Recherche avancée

Sur d’autres sites (7192)

• Evolution #3000 (Nouveau) : box-sizing et mediabox

  13 mai 2013, par Valéry -

  Je signale un problème que j’ai rencontré à deux reprises. Certains intégrateurs ou frameworks css (knacss par ex.) choisissent dans le reset box-sizing:border-box ; or ceci casse l’affichage du plugin mediabox.

  On peut naturellement dans le css surcharger les styles de mediabox mais ne serais-ce pas plutôt aux plugins SPIP de préciser systématiquement le modèle de boîte qu’ils utilisent pour que les utilisateurs n’aient pas à rechercher la solution (sachant que seuls les utilisateurs avancés vont la trouver).

• Convert MP3 bitrate from a stream into another stream with ffmpeg

  10 juillet 2013, par Léon Pelletier

  Using ffmpeg, I would like to know if it's possible to convert mp3 bitrate as data chunks are received ?

  That means I would send slowly chunks to ffmpeg so that it outputs a mp3 with another bitrate.

  So in very-pseudo-code, it looks like it :

  1. MP3 Request from user

  2. Send the default mp3 to ffmpeg with parameters to convert to the desired bitrate.

  3. As it's writing a new file, write what as been writen so far in the Response outputstream (I'm in ASP.Net)

  Is that feasable or I need to switch to another technology ?

  [EDIT]

  For now, I'm trying a solution like this : Convert wma stream to mp3 stream with C# and ffmpeg

  [EDIT 2]

  I answered my question, and it is feasible with an url as input and standard output as output. Using an url allows to process a file chunk by chunk, and using stdout, we can access data while it is processed.

• How to use ffmpeg / x264 2-Pass encoding for multiple bitrate output files

  2 décembre 2014, par Jonesy

  While performing a 2-Pass encode to multiple output files I was receiving the error

  ratecontrol_init: can't open stats file 1 ffmpeg2pass-2.log

  My setup is to do a single first pass and then multiple second pass encodes to output files with different target bitrates using the same first pass results.

  ffmpeg -y -i $INPUT_FILE -an -vcodec libx264 -pass 1 -b:v 700k -f rawvideo /dev/null
  
  
  
  ffmpeg -y -i $INPUT_FILE -i out-aud.mp4 \
  
  $AUDIO_OPTIONS_P2 -vcodec libx264 -pass 2 -b:v 250k -f mp4 out-250.mp4 \
  
  $AUDIO_OPTIONS_P2 -vcodec libx264 -pass 2 -b:v 500k -f mp4 out-500.mp4 \
  
  $AUDIO_OPTIONS_P2 -vcodec libx264 -pass 2 -b:v 700k -f mp4 out-700.mp4

  This sequence resulted in the error listed above. What I discovered thru code-inspection is that ffmpeg/x264 looks for a different set of first-pass files for each second-pass encoding path. The first encoding path uses the set of files originally created

  ffmpeg2pass-0.log
  
  ffmpeg2pass-0.log.mbtree

  The second encoding path requires first-pass files with the names

  ffmpeg2pass-2.log
  
  ffmpeg2pass-2.log.mbtree

  The third encoding path requires first-pass files with the names starting with ffmpeg2pass-4*, etc.

  My solution was to create soft-links to the originally created set of files with the new names that were required for each pass before running the second-pass command.

  ln -s ffmpeg2pass-0.log ffmpeg2pass-2.log
  
  ln -s ffmpeg2pass-0.log.mbtree ffmpeg2pass-2.log.mbtree
  
  ln -s ffmpeg2pass-0.log ffmpeg2pass-4.log
  
  ln -s ffmpeg2pass-0.log.mbtree ffmpeg2pass-4.log.mbtree

  This seems to work as it results in the output encodes that I needed. However, I don’t know if this method is legitimate. Am I getting sub-optimal encoding results by using a first-pass output for one bitrate (700k) as the input to second-pass encodings for other bitrates ?