Recherche avancée

Sur d’autres sites (8660)

• Convert .flac to .mp3 with ffmpeg, avoid small silence ?

  19 septembre 2017, par RocketNuts

  I’m converting a bunch of .flac audio samples to .mp3 with ffmpeg. Some of these samples are supposed to loop, i.e. the end of the sample fits seamlessly to the beginning (no tick).

  Now no matter how high the mp3 quality setting I use, I keep getting about 0.05 sec of silence at the beginning of the resulting .mp3 which is not there in the original flac.

  Here’s a zoomed in Audacity view of the beginning and ending of both audio samples :

  Original flac :

  

  Notice how the beginning and ending are both ’flat’, so they fit smoothly.

  Converted to mp3 :

  

  Notice how there is now suddenly a 0.05 sec silence, and the end doesn’t end exactly flat (on the baseline). Hence they don’t fit, and the tick and silence are clearly audible when playing in a loop.

  I’m using this for the conversion :

  ffmpeg -i foo.flac -acodec libmp3lame -ab 320k bar.mp3

  or this : (with the -q:a VBR setting)

  ffmpeg -i foo.flac -acodec libmp3lame -q:a 0 bar.mp3

  Either -q:a 0 or -ab 320k is supposed to give the highest possible quality, but I can’t get rid of that strange silence in the beginning and small cutoff at the end, which causes a tick when playing in a loop.

• Google Cloud Platform Storage JSON API upload breaks audio files..?! How to fix ?

  13 octobre 2017, par Zolai

  I did manage to upload files to Google storage with GCP resumable upload (json api) :
  https://cloud.google.com/storage/docs/json_api/v1/how-tos/resumable-upload

  Using jQuery and Plupload http://www.plupload.com

  But now it seems that audio files loses their codecs in upload and those won’t play or ffmpeg can’t probe those. More testing showed that actually .flac files are ok to ffmpeg, but for example .m4a files are not.

  I need to get most of the audio files uploaded to google storage and be able to probe those with ffmpeg. And must be able upload big files over 100Mb. That’s why I believe JSON API is best or only possible way.

  I think I have metadata also added correctly, but now I’m confused how to continue..
  Any ideas what could be the problem here ?

  EDIT (more info) :

  I had working software :

  • Using regular POST > /temp > Google Cloud Storage Client (php).
  • GCP Storage Client uploaded files with perfection.
  • Even that I named uploaded files with random unique name like "woeiwehf.tmp", those did work and ffmpeg could probe them fine.

  THEN something happened, I guess uploading like this was limited (or something) :

  • What meant that my solution couldn’t upload anymore big files, over 32Mb.

  • Well that wasn’t best solution anyway, so that’s okey, that this changed.

  NOW :

  • I have changed software so it uses GCP JSON API upload resumable, and this mainly works
  • But I had to change files to look like "woeiwehf.flac" or what ever type is. ".tmp" didn’t work anymore.
  • For example .flac files work nicely, but .m4a will not.
  • I’ve tested uploading many different ways, and it seems that JSON API upload loses at least codecs and bit rate.

  —

  I would be very grateful if I could get help how to resolve this.
  I guess one solution may could be to fork ffmpeg client and manually set audio codec, but I’m saving that for the last solution.

• How the ffmpeg astats crest factor is calculated

  30 août 2017, par FranGar

  I’m scripting a ffmpeg chain process for my work. The aim is normalizing/compressing lot of audio files (mp3’s).
  It’s done in Python and the critical part is the line :

  ffmpeg -y -i "Input.mp3" -codec:a libmp3lame -b:a 96k -af acompressor=threshold=-15dB:ratio=5:attack=0.01:release=1000:knee=2,dynaudnorm=g=3:m=2:p=0.95 "Output.mp3"

  The python script it’s complete and working BUT the nature of the audios (voice recordings) are very different so I can’t use the same params for all of them.

  I make some experimenting with the values of the ffmpeg filter astats and i discovered that the crest factor (Standard ratio of peak to RMS level ) gave a good reference to programatically get the better params.

  In fact I saw that a recording with a nice dynamic range sound and smooth in shape, get crest values around 9-15 (the compress/normlz params will be somehow conservative). But audios with crest around 22-30 need more aggressive processing.
  (All empirically)

  Somebody can clarify how the crest values are really calculated ? Which are the peaks taken to account ? (Why the flat factor is always 0 ?)
  Or if somebody knows how to get a value representing the sound ’smoothness’ will be nice also.

  Thanks for the ideas.