Recherche avancée

Sur d’autres sites (13183)

• ffmpeg split_by_silence output flac have broken metadata [duplicate]

  20 mars 2023, par Martin

  Input flac File : https://file.io/D1EXEgJ0OtgM

  


  I have a shell script split_by_silence.sh which takes an input audio file, detects split points where silence (or very low volume audio) occurs, and then splits the audio file using ffmpeg into the individual split point segments.

  


  So if 10 split points = 11 audio files exported.

  


  When I run it on an mp3, it produces 11 files (as expected with my input file) and each file has correct length metadata.

  


  But when ffmpeg splits a flac file, the metadata output for 'length' is broken

  


  I've tried following the closest question, marked as a duplicate, here :
Cutting FLAC using ffmpeg does not change timestamps accordingly

  


  By adding a step to re-encode my flac file in ffmpeg :

  


  ffmpeg -i full.flac -ss 0 -t 2241 fixed.flac

  


  Which I'm not sure if is working, but outputs a fixed.flac file. Which when I run with my split command :

  


  ffmpeg -i "inputFile.flac" -c copy -map 0 -f segment -segment_times "22,441,5567,1122,etc..." "%d_output.flac"

  


  The output files still have broken flac 'length' metadata, like you can see in my image, even though when opened in audacity, the file does have a real length. ffmpeg just encoded the wrong value ?

  


  

  


  This is my current ffmpeg version :

  


  $ ffmpeg
ffmpeg version 4.4.2-0ubuntu0.22.04.1 Copyright (c) 2000-2021 the FFmpeg developers
  built with gcc 11 (Ubuntu 11.2.0-19ubuntu1)


  


  And here is my .sh file I run with

  


  $ ./split_by_silence.sh

  


  # -----------------------
# SPLIT BY SILENCE
# Requirements:
#    ffmpeg
#    $ apt-get install bc
# How To Run:
# $ ./split_by_silence.sh "full_lowq.flac" %03d_output.flac

# output title format
OUTPUTTITLE="%d_output.flac"
# input audio filepath
IN="/mnt/c/Users/marti/Documents/projects/split_by_silence2/full.flac"
# output audio filepath
OUTPUTFILEPATH="/mnt/c/Users/marti/Documents/projects/split_by_silence2"
# ffmpeg option: split input audio based on this silencedetect value
SD_PARAMS="-11dB"
MIN_FRAGMENT_DURATION=120 # split option: minimum fragment duration
export MIN_FRAGMENT_DURATION

# -----------------------
# step: ffmpeg
# goal: get comma separated list of split points (use ffmpeg to determine points where audio is at SD_PARAMS [-18db] )

echo "_______________________"
echo "Determining split points..."
SPLITS=$(
    ffmpeg -v warning -i "$IN" -af silencedetect="$SD_PARAMS",ametadata=mode=print:file=-:key=lavfi.silence_start -vn -sn  -f s16le  -y /dev/null \
    | grep lavfi.silence_start= \
    | cut -f 2-2 -d= \
    | perl -ne '
        our $prev;
        INIT { $prev = 0.0; }
        chomp;
        if (($_ - $prev) >= $ENV{MIN_FRAGMENT_DURATION}) {
            print "$_,";
            $prev = $_;
        }
    ' \
    | sed 's!,$!!'
)
echo "split points list= $SPLITS"

# add '5.5' to each split\
IFS=',' read -ra SPLITS_ARRAY <<< "$SPLITS"
for i in "${!SPLITS_ARRAY[@]}"; do
  SPLITS_ARRAY[i]=$(echo "${SPLITS_ARRAY[i]}+5.5" | bc)
done

SPLITS=$(IFS=','; echo "${SPLITS_ARRAY[*]}")
echo "SPLITS=$SPLITS"

# using the split points list, calculate how many output audio files will be created 
num=0
res="${SPLITS//[^,]}"
CHARCOUNT="${#res}"
num=$((CHARCOUNT + 2))
echo "_______________________"
echo "Exporting $num tracks with ffmpeg"

ffmpeg -i "$IN" -c copy -map 0 -f segment -segment_times "$SPLITS" "$OUTPUTFILEPATH/$OUTPUTTITLE"

echo "Done."

echo "------------------------------------------------"
echo "$num TRACKS EXPORTED"


  


  I think this is an open defect ?
https://trac.ffmpeg.org/ticket/4905

  


• How to format the output of ffmpeg

  1er mars 2023, par Default

  I am trying to use ffmpeg to get the list of audio input devices for my machine(windows 11).

  


  This is the command that I use to achieve this :

  


  ffmpeg -list_devices true -f dshow -i dummy -hide_banner


  


  This is the output :

  


  [dshow @ 00000282627590c0] "OBS Virtual Camera" (video)
[dshow @ 00000282627590c0]   Alternative name "@device_sw_{860BB310-5D01-11D0-BD3B- 
00A0C911CE86}\{A3FCE0F5-3493-419F-958A-ABA1250EC20B}"
[dshow @ 00000282627590c0] "Microphone (Realtek(R) Audio)" (audio)
[dshow @ 00000282627590c0]   Alternative name "@device_cm_{33D9A762-90C8-11D0-BD43- 
00A0C911CE86}\wave_{2409F2C8-F2CB-4A6D-AEA7-DAD79C4EE9DF}"


  


  I want to modify the output so it only shows this :

  


  "OBS Virtual Camera" (video)
"Microphone (Realtek(R) Audio)" (audio)


  


  How can I remove the brackets at the beginning as well as the alternative names to achieve this ? What command do I use ?

  


• FFMPEG, convert video same as input size with watermark in Batch

  11 août 2023, par Deex

  I would like to watermark some videos in a directory. I use a png watermark of 3840x2160 in size. The watermark should go in fullsize on the video. Because my videos have different resolutions I use this batch script

  


  set /p add_watermark=Wasserzeichen Addieren? (Standard: ja/nein):
if /i "%add_watermark%" == "" set add_watermark=ja

if /i "%add_watermark%" == "ja" (
    for %%f in (%cd%\output\tmp\*.mp4) do (
        set "filename=%%~nf"
        ffmpeg -i "%%f" -i %cd%\watermark\watermark.png -filter_complex "[0:v][1:v]scale2ref=oh*mdar:ih[v0][v1];[v0][v1]overlay=main_w-overlay_w-0:0" -c:a copy -map 0 -y "%cd%\output\%%~nf.mp4"
    )
)


  


  That works well with videos in 4k but if my input is a HD video it will scale the output up to 4k in the size of the watermark.

  


  If I try to add
-c:v copy FFMPEG is giving out the error

  


  Streamcopy requested for output stream fed from a complex filtergraph. Filtering and streamcopy cannot be used together.


  


  My question is, how can I set the output as same size as the input video with the scaled watermark without doing this in two steps.