How to Split a Long Recording and remove the Silences of eac

Audacity 2.1.3
Windows 7

Hi all,
first time poster and noob at audacity. I’ve searched all over this forum and google but can’t find a solution to my issue below:

OBJECTIVE: I’m trying separate a very large (over 10 hour) recording into multiple files (over 2000 separate files…yes a lot, which is why efficiency is key), with each file not having more than 0.1 second silence either at the start or the end. The separation points are predetermined by me, so I using Sound Finder is not an option, also I don’t want to get rid of all the silences (only those after the start marker label and before the end marker label).

WHAT I’VE DONE SO FAR: I play the audio at 2x to 3x the speed and I zoom, then I play the recording and Add Label at Playback Position (CTRL + M) when I hear when the recordings needs to be separated. However, frequently there is too much silence from the starting marking point to when the audio begins and too much from when the audio ends and the ending marking label (essentially, between the two label markers). I tried using Trim Silence plug-in (this plug is does exactly what I want, it only removes the silences at the start and end and not in between), but this only does it for the full recording (the first seconds of the full recording and the last seconds), whereas I need Trim Silencer look at each audio between two label markers as a separate file and apply to all 2000 splits.

First, am I in the right track with what I’ve done so for, or is there an easier/faster way of accomplishing this?
Second, how do I get rid EASILY of the silences in front and at the end of each split (note there will be over 2000 separate files, so efficiency is key here)?

Note, I’m fairly new so don’t assume i know anything. I think it can be done, i’m just missing basic concepts.

Thank you in advance.

I can’t help you with any automation/batch processing… I don’t do that kind of thing, because just about every project is different and some human interaction/judgment is usually needed, and it would take more time to “program” & debug than to jump-in and just get it done.

I tried using Trim Silence plug-in (this plug is does exactly what I want, it only removes the silences at the start and end and not in between),

Try [u]Truncate Silence[/u].

Audio editing (and video editing) is time consuming! If you listen-through once before editing and once-again after editing, that’s 2x real-time before you do anything, and without including the recording time. I’d typically expect to spend a MINIMUM of about 30 hours of “work” editing a 10-hour program… (Sometimes I can go faster if I’m only splitting a recording and I know it’s good so I don’t have to listen-through the whole thing.)

Doug, thanks for the reply. Conceptually, the solution is rather easy…I simply need Audacity to recognize each split and apply the Trim Silencer to each split.
I honestly believe there has to be some easy workaround to my issue.

The idea of opening over 2000 files and applying Trim Silencer is unfathomable.

I’ll reply here if i figure it out, but I’m hoping someone knows how to.

Thank you.

It is often helpful in finding a solution, if you describe what the job is.
What is this 10 hour recording?
On what basis are you (manually) selecting 2000 split points?
Why do you need to do this?

Hi Steve, the recording is a foreign class language room recording.
The basis of splitting points is where a paragraph or sentence or some other idea ends (this needs to be done manually, until they can invent software that can read my mind).

I think I’m very close to a solution:
Instead of removing the silences from the main 10-hour-long recording THEN exporting each split into an individual file, I’m instead going to FIRST export each split into an individual file THEN use Chains to remove the silences from all the 2000 recordings.
I just hope over 2000 files won’t overwhelm audacity (not sure what the file limit of apply chains is).

Unless anyone has any other ideas.

I’ll let everyone know if this is successful.

The Audacity Manual says: “It is recommended not to process more than 500 files at a time.”

See this page:

In the alpha Manual for the next release I’ve just made that advice a bit more obvious :sunglasses:



That’s exactly what I was going to suggest, but as waxcylinder noted, 2000 files is probably too ambitious.
Test the chain on one or two files first to check that it is working correctly, then try a batch of say 50 files, and work up from there, but not more than about 500 files at a time.