GitLab

+ put access and demux in the right place

  (what the hell is i_duplicate ?)

* Each module can declare a "human-readable short name" with set_name
* Modules are sorted by category (set_category, set_subcategory).
  Modules configs can be separated by set_section()
* Separated audio-filter and audio-visual
* Separated extraintf and control
* New command and widget : add_module_list() for comma-separated modules
* Vfilters now use "," as separator

* modules/stream_out/transcode.c: trying to make the audio/fps resync work even when an ES is not transcoded.

  original author) email address.

* modules/stream_out/transcode.c: fixed subpictures overlaying when the audio-sync or fps options are used.

* modules/stream_out/transcode.c: new sfilter option for using subpictures filters during the transcoding.

* src/libvlc.h: new --sub-filter config option + added a new subpictures category that regroups all the subpictures options.

* modules/stream_out/description.c: bug fix.

* modules/stream_out/description.c: new "description" stream output module which gathers information on elementary streams (pre-parsing).
* src/misc/vlm.c: pre-parse vod media items.

* modules/stream_out/transcode.c: re-use the SPU engine.
* src/video_output/vout_subpictures.c: support for resizing subpictures.

* modules/stream_out/transcode.c: use picture ring buffers to avoid leaks due to decoders not releasing pictures properly and to avoid uncessary mallocs.

* modules/stream_out/transcode.c: when overlaying subtitles, duplicate frames if they are referenced by the decoder.

* stream_out/transcode.c: deinterlacing works again.

* modules/stream_out/transcode.c: respect input requested by encoder.

  + correct cropping of the decoded video when specified.
* stream_out/transcode.c: make sure the size of the encoded video is at least a multiple of 2.

  - Re-use our audio decoder modules instead of using libavcodec directly.
  - No more dependance on libavcodec.
  (downmixing is currently broken + you have to force the ffmpeg codec for a52)
* modules/audio_filter/format.c:
  - PCM audio format conversion filter using the new common filter architecture.
* modules/audio_filter/converter/mpgatofixed32.c:
  - implements both the old and new filter architecture (ie. useable in the transcoder).
* modules/codec/ffmpeg/audio.c:
  - fixes and cleanup.

  - chroma + resizing video filter (using the filter_t architecture).
* modules/codec/ffmpeg/*:
  - cleanup + small updates.
* modules/codec/speex.c, theora.c, vorbis.c:
  - got rid of pf_header() in the encoder.
  - store the headers in fmt_out.p_extra (this will break the ogg muxer for now).
* modules/codec/libmpeg2.c, modules/codec/ffmpeg/video.c:
  - added a p_dec->b_pace_control field to signal if the decoder is allowed to drop frames.
* modules/stream_out/transcode.c:
  - heavy cleanup.
  - re-use video decoder modules and got rid of the duplicated ffmpeg video decoder.
  - use video filters for chroma conversion and resizing.
  (a few things are broken now like deinterlacing but I'll repair them asap).

    + Added macros for AltiVec versions of these modules. Will not work at all
      for the moment.

* modules/stream_out/transcode.c: small cleanup + added an fps (to specify the output frame rate) and an audio-sync option (to add/drop video frames to keep a constant framerate while being synchronised on the audio track).