lavc: use a separate field for exporting audio encoder padding

Currently, the amount of padding inserted at the beginning by some audio
encoders, is exported through AVCodecContext.delay. However
- the term 'delay' is heavily overloaded and can have multiple different
  meanings even in the case of audio encoding.
- this field has entirely different meanings, depending on whether the
  codec context is used for encoding or decoding (and has yet another
  different meaning for video), preventing generic handling of the codec
  context.

Therefore, add a new field -- AVCodecContext.initial_padding. It could
conceivably be used for decoding as well at a later point.

1 files changed, 18 insertions, 10 deletions

diff --git a/libavcodec/avcodec.h b/libavcodec/avcodec.h
index f0fa7a959f..a24ce407c9 100644
--- a/libavcodec/avcodec.h
+++ b/libavcodec/avcodec.h
@@ -1191,16 +1191,7 @@ typedef struct AVCodecContext {
      *   encoded input.
      *
      * Audio:
-     *   For encoding, this is the number of "priming" samples added by the
-     *   encoder to the beginning of the stream. The decoded output will be
-     *   delayed by this many samples relative to the input to the encoder (or
-     *   more, if the decoder adds its own padding).
-     *   The timestamps on the output packets are adjusted by the encoder so
-     *   that they always refer to the first sample of the data actually
-     *   contained in the packet, including any added padding.
-     *   E.g. if the timebase is 1/samplerate and the timestamp of the first
-     *   input sample is 0, the timestamp of the first output packet will be
-     *   -delay.
+     *   For encoding, this field is unused (see initial_padding).
      *
      *   For decoding, this is the number of samples the decoder needs to
      *   output before the decoder's output is valid. When seeking, you should
@@ -2780,6 +2771,23 @@ typedef struct AVCodecContext {
      * use AVOptions to set this field.
      */
     int side_data_only_packets;
+
+    /**
+     * Audio only. The number of "priming" samples (padding) inserted by the
+     * encoder at the beginning of the audio. I.e. this number of leading
+     * decoded samples must be discarded by the caller to get the original audio
+     * without leading padding.
+     *
+     * - decoding: unused
+     * - encoding: Set by libavcodec. The timestamps on the output packets are
+     *             adjusted by the encoder so that they always refer to the
+     *             first sample of the data actually contained in the packet,
+     *             including any added padding.  E.g. if the timebase is
+     *             1/samplerate and the timestamp of the first input sample is
+     *             0, the timestamp of the first output packet will be
+     *             -initial_padding.
+     */
+    int initial_padding;
 } AVCodecContext;
 
 /**