barks. I’m betting none of the authorities has any way to listen to a sound file at work.
Audacity tends to fall apart when asked to critically analyze sound signatures and then take some action. The content analysis tools, what there is of them, don’t lend themselves to programming and automation. Even something as simple as recognizing the spaces between songs on a record fails enough of the time that a lot of people give up and do it manually.
So no. Sorry.
How many neighbors call the authorities? Can you get up a little party, serve strong drink and make everyone promise to call, staggered so it looks natural?
Just from being able to speak a little dog, I can tell you that one thinks there is something very wrong which needs barking at. Do the owners go to the shopping center for hours and just leave him out there?
There are different complaint levels. Level 1 is a typical phone call. To elevate and accelerate the process, a video or audio recording is required by our county’s animal control dept. Based on the audio or video they send an animal control officer to the offenders property, and issue the owner with an ultimatum. That when the fines start, and they get larger for each offense.
Unfortunately for the dog, the owners are young, uneducated and leave the dog outside for most of the day.
Parties are allowed, but continuous unsupervised noise lasting for months is something else entirely.
Thanks for your advice, I’ll do some more digging.
You can try my syllable counter plug-in (Speech-rate).
put the the attached file into the plug-in folder and enable it with “Add/Remove Plug-ins” under effects, Generate, or Analyze. It will appear in the latter menu. rjh-syllable-count.ny (1.37 KB)
make the track mono(Tracks menu)
Amplify or normalize to 0 dB.
select a few seconds and count the barks manually.
Call the plug-in “Speech Rate” under Analyze and set the parameters. start with a threshold of -12 dB–this works at least with your sample file. The interval should be OK, I guess.
As soon as the output shows the correct number of manually counted barks, you can select the whole file and do the analysis.