Home / ubuntu / Questions / 1342540
Accepted
Jags
Asked: 2021-06-04 12:45:40 +0800 CST

Bash script to run FFmpeg concat on MP4 files in batches of 5

  • 772

I like to create a Bash script to concatenate MP4 files using FFmpeg, in batches of 5 files at a time for a directory with 100 MP4 files, so that afterwards there would be 20 files like:

001_005.mp4, 006_010.mp4, and so on...

instead of just 1 file consisting of all 100 files.

Contents of mylist.txt:

file 001.mp4
file 002.mp4
file 003.mp4
file 004.mp4
file 005.mp4
............
file 099.mp4
file 100.mp4

Though I've found a command that works just fine (from this StackOverflow thread), it would create only 1 file consisting all 100 files.

#!/bin/bash

cd /home/admn/Downloads/MP4_Files;

# Create mylist.txt:
for f in *.mp4
do
    echo "file $f" >> mylist.txt
done;

# Concatenate files:
ffmpeg -f concat -safe 0 -i mylist.txt -c copy output.mp4;

So, how do I modify the ffmpeg command so that it concat in batches of 5 files at a time.

All the files have exact same resolution (1080p), audio, and video codecs.

OS: Ubuntu MATE 21.04

ffmpeg version: 4.3.2-0+deb11u1ubuntu1

command-line

3 Answers

  1. Best Answer

    I think the following script will work.

    • First try it as it is and check that it seems to do what you want
    • Then remove echo from the line with ffmpeg to make it do its thing.

    Check that the content of the temporary files xaa' ... xat` matches the names (and content) of the output files.

    #!/bin/bash

> mylist.txt
for f in *.mp4
do
 echo "file '$f'" >> mylist.txt
done

< mylist.txt sort -t \' -n -k2 | split -l 5

k=1
for j in x*
do
 inc=$(wc -l "$j" | cut -d ' ' -f 1)
 m=$(printf "%03d" $((k)))
 n=$(printf "%03d" $((k+inc-1)))
 name="${m}_${n}.mp4"
 echo ffmpeg -f concat -safe 0 -i "$j" -c copy "$name"
 k=$((k+5))
done
    • 3

  2. Another solution is to use nested loops.

    As long as there are files remaining, the inner loop uses ${@:1:5} to take the next slice of (up to) 5 files.

    #!/bin/bash
cd /home/admn/Downloads/MP4_Files;

shopt -s failglob
set -- *.mp4
while [[ $# -gt 0 ]]; do
    from=$(basename "$1" .mp4)
    for f in "${@:1:5}"; do
        #not essential in your case, but use @Q to quote/escape special characters
        echo "file ${f@Q}" >> mylist.txt
        shift
    done
    to=$(basename "$f" .mp4)
    ffmpeg -f concat -safe 0 -i mylist.txt -c copy "${from}_${to}.mp4"
    rm mylist.txt
done
    • 1

  3. Assuming mylist.txt exists in your current working directory, you can do something like this,

    LINES=$(cat mylist.txt | wc -l)
BATCH_SIZE=5
BATCHES=$(($LINES/$BATCH_SIZE))             

for ((i=0;i<$BATCHES;i++))
do
  cat mylist.txt | head -n $((($i+1)*$BATCH_SIZE)) | tail -n $BATCH_SIZE > slist.txt
  ffmpeg -f concat -safe 0 -i slist.txt -c copy $(($i*$BATCH_SIZE+1))_$((($i+1)*$BATCH_SIZE)).mp4
done
rm slist.txt
    • 0