Recherche avancée

Médias (91)

Autres articles (77)

  • Websites made ​​with MediaSPIP

    2 mai 2011, par

    This page lists some websites based on MediaSPIP.

  • Creating farms of unique websites

    13 avril 2011, par

    MediaSPIP platforms can be installed as a farm, with a single "core" hosted on a dedicated server and used by multiple websites.
    This allows (among other things) : implementation costs to be shared between several different projects / individuals rapid deployment of multiple unique sites creation of groups of like-minded sites, making it possible to browse media in a more controlled and selective environment than the major "open" (...)

  • Personnaliser les catégories

    21 juin 2013, par

    Formulaire de création d’une catégorie
    Pour ceux qui connaissent bien SPIP, une catégorie peut être assimilée à une rubrique.
    Dans le cas d’un document de type catégorie, les champs proposés par défaut sont : Texte
    On peut modifier ce formulaire dans la partie :
    Administration > Configuration des masques de formulaire.
    Dans le cas d’un document de type média, les champs non affichés par défaut sont : Descriptif rapide
    Par ailleurs, c’est dans cette partie configuration qu’on peut indiquer le (...)

Sur d’autres sites (12352)

  • lavd/x11grab : fix vertical repositioning

    28 mars 2019, par Octavio Alvarez
    lavd/x11grab : fix vertical repositioning
    

    There is a calculation error in xcbgrab_reposition() that breaks
    vertical repositioning on follow_mouse. It made the bottom
    reposition occur when moving the mouse lower than N pixels after
    the capture bottom edge, instead of before.

    This commit fixes the calculation to match the documentation.

    follow_mouse : centered or number of pixels. The documentation says :

    When it is specified with "centered", the grabbing region follows
    the mouse pointer and keeps the pointer at the center of region ;
    otherwise, the region follows only when the mouse pointer reaches
    within PIXELS (greater than zero) to the edge of region.

    • [DH] libavdevice/xcbgrab.c
  • ffmpeg extract segment from video on-the-fly

    29 novembre 2022, par brunoais

    Context

    


    I want to make a service that hosts mp4 files but also provides video streaming.

    


    The server side is made of 2 "small" edge servers with capacity to cache about 0.1% of the content and one main server with a fraction of the bandwidth of the edge servers but much more robust storage.

    


    Recent status

    


    With the help of the ffmpeg manual,
    
Recently, when a segment is requested by an edge server, I run this command in ffmpeg

    


    ffmpeg -i in.mp4 -f hls -hls_list_size 0 -hls_playlist 1 -hls_time 60 -strftime 1  -hls_flags independent_segments+second_level_segment_index+second_level_segment_duration -hls_base_url '/path/to/file/' -hls_segment_filename 'name_%%03d.ts'  -c copy -copyts -hls_segment_type fmp4 out.m3u8


    


    Then obtain the requested segment (plus 3 subsequent ones) then delete everything.

    


    Current status

    


    Currently, I tried to get some level of optimization using -to :

    


    ffmpeg -i in.mp4 -to  -f hls -hls_list_size 0 -hls_playlist 1 -hls_time 60 -strftime 1  -hls_flags independent_segments+second_level_segment_index+second_level_segment_duration -hls_base_url '/path/to/file/' -hls_segment_filename 'name_%%03d.ts'  -c copy -copyts -hls_segment_type fmp4 out.m3u8


    


    Then sending the requested segments to the edge that requested them, then delete the result.
Do note that /path/to/file/ is a tmpfs with quite modest capacity (files can't stay there for too long).

    


    The current setback

    


    The main issue I get with the current process is that it takes a long time to obtain the last segments (2-4s).
That creates a bottleneck in how long segments take to be served. I can increase the buffer from 3 videos to 5 (or even more) but that doesn't solve the actual problem and, instead, will bring more strain on other areas.

    


    Cutting is not reliable

    


    There doesn't seem to exist an option to select what segments to generate using ffmpeg.

    


    Using cutting argument for start time (-ss) has shown to usually work but regularely causes the first keyframe used to be wrong. However, I believe I have all segment files in variable length and cut at the keyframe of the original file.

    


    Help needed

    


    How to extract an arbitrary segment of an mp4, described in the m3u8 file (which was done in a previous extraction) ?

    


    I have full control over ffmpeg and I can automate edits to the m3u8 file after generating them, as required. However, I need to understand and see how can this be solved.

    


  • Embedding pure python(Cpython) on android

    13 août 2020, par KiYugadgeter

    I want to create movie download app for android for learn.
To make develop easily, I would like to use youtube-dl for downloader backend.

    


    So I want to embed Cpython runtime and ffmpeg (for convert movie format) to Android app.
Is it able to do with android NDK ?

    


    Note that I know more better ways are exist.(like use java-friend python runtime or downloader implement as online server)

    


    But I want to try to embed python and ffmpeg in app for learn.

    


    Can it with Android NDK ?