Recherche avancée

Sur d’autres sites (11654)

• ffplay startup time proportional to specified framerate

  17 décembre 2019, par bremen_matt

  I am playing a video over http using ffplay. The call I am using looks like this :

  ffplay -framerate 30 -fflags nobuffer -flags low_delay -autoexit -i http://localhost:8880

  The video is an H.264 encoding where (my understanding is a bit unclear here) it is something like a "raw" H.264 stream without timestamps.

  My primary concern is to get video displayed with low latency. In that regard, the video is fine.

  The issue is with the framerate and with the startup time.

  The video source is emitting frames as soon as they are processed, so the frame rate is not constant. However, my experience is that as long as you specify a framerate larger than the max achievable frame rate of the source, then the viewer still looks fine. On the flipside, if the video source starts emitting frames at 60 fps, but I specify a framerate of 30, then the delay just sort of builds up in ffplay to the point where after 10 seconds, the video is 20 seconds behind. So the first question would be whether there is a way to get ffplay to use a variable framerate. The behavior I am looking for is "display a frame as soon as it is received over http".

  In light of the aforementioned issue, the approach I have been taking is to simply specify a high framerate, which seems to work. However, there is an issue with this approach in the form of startup time. When I set the framerate to 10, the ffplay window starts in approx 3 seconds, but then quickly starts accruing a lag (so I can’t do this). When I set the framerate to 100, the ffplay window takes 30 seconds (literally 30 seconds) to start, but then will not have any lag.

  I have seen that ffmpeg has a vsync option that on the surface seems like it would allow you to set a variable framerate. However, ffplay doesn’t seem to recognize this. I would also be willing to pipe the output of ffmpeg to a different window (I am running Ubuntu 18.04) if that is what must be done, but I would prefer not to have to recompile ffplay.

• How to accurately detect the start of the main beat and soundtracks in diverse audio tracks ?

  18 juin 2024, par SnoofFloof

  I'm working on a project where I need to edit soundtracks. The challenge is to detect when the main beat and melody of any given soundtrack is properly developed. I am certain there is better terminology to describe what I am aiming for, but ideally, I want to skip the "build-up" and immediately have the song starting at the "main part". This needs to work for various songs across different genres, which often have different structures and onset patterns, making it difficult to streamline the process.

  


  For example :

  


  https://www.youtube.com/watch?v=P77CNtHrnmI -> I would want to my code to identify the onset at 0:24

  


  https://www.youtube.com/watch?v=OOsPCR8SyRo -> Onset detection at 0:12

  


  https://www.youtube.com/watch?v=XKiZBlelIzc -> Onset detection at 0:19

  


  I've tried using librosa to analyze the onset strength and detect beats, but the current implementation either detects the very beginning of the song or fails to consistently identify when the beat is fully developed.

  


  This was my approach ;

  


  def analyze_and_edit_audio(input_file, output_file):
    y, sr = librosa.load(input_file)
    tempo, beat_frames = librosa.beat.beat_track(y=y, sr=sr)
    beat_times = librosa.frames_to_time(beat_frames, sr=sr)
    main_beat_start = beat_times[0]


  


  I have very little experience with librosa/audio editing, so I would appreciate any suggestions you might have !

  


• Stuck in deadend while compiling ffmpeg with nvidia CUDA support on Ubuntu 18.04 and GeForce GT 730

  24 février 2019, par Marko

  I want to compile ffmpeg with CUDA support, on Ubuntu 18.04, and my graphics card is GeForce GT 730 (so far I have not CUDA programming experience).
  Here nvidia claims that it has at least some transcoding support.

  I have cloned https://git.ffmpeg.org/ffmpeg.git and as I understand I also need to build and install nv-codec-headers.

  But there in the README, they say that minimum driver version for Linux is 418.30 or newer. So I went to nvidia site to download driver and after selecting my card I got this download link. So the driver version is high enough it seems. But when I try to install it I get message that this device will be ignored by that driver and that it is supported by legacy driver 390.xx (which is version I can install with apt).
  Is something wrong with nVidia, why are they offering me driver that is cannot be used for this device ?

  I’ve installed 390.xx driver using ubuntu-drivers autoinstall and built ffmpeg, but when I try to run it I get this error : Cannot load cuvidGetDecodeStatus [h264_cuvid @ 0x562a86dbc680] Failed loading nvcuvid.

  I suppose it is because of driver ?

  So, what can I do to unstuck myself ?