Recherche avancée

Sur d’autres sites (9850)

• ffmpeg audio ducking with stereo and surround input

  6 février 2021, par cyptus

  I got two input audio files for ffmpeg :

  


  

  1. 5.1 channels with background music (background.mp3)

  


  2. Stereo with voice (voice.mp3)

  


  


  I want to merge this files into one ouput file (5.1), while reducing the background music (file 1) when voice (file 2) is played (called audio ducking). sidechaincompress should be the correct filter for this.

  


  ffmpeg -i background.mp3 -i voice.mp3 -filter_complex "\
[1:a]asplit=2[voice][voice2];\
[0:a][voice]sidechaincompress[mix];\
[voice2][mix]amerge" \
mixed.mp3


  


  Is kind of able to produce this output, but is down-mixing the background.mp3 to a stereo output (ffprobe results in channels=2 channel_layout=stereo for the generated mixed.mp3).

  


  How can i merge the streams voice2 and mix while keeping all channels ?

  


• Revision 1f4bf79d65 : Added per pixel inter rd hit count stats Added some code to output normalized r

  21 août 2013, par Paul Wilkins

  Changed Paths :
   Modify /vp9/encoder/vp9_onyx_if.c

  
   Modify /vp9/encoder/vp9_onyx_int.h

  
   Modify /vp9/encoder/vp9_rdopt.c

  
  
  Added per pixel inter rd hit count stats

  Added some code to output normalized rd hit count stats.
  In effect this approximates to the average number of rd
  operations/tests per pixel for the sequence.

  The results are not quite accurate and I have not bothered
  to account for partial SB64s at frame edges and for key frames
  However they do give some idea of the number of modes /
  prediction methods being tested for each pixel across the
  different partition sizes. This indicates how much scope their
  is for further gains either by reducing the number of partitions
  examined or the modes per partition through heuristics.

  Patch 3 moved place where count incremented so partial rd
  tests that are aborted with INT_MAX return are also counted.

  Example numbers for first 50 frames of Akiyo.
  Speed 0 84.4 rd operations / pixel
  Speed 1 28.8
  Speed 2 11.9

  Change-Id : Ib956e787e12f7fa8b12d3a1a2f6cda19a65a6cb8

• Improving Performance and Quality of Screen Streaming(Headless) with FFMPEG and NVIDIA Hardware Acceleration [closed]

  24 avril 2024, par Dhairya Verma

  I am attempting to stream my screen to an RTMP URL using FFMPEG with X11 for screen capture and NVIDIA's hardware acceleration to enhance performance. Despite using NVIDIA acceleration, the stream is still experiencing lags and low-quality output. I've noticed that FFMPEG is utilizing only about 100MB of GPU memory, which seems low. Here's the command I'm currently using :

  


  ffmpeg -hwaccel cuvid -f x11grab -s 1920x1080 -i :1 -f pulse -i VirtualSink.monitor -c:v h264_nvenc -preset:v p1 -b:v 2500k -maxrate 2500k -bufsize 5000k -vf "fps=30,crop=1280:720:320:180,format=yuv420p" -g 60 -c:a aac -b:a 128k -ar 44100 -f flv rtmps://[RTMP_URL]


  


  Questions :

  


  


  1. Are there any specific settings or tweaks I should consider to fully utilize the GPU for better performance and video quality ?

     


  


  2. Is there a more optimal way to configure the bitrate or buffer size to improve the stream quality without increasing lag ?

     


  


  3. Would adjusting the preset or using different ffmpeg flags help in reducing the load and improving the output ?

     


  


  4. Any advice on optimizing FFMPEG for smoother streaming with hardware acceleration would be greatly appreciated !