Recherche avancée

Sur d’autres sites (3673)

• [FFmpeg on Windows] : help me converting a set of .png files to a video file

  31 octobre 2016, par Stefano Fedele

  Hi everybody I am on windows 10, I installed FFMpeg and I would like to use it to convert 100 .png image files into a video file.

  I am using it via the prompt, which is the Windows console, and I am struggling with the syntax (which probability comes from Linux). I set the folder in which ffmpeg.exe is contained as a windows system path as suggested on many websites and I moved the prompt to the folders in which the image files are contained which is reported here :

  cd C:\video\pnts + ag 40 nm- during-tmpyp 10-12- laser power 7mw- kinetic100- exposure time 1s- gain

  I would like to convert the image files contained in this folder into a video file, the names of those image files are reported here :

  pnts + ag 40 nm- during-tmpyp 10-12- laser power 7mw- kinetic100- exposure time 1s- gain 20 001.png

  pnts + ag 40 nm- during-tmpyp 10-12- laser power 7mw- kinetic100- exposure time 1s- gain 20 002.png

  ...

  ...

  pnts + ag 40 nm- during-tmpyp 10-12- laser power 7mw- kinetic100- exposure time 1s- gain 20 100.png

  The number of frames per second whould be 2.

  I read on internet that I should have to type something like this

  ffmpeg -i image-%03d.png video.webm

  But it looks like there are a lot of variables to consider and all the times I try to set them and run FFmpeg I get an error from the prompt.
  Is there anyone who could suggest me the proper syntax, please ?

• Using FFMPEG To Fill in Gaps of a Raw Audio UDP Stream

  14 février 2020, par Wallace

  I have a software defined radio (SDR) that picks up audio from emergency services and with the help of software, streams raw audio using UDP. The audio is PCM signed 16-bit little-endian. The UDP stream is also not constant and only has data when audio is detected.

  The problem I’m trying to solve is that I would like the gaps in recorded audio to be filled with silent or Null audio. Below are just a couple of my attempts at resolving this :

  1. ffmpeg -f s16le -ar 8000 -i udp ://127.0.0.1:23456 -af aresample=async=1 -acodec libmp3lame -
     f rtp rtp ://127.0.0.1:1234
  2. ffmpeg -re -f lavfi -i anullsrc -f s16le -ar 8000 -i udp ://127.0.0.1:23456 -filter_complex
     amix=inputs=2:duration=first -acodec libmp3lame -f rtp rtp ://127.0.0.1:1234

  I guess my questions are regarding the best way to resolve this and whether or not ffmpeg can be used in a way to accomplish this ?

• ffmpeg : Make output file duration that of the file with the longest length [closed]

  27 février 2024, par losercantcode

  I'm going crazy trying to figure this out.

  


  Essentially I'm trying to apply sidechain compression using ffmpeg (for the sake of speed and automation). a2.mp3 in this example is the full length track and a1.wav is a short 5 second, spoken-word clip. The code currently applies the sidechain compression well, but the output file is only that of the spoken-word clips length, so 5 seconds. Is it at all possible to make the output match that of a2.mp3's length instead ? Switching around the input order is not something that's possible as it'll apply the compression to the incorrect track.

  


  Essentially, i'm trying to make it so i'm just layering the spoken-word file on top of the audio track without having to pre-pad the file. I'd rather, if it's possible, this just happen with one block of code within the in one program. I'm interested to see if this is at all possible within ffmpeg.

  


  ffmpeg -i a2.mp3 -i a1.wav -filter_complex "[1:a]asplit=2[sc][mix];[0:a][sc]sidechaincompress=threshold=0.000976563:ratio=10:level_sc=0.015625:release=100:attack=10[compr];[compr][mix]amix=duration=longest[aout]" -map "[aout]" output.wav


  


  a1.wav is mon, a2.mp3 is stereo.

  


  Attempted to implement amix=duration=longest to no avail. Research similar problems on the Stack Exchange that gave unrelated answers.