Split by Silence, ffmpeg
24. 10. 2022Split by silence (very fragile) ↓
silence detection v1
input="80s.m4a"
quiet=(-hide_banner -loglevel info -nostats)
ffmpeg "${quiet[@]}" -i "${input}" -af silencedetect=n=-50dB:d=1 -f null - 2> | awk '{ print $4, $5}' | grep "silence_" > detect.log
Returns
silence_start: 447.672
silence_end: 449.415
silence_start: 674.288
silence_end: 677.722
silence_start: 677.899
...
Now, my 1st song start is probably 1st silence_end (-ss), having the spliting command in mind
ffmpeg "${quiet}" -i "${input}" -ss 449.415 -to 674.288 -c:a copy test.m4a
Rename to -ss and -to
sed -i 's/silence_end:/-ss/g' detect.log
sed -i 's/silence_start:/-to/g' detect.log
silence detection v2, real one
ffmpeg "${quiet[@]}" -i "${input}" -af silencedetect=n=-50dB:d=0.5 -f null - 2>&1 | \
awk '{ print $4, $5}' | grep "silence_" | sed 's/silence_end:/-ss/g' | sed 's/silence_start:/-to/g' \
| tail -n +2 > detect.log
Returns
-ss 449.415
-to 674.288
-ss 677.722
-to 677.899
...
Tail shall skip 1st line.
Add empty line
echo >> detect.log
Note: If 1st is to be ignored, that means that the file is required to start with some silence
Test loop
awk_round () {
awk 'BEGIN{printf "%."'$1'"f\n", "'$2'"}'
}
c="1"
cat detect.log | while read _ ssvar; read _ tovar; do
#pad name
printf -v name "%03d" "$c"
# action
ssvar="$(awk_round "5" "$ssvar")"
tovar="$(awk_round "5" "$tovar")"
echo "-ss $ssvar -to $tovar $name.m4a"
(( c++ ))
done
The split loop, real one
awk_round () {
awk 'BEGIN{printf "%."'$1'"f\n", "'$2'"}'
}
round () {
ssvar="$(awk_round "5" "$ssvar")"
tovar="$(awk_round "5" "$tovar")"
}
c="1"
quiet=(-hide_banner -loglevel info -nostats -nostdin)
cat detect.log | while read _ ssvar; read _ tovar; do
#pad name
printf -v name "%03d" "$c"
# action
echo; echo "-ss $ssvar -to $tovar"
echo "--------------------------------------"
if [ -z "$tovar" ]; then
round && ffmpeg "${quiet[@]}" -i "${input}" -ss "$ssvar" -c:a copy "${name}".m4a
else
round && ffmpeg "${quiet[@]}" -i "${input}" -ss "$ssvar" -to "$tovar" -c:a copy "${name}".m4a
fi
(( c++ ))
done
Split by yt ‘chapters’ file ↓
Example ‘list.txt’
0:00 As It Was - Harry Styles
2:39 Stay - The Kid LAROI with Justin Bieber
4:57 Bad Habits - Ed Sheeran
8:42 Unstoppable - Sia
12:17 Memories - Maroon 5
Add some empty lines
echo "\n\n" >> list.txt
Test loop (dual) - should return the split commands
loop () {
list="$1"
tail -n +${c} "$list" | while read -r ssvar name ; read -r tovar _ ; do
# action
[ -z "${ssvar// }" ] && break
#pad counter
printf -v ccc "%03d" "$c"
if [ -z "${tovar// }" ]; then
#echo "empty tovar"
echo "-ss $ssvar ${c}_${name}.m4a"
else
echo "-ss $ssvar -to $tovar ${ccc}-${name}.m4a"
fi
(( c = c + 2 ))
done
}
c="1"; loop list.txt; c="2"; loop list.txt
Actual spliter script
#!/bin/bash
# spliter
# split local 'm4a audio file' by local 'list.txt' (yt chapters thing)
# example list.txt
# 0:00 As It Was - Harry Styles
# 2:39 Stay - The Kid LAROI with Justin Bieber
# 4:57 Bad Habits - Ed Sheeran
# 8:42 Unstoppable - Sia
# 12:17 Memories - Maroon 5
# 15:22 Dance Monkey - Tones and I
# 18:49 Girls Like You - Maroon 5
# usage: spliter master.m4a list.txt
input="$1"; list="$2"
# Add some empty lines
echo "\n\n" >> "$list"
loop () {
tail -n +${c} "$list" | while read -r ssvar name ; read -r tovar _ ; do
# action
[ -z "${ssvar// }" ] && break
#pad counter
printf -v ccc "%03d" "$c"
if [ -z "${tovar// }" ]; then
#echo "empty tovar"
echo "-ss $ssvar ${c}_${name}.m4a"
ffmpeg "${quiet[@]}" -i "${input}" -ss "$ssvar" -c:a copy "${ccc}_${name}".m4a
else
echo "-ss $ssvar -to $tovar ${ccc}-${name}.m4a"
ffmpeg "${quiet[@]}" -i "${input}" -ss "$ssvar" -to "$tovar" -c:a copy "${ccc}_${name}".m4a
fi
(( c = c + 2 ))
done
}
quiet=(-hide_banner -loglevel quiet -nostats -nostdin)
c="1"; loop; c="2"; loop
