Batch find silences in talk radio programs


I manage a Christian talk radio station, and sometimes we find (after the fact) that a program we acquired and aired had a production error: there may be a gap of silence in the middle or it might play a while and then the rest is silence. We don’t have the staff to screen every program, so I’m looking for a batch process that will find silences (say, > 5 seconds) in mp3 files.

In Audacity, I have played with Silence Finder and with the Extend Silences plugin and developed two possible workflows, but neither is very good:

  1. Create a chain that contains only Silence Finder and End. Print a list of all mp3 files to be analyzed. Run the chain on the files. (Each time Audacity analyzes a file and does not find silence, it pops up an error message that must be clicked on to dismiss. If it does find silence in a file, it adds labels, then closes the file with no message and moves on to the next one.) Every time Audacity doesn’t find silence and pops up the error message, cross that file out of the list. So we’re left with a list of the files with silence.

  2. Create a chain that contains Extend Silence and Export mp3. Files that contained silence will be exported to the Cleaned folder, but the first time it encounters a file with no silence, the Chain stops and does not process any subsequent files. Since most of our files don’t contain silence, we’d be restarting the chain over and over.

Can anyone suggest improvements on either of these ideas, or, better yet, something much better?

Thanks! :slight_smile:

Using Audacity 2.1.2 on Windows 10. I don’t remember if it was installed with the .exe or .zip.

What do you want to do with those that have silences?
If you want to reduce the amount of silence, why not just use Truncate Silence in a Chain? (

That won’t work because Audacity’s batch processing (with Chains only works on audio files, not projects, but labels only work with Audacity projects, not audio files.

Well, it doesn’t work very elegantly, but it gets the job done in a labor-intensive way. We would end up with a list of files that contain long silences.