Recherche avancée

Sur d’autres sites (5170)

• Real-Time Buffer Too Full (FFMPEG)

  25 janvier 2018, par Nimble

  So I’ve been having this issue with ffmpeg, it has been a journey getting the hardware and command to actually do what I want, but I still have one problem.

  Sometimes when I’m recording I just start dropping frames like crazy, this can be after an hour of recording or even ten hours in... Everything will be working fine and then suddenly I’ll start dropping frames due to "real-time buffer too full or near too full". This happens regardless of how low I put the bitrate, and the buffer size is high as it will allow, eventually I’ll just start dropping frames. Almost seems like it could be like a power saving feature kicking in but it’s too inconsistent it seems. Like I said sometimes I can go 10 hours without having this issue.

  Any ideas ?

  Here is my block of code :

  ffmpeg -guess_layout_max 0 -y -f dshow -video_size 3440x1440 -rtbufsize 2147.48M -pixel_format nv12 -framerate 200 ^
  
  -i video="Video (00 Pro Capture HDMI 4K+)":audio="SPDIF/ADAT (1+2) (RME Fireface UC)" -map 0:0,0:1 -map 0:1 ^
  
  -preset: llhp -codec:v h264_nvenc -pix_fmt nv12 -b:v 250M -maxrate:v 250M -minrate:v 250M -bufsize:v 250M -b:a 320k ^
  
  -ac 2 -r 100 -async 1 -vsync 1 -segment_time 600 -segment_wrap 9 -f segment C:\Users\djcim\Videos\PC\PC%02d.mp4 ^
  
  -guess_layout_max 0 -f dshow -rtbufsize 2000M -i audio="Analog (3+4) (RME Fireface UC)" -map 1:0 -b:a 320k -ac 2 ^
  
  -af "adelay=200|200" -segment_time 600 -segment_wrap 9 -f segment C:\Users\djcim\Videos\PC\Voices\Theirs\TPC%02d.wav ^
  
  -guess_layout_max 0 -f dshow -rtbufsize 2000M -i audio="Analog (5+6) (RME Fireface UC)" -map 2:0 -b:a 320k -ac 2 ^
  
  -af "adelay=825|825" -segment_time 600 -segment_wrap 9 -f segment C:\Users\djcim\Videos\PC\Voices\Mine\MPC%02d.wav

  Here is the error, it repeated around 300 times before locking up ffmpeg forcing my to quit before starting the recording again :

  [dshow @ 0000019a596bdcc0] real-time buffer [Video (00 Pro Capture HDMI 4K+)] [video input] too full or near too full (62% of size: 2147480000 [rtbufsize parameter])! frame dropped!

• Input/output error while using ffmpeg with pulse

  2 septembre 2017, par user116522

  i want to record my audio output. Yesterday, i used the following command

  ffmpeg -f pulse -ac 2 -i 1 -c:a libfdk_aac -b:a 192k test4.m4a

  and it was working. But now, i receive the output :

   ffmpeg version N-83415-g012dd8c Copyright (c) 2000-2017 the FFmpeg developers
  
    built with gcc 4.8 (SUSE Linux)
  
    configuration: --enable-gpl --enable-libx264 --enable-libx265 --enable-nonfree --enable-libfdk-aac --enable-libpulse
  
    libavutil      55. 46.100 / 55. 46.100
  
    libavcodec     57. 75.100 / 57. 75.100
  
    libavformat    57. 66.101 / 57. 66.101
  
    libavdevice    57.  2.100 / 57.  2.100
  
    libavfilter     6. 73.100 /  6. 73.100
  
    libswscale      4.  3.101 /  4.  3.101
  
    libswresample   2.  4.100 /  2.  4.100
  
    libpostproc    54.  2.100 / 54.  2.100
  
  1: Input/output error

  The command

  pactl list sources

  says :

  Source #0
  
          State: SUSPENDED
  
          Name: alsa_output.pci-0000_01_00.1.hdmi-stereo.monitor
  
          Description: Monitor of HDA NVidia Digital Stereo (HDMI)
  
          Driver: module-alsa-card.c
  
          Sample Specification: s16le 2ch 44100Hz
  
          Channel Map: front-left,front-right
  
          Owner Module: 6
  
          Mute: no
  
          Volume: front-left: 65536 / 100% / 0,00 dB,   front-right: 65536 / 100% / 0,00 dB
  
                  balance 0,00
  
          Base Volume: 65536 / 100% / 0,00 dB
  
          Monitor of Sink: alsa_output.pci-0000_01_00.1.hdmi-stereo
  
          Latency: 0 usec, configured 0 usec
  
          Flags: DECIBEL_VOLUME LATENCY 
  
          Properties:
  
                  device.description = "Monitor of HDA NVidia Digital Stereo (HDMI)"
  
                  device.class = "monitor"
  
                  alsa.card = "1"
  
                  alsa.card_name = "HDA NVidia"
  
                  alsa.long_card_name = "HDA NVidia at 0xf7080000 irq 17"
  
                  alsa.driver_name = "snd_hda_intel"
  
                  device.bus_path = "pci-0000:01:00.1"
  
                  sysfs.path = "/devices/pci0000:00/0000:00:01.0/0000:01:00.1/sound/card1"
  
                  device.bus = "pci"
  
                  device.vendor.id = "10de"
  
                  device.vendor.name = "NVIDIA Corporation"
  
                  device.product.id = "0fb9"
  
                  device.string = "1"
  
                  module-udev-detect.discovered = "1"
  
                  device.icon_name = "audio-card-pci"
  
          Formats:
  
                  pcm

• ffmpeg 180 degree panoramic fisheye image to equirectangular / flat

  7 juillet 2024, par Willy62

  I am trying to get my Hikvision Panovu image of a sportsfield to look like a standard camera image, similar to what would be seen with a Veo solution / traditional camera.

  


  This is what the image would ideally look like with a little bit of zoom. Note the players are all upright and it looks "correct" and not skewed with the far end of the field in line with the horizon.

  


  veo sample image

  


  The original image looks like this (same field but other side). This is a 180 degree panoramic image from a Hikvision camera as found here.

  


  It provides the following output natively.

  


  hikvision native view

  


  I have had some luck converting the image with ffmpeg using the v360 filter. Note there is a downward tilt meaning I have to apply some yaw to correct it.

  


  v360=input=fisheye:output=rectilinear:ih_fov=180:iv_fov=87.5:d_fov=87.5:pitch=20:yaw=5:w=3840:h=2160


  


  And this gets the following output :

  


  deskewed v360 image

  


  So the challenge here to make the original image flat/equirect but to address the skew such that :

  


  

  • the players are orientated "upright"

  


  • the far sideline of the field looks like a straight line in line with the horizon

  


  • the image quality is preserved as best as possible

  


  


  With these cameras the image is 32MP so there is the opportunity to do an ePTZ into the area of interest.

  


  I suspect v360 isnt the right choice here and it is some remap-style filter, or perhaps I am best going across to gstreamer or similar.

  


  I tried an ffmpeg v360 filter and it partially works, but the players are still skewed because the top of the image is not wide enough. The issue can possibly be solved by correctly applying a couplex perspective filter, but I think this will only mask the issue and perspective requires a complex filter that hasn't worked for me so far.