Recherche avancée

Sur d’autres sites (7785)

• Evolution #4103 : Autoriser /local/cache-gd2/ et /local/cache-vignette/ dans robots.txt

  28 mai 2018, par b b

  Hop,

  Dans mon cas, à partir du moment où une image est visible sur une page, je considère qu’elle doit être indexée car c’est du contenu et un point d’entrée au même titre que le texte (recherche par image).

  C’est plus fin que ça, autoriser l’indexation de local/ ferait que toutes les tailles de vignettes d’une images seraient indexées par les moteurs alors que l’image originale l’est déjà (ce qui suffit amplement). De plus, cela pourrait générer pas mal de trafic en plus sur les sites uniquement pour indexer des images "non utiles aux visiteurs", bref ça ne serait pas super écolo :p

• Split audio file into multiple files using ffmpeg on macOS

  3 avril 2018, par Tony Hegr

  My idea was to save some time and split long audio file into multiple audio files and reading the data (startTime, endTime, artist, title) from a text file.
  Using terminal on MacOS (UNIX) and ffmpeg (audio).

  Input put file example with 3 lines :

  00:00 - 06:16 - art - tit
  
  06:16 - 09:22 - arti - titl
  
  09:22 - 13:13 - artist - title

  My unix code, shell file : "split.command"

  inputFileName=audiofile
  
  input=test.txt
  
  
  
  
  
  while IFS= read -r var
  
  do
  
  
  
  startTime=$(echo "$var" | cut -f1 -d"-")
  
  endTime=$(echo "$var" | cut -f2 -d"-")
  
  artist=$(echo "$var" | cut -f3 -d"-")
  
  title=$(echo "$var" | cut -f4 -d"-")
  
  
  
  echo ffmpeg -i ~/Downloads/$inputFileName.mp3 -metadata artist="$artist" -metadata title="$title" -ss $startTime -to $endTime -acodec copy "$title".mp3 -vsync 2
  
  
  
  wait 
  
  
  
  done < "$input"

  When I ECHO it, everything looks fine.

  $ sh split.command
  
  ffmpeg -i /Users/tony/Downloads/audiofile.mp3 -metadata artist= art  -metadata title= tit -ss 00:00 -to 06:16 -acodec copy  tit.mp3 -vsync 2
  
  ffmpeg -i /Users/tony/Downloads/audiofile.mp3 -metadata artist= arti  -metadata title= titl -ss 06:16 -to 09:22 -acodec copy  titl.mp3 -vsync 2
  
  ffmpeg -i /Users/tony/Downloads/audiofile.mp3 -metadata artist= artist  -metadata title= title -ss 09:22 -to 13:13 -acodec copy  title.mp3 -vsync 2

  When ECHO is deleted, it should work as command. The first audio is fine, the second throws an error, and the third work but has missing artwork .

  1st *works*
  
  2nd *error* Invalid duration specification for to: arti
  
  3rd *works but..* [mp3 @ 0x7ff7e1800000] No packets were sent for some of the attached pictures.

• NodeJS/ffmpeg : Can't create screenshot of stream file

  11 septembre 2017, par user3142695

  I need to create a screenshot via ffmpeg (node-fluent-ffmpeg : https://github.com/fluent-ffmpeg/node-fluent-ffmpeg) of a stream, as my data is stored in a gridFS.

  This is how I’m doing that with a local file - and it works so far

  ffmpeg('/Users/Anybody/Downloads/1test.mp4')
  
    .on('error', (error) => {
  
      console.error(error)
  
    })
  
    .on('end', () => {
  
      console.log('Screenshots taken')
  
    })
  
    .screenshots({
  
      folder: '/Users/Anybody/Downloads/',
  
      timestamps: ['00:01.000']
  
    })

  But if I change the code to use a stream (gridfs-stream), no file is created, although the output gives me a successful message :

  import Grid from 'gridfs-stream'
  
  
  
  const gfs = Grid(
  
    MongoInternals.defaultRemoteCollectionDriver().mongo.db,
  
    MongoInternals.NpmModule
  
  )
  
  
  
  ffmpeg(gfs.createReadStream({ _id: sourceId }))
  
    .on('filenames', (filenames) => {
  
      console.log('Will generate ' + filenames.join(', '))
  
    })
  
    .on('codecData', (data) => {
  
      console.log('Input is ' + data.audio + ' audio ' + 'with ' + data.video + ' video')
  
    })
  
    .on('error', (error) => {
  
      console.error(error)
  
    })
  
    .on('end', () => {
  
      console.log('Screenshots taken')
  
    })
  
    .screenshots({
  
      folder: '/Users/Anybody/Downloads/',
  
      timestamps: ['00:01.000']
  
    })

  Output

  Will generate tn.png
  
  Input is aac (mp4a / 0x6134706D) audio with h264 (avc1 / 0x31637661) video
  
  Screenshots taken