Recherche avancée

Sur d’autres sites (10505)

• How to tell from a media file if it's a video or an audio ? [duplicate]

  25 février 2021, par Random Name

  I have some files - audio and video files - stored in an aws s3 bucket, which can have different containers and stuff, and I would like to send them to the aws mediaConvert(MC). My issue is that I need to know whenever the file is an audio or a video file, because I need to configure the MC accordingly. The main app on the server is php based, but this information might not be that interesting.

  


  I have ffmpeg, and I already use ffprobe to get the file duration, however I can't tell from the returning data if it's a video or audio-only file. And in an another question someone said that maybe I can get this info by using the ffmpeg, but as far as I know that requires me to download the file from the s3 to the server.
There is also an s3 function that I can use to get the header of a specific object, which contains the mime type of the file, however this sometimes returns as audio/mpeg for video files, which then configures the MC incorrectly.

  


  I could also check the error message from the first job create of MC, which then tells me about the issue, and I can resend the file with the correct configuration, but then the job is marked as failed with error, which doesn't feel right to do.

  


  So my question is ; what would be the best approach to tell from a media file if it's a video or audio only ? It's possible that ffmpeg or ffprobe can do it remotely, but I just don't know the tools deeply enough, so I'm not entirely sure that it can only work in that way I described it.

  


• ffmpeg seek by position, not timestep [closed]

  23 octobre 2023, par user112346

  I'm implementing an application which requires extracting, say, 100 frames from a video stream, processing them, then extracting the next 100 and repeating this process until the end of the video.

  


  I've been doing this using :

  


  ffmpeg -i input.mkv -vf "select=gt(n\,$MYMIN_MINUS_ONE)*lt(n\,$MYMAX_PLUS_ONE)" -vsync 0 frames/frame%08d.png


  


  However, this approach encodes everything unnecessarily and is slow. I'm not sure if ffmpeg can seek based on the sequential frame position. Therefore, I changed the code to use timesteps by first doing :

  


  ffmpeg -itsscale $FRAME_RATE -i input.mkv -vf "fps=1" -q:v 1 -vsync vfr -c:v libx265 input_stretched.mkv


  


  Which stretches the input such that each frame in the original sequence corresponds to a different second. Then frame access is done by using :

  


  ffmpeg -ss $MYMIN -i input_stretched.mkv -frames:v $NFRAMES frames/frame%08d.png


  


  This solution seems to work and, after the initial encoding, access is much faster than using the filtergraph. I can probably alternatively divide the timestamps when accessing, instead of stretching the video, although I have not yet tested this.

  


  I'd like to know if there is a less "hacky" way of doing this, such as to use the frame sequence numbers directly, instead of converting to timesteps.

  


• FFMPEG : new m3u8 playlist in every hour from the same segmented stream

  23 mars 2018, par iPog

  I’m using the latest Windows build of FFMPEG (by Zeranoe).

  Currently I create 2 outputs from the same raw live video input. I use the tee pseudo-muxer to give the user the following outputs :

  • PRIMARY OUTPUT : transcoded live stream ; using the HLS or the segment muxer ; 14 pieces of 10-second-long segments on a ramdisk.
  • SECONDARY OUTPUT : from the same transcoding I store every segment on the hard drive ; to make the hard drive versions watchable, I create m3u8 playlists with an automated batch script in every hour ; all individual playlist files contain 1 hour of content.

  Is it possible to achieve the same result with FFMPEG only ? I.e. the secondary output should be able to finish the current m3u8 playlist, and start a new one with a new filename at every hour o’clock.

  (My batch-based solution works fine, so it isn’t that important, but it would be nice to know if it is possible at all. I could not find a similar approach in the documentation/wiki.)