Using lossless cut to batch remove silence at the end.

[_Al_]

It is for mp3 files. Using lossless cut.
Is it possible to do that for many files at once?
In tools/detect silent scenes it can be set. But lossless cut somehow uses all segments after.
And no idea how to export in batch. Also possibly lossless. If anyone uses this soft and knows how to do it, thanks.

[videoAI]

Shouldn't AI answer questions instead of asking them ..

LosslessCut does not have a direct, built-in batch silence removal feature for
multiple MP3 files. However, there are several alternative approaches:

Workaround Options:
1. Manual Batch Processing
- Open multiple files in LosslessCut's batch list
- Cut silence from each file individually
- Export each file separately

2. FFmpeg Script Method
- Use a bash/PowerShell script to automate silence removal
- Leverage the FFmpeg command that LosslessCut uses internally

Alternative Tools for Batch Silence Removal
- Audacity: Has macro capabilities for silence trimming
- SoX (Sound eXchange): Command-line audio processing tool
- GoldWave: Professional audio editor with batch processing
- mpTrim: Specialized MP3 trimming software

I recommend creating a PowerShell script to batch remove silence.

It is for mp3 files. Using lossless cut.
Is it possible to do that for many files at once?
In tools/detect silent scenes it can be set. But lossless cut somehow uses all segments after.
And no idea how to export in batch. Also possibly lossless. If anyone uses this soft and knows how to do it, thanks.

[_Al_]

It is _Al_ not _AI_
anyway I can google too, thanks

I was using:

ffmpeg -y -i input.mp3 -af "silenceremove=stop_periods=-1:stop_duration=1:stop_threshold=-50dB"  output.mp3

but that was not lossless, sure if there is ffmpeg cmd that would be helpfull, no problem to construt batch.
It looks trickier, maybe a timecode needs to be found, and then just using ffmpeg to just slplit file. Before Id try to do this using ffprobe, maybe that losslesscut can do that as well. Not sure.

[ricardouk]

Hi, im going to suggest a differente aproach

1- install the free davinci resolve video editor
2- drag and drop all of the mp3s on the timeline, select all the clips, right click on a clip and select create "new compound clip" to join all clips in one clip.
3- go to "menu"--->audio operations and select "ripple delete silence", it will detect all the silence "bits" and a pop up menu will show up where you can customize the silence duration etc etc
4- export the timeline to a big wav file

hope it helps.

[_Al_]

Videoeditor is not a problem. That recommended davinci resolve silence remover feature is interesting, but not usable. Separate mp3's are needed.

I kind of solved it in that lossless cut, figuring out that app segment logic, exporting losslessly! an mp3 without silent end. But this way all files need to be treated separately. Wondering if there is a batch capability somewhere. Though thinking it is not possible, the way that lossless cut effect works. It finds silent segments and creates segments. In our case silence at the end. That needs to be removed before export. There might be silence at the beginning or even in the middle and those segments should be kept. It needs manual approach. But we need just to cut silence at the end for a mp3.

[videoAI]

I came up with a solution for you ..
What is it?
I am sure you know how to Google stuff ..

anyway I can google too, thanks

[_Al_]

The way lossless cut does it, it gets silent parts and all timecodes, it gets segments. It shows those segments in the right with timecodes as well, it looks really nice. Those timecodes are stored in app's project file in seconds (llc). Then allows to export needed segments. This needs to be done file by file. That's not possible to work with huge mp3 folders. But anyway, losslessCut looks like a nice app.

So to automatize it, not using losslessCut, using whatever windows batch, python, I need to come up with ffprobe command lines to get those timecodes and then using ffmpeg to get those segments or maybe not needing ffprobe and getting those segments right away. I have to look into this.

Or if someone can help with those commands, that be nice. Only getting rid of silence at the end needed.

[pcspeak]

I did a very short test!
The silence duration and decibel parameters may need tweaking.
Does this work for your needs?

if not exist New\ md New
for %%a in (*.mp3) do ffmpeg -i "%%a" -af silenceremove=start_periods=1:start_silence=0.1:start_threshold=-50dB,areverse,silenceremove=start_periods=1:start_silence=0.1:start_threshold=-50dB,areverse "New\%%a"

[videoAI]

Command Breakdown

for %%a in (*.mp3) do ffmpeg -i "%%a" -af silenceremove=start_periods=1:start_silence=0.1:start_threshold=-50dB,areverse,silenceremove=start_periods=1:start_silence=0.1:start_threshold=-50dB,areverse "New\%%a"

Key Components

1. Batch Loop
- `for %%a in (*.mp3)`: Iterates through all MP3 files in the current directory
- `do`: Executes the following command for each file

2. FFmpeg Audio Filters
- `-af`: Applies audio filters in sequence
- `silenceremove`: Removes silent parts from the audio

3. Silence Removal Process
- First `silenceremove`:
- `start_periods=1`: Remove silence at the beginning of the file
- `start_silence=0.1`: Minimum silence duration of 0.1 seconds
- `start_threshold=-50dB`: Silence threshold at -50 decibels

- `areverse`: Reverses the audio temporarily

- Second `silenceremove`: Removes silence from the end (now the beginning after reversal)

- Final `areverse`: Restores the audio to its original direction

4. Output
- `"New\%%a"`: Saves processed files in a "New" subdirectory with original filename

- The script will process each MP3
- Remove silence from the start and end
- Save processed files in a "New" subfolder

[_Al_]

thanks, but that is a filter I was using, it is not lossless

[_Al_]

It looks right filter is silencedetect, that finds time for a silence to start and end (perhaps no end, if it is end of a song).

ffmpeg -i "%~1" -af silencedetect=noise=-50dB:d=0.5,ametadata=print:file=log.txt -f null -

It prints a log, that log can be parsed for last start and end segment. If there is no end, last entry is a start of silence, so that is a que perhaps we have a silent end for trimming.
I will make something in python to batch auto it.