Recherche avancée

Sur d’autres sites (5037)

• Having ffmpeg add a repeating text overlay on a video

  30 mai 2020, par Caius Jard

  I'm looking to create an overlay that cycles through the characters in a string over and over. I've succeeded in using a sendcmd file to put A, B, C, D, E on the first 5 seconds of a video

  



  0  drawtext reinit 'text=A';
1  drawtext reinit 'text=B';
2  drawtext reinit 'text=C';
3  drawtext reinit 'text=D';
4  drawtext reinit 'text=E';


  



  But it doesn't cycle and I haven't been able to find a way to make it, because sendcmd looks like it just takes a simple timecode. I could make a command file 3600 lines long for my hour video, with those commands in over and over (the command file would be generated programmatically so not onerous)

  



  ---

  



  After some considerable experimenting I was able to do it with 5 separate drawtext :

  



  drawtext=fontfile=/Windows/Fonts/bauhs93.ttf:fontsize=1024:fontcolor=white@0.1:bordercolor=black@0.1:borderw=10:r=250:text='A':x=if(trunc(mod(t\,5))\,-2000\,(w-tw)/2),
drawtext=fontfile=/Windows/Fonts/bauhs93.ttf:fontsize=1024:fontcolor=white@0.1:bordercolor=black@0.1:borderw=10:r=250:text='B':x=if(trunc(mod(t\,5))-1\,-2000\,(w-tw)/2),
drawtext=fontfile=/Windows/Fonts/bauhs93.ttf:fontsize=1024:fontcolor=white@0.1:bordercolor=black@0.1:borderw=10:r=250:text='C':x=if(trunc(mod(t\,5))-2\,-2000\,(w-tw)/2),
drawtext=fontfile=/Windows/Fonts/bauhs93.ttf:fontsize=1024:fontcolor=white@0.1:bordercolor=black@0.1:borderw=10:r=250:text='D':x=if(trunc(mod(t\,5))-3\,-2000\,(w-tw)/2),
drawtext=fontfile=/Windows/Fonts/bauhs93.ttf:fontsize=1024:fontcolor=white@0.1:bordercolor=black@0.1:borderw=10:r=250:text='E':x=if(trunc(mod(t\,5))-4\,-2000\,(w-tw)/2)


  



  But as can be seen, I have to repeat a lot of stuff here. Is there any slicker way ? It does seem to have a noticeable effect on encoding speed the more chars are added

  



  I was hoping that text expressions would help but it seems I can only return numerics from the values, so this expression didn't work out :

  



  %{e:if(trunc(mod(t,5)),'A', '')%{e:if(trunc(mod(t,5))-1,'B', '') ...


  


• FFMPEG concat videos with filter_comlex, getting memory allocate Error

  8 août 2019, par Cherry_Hunter

  I m trying to concat few videos and output 1 combined video.
  I have following codes, and it runs :

  ffmpeg.exe -i 1.mov -i
  
  2.mp4 -f lavfi -i color=s=2000x2000 -f lavfi -t 1 -i anullsrc -filter_complex "[0:v]scale=2000x2000:force_original_aspect_ratio=decrease[v0];[v0]pad=2000:2000:(W-w)/2:y=(H-h)/2[v0];[1:v]scale=2000x2000:force_original_aspect_ratio=[v1];[v1]pad=2000:2000:(W-w)/2:y=(H-h)/2[v1];[v0][3:a][v1][1:a] concat=n=2:v=1:a=1[v][a]" -map "[v]" -map "[a]"  out.mp4

  However, it shows following errors half way :

  Error while filtering: Cannot allocate memory44kB time=00:03:55.63 bitrate=1326.1kbits/s speed=0.493x
  
  Failed to inject frame into filter network: Cannot allocate memory
  
  Error while processing the decoded data for stream #1:0

  Target :

  combine 1.mov and 2.mp4 into one mp4 video, output resolution is 2000x2000 without changing input videos’ aspect ratios.

  2.mp4 has a audio stream while 1.mov has only video stream, I m trying to add a dummy audio to 1.mov and keep audio stream from 2.mp4.

  It keeps getting memory allocate errors, both input videos are small and it takes up 16GB RAM.

  Help !!!

  My temporary solution is to split the process into smaller parts :

  1. add dummy audio to 1.mov and modify resolution -> output1.mp4
  2. modify resolution of 2.mp4 -> output2.mp4
  3. concat output1.mp4 and output2.mp4

  But it takes much more longer time, it is possible to improve the first code so that avoid memory errors !

• Poweramp Equalizer : how they intercept and process the audio coming from another app ?

  25 mai 2021, par Fabio Fracassi

  I'm spending a lot of time trying to understand how Poweramp Equalizer works. Seems that they have a service that intercepts the audio using the audio session ID and after, they process the audio using the FFmpeg.
My question is : how is possible to intercept and substitute the audio streamed by another app knowing the audio session ID ?

  


  Legenda : They -> Maxim Petrov :-)