From: Anton Khirnov <anton@khirnov.net>
To: ffmpeg-devel@ffmpeg.org
Subject: [FFmpeg-devel] [PATCH 18/30] lavc/encode: pass through frame durations to encoded packets
Date: Sun, 27 Nov 2022 18:03:39 +0100
Message-ID: <20221127170351.11477-18-anton@khirnov.net> (raw)
In-Reply-To: <20221127170351.11477-1-anton@khirnov.net>
The generic code can only handle the no-delay case. Encoders with delay
need to be handled individually, which will be done in the following
commits.
---
doc/APIchanges | 2 +-
libavcodec/avcodec.h | 7 +++++++
libavcodec/encode.c | 14 ++++++++++++--
libavcodec/options_table.h | 1 +
4 files changed, 21 insertions(+), 3 deletions(-)
diff --git a/doc/APIchanges b/doc/APIchanges
index a3c9819e32..4ae5571f71 100644
--- a/doc/APIchanges
+++ b/doc/APIchanges
@@ -15,7 +15,7 @@ libavutil: 2021-04-27
API changes, most recent first:
2022-12-xx - xxxxxxxxxx - lavc 59.55.100 - avcodec.h
- Add AV_CODEC_FLAG_COPY_OPAQUE.
+ Add AV_CODEC_FLAG_COPY_OPAQUE and AV_CODEC_FLAG_FRAME_DURATION.
2022-11-xx - xxxxxxxxxx - lavu 57.43.100 - tx.h
Add AV_TX_FLOAT_DCT, AV_TX_DOUBLE_DCT and AV_TX_INT32_DCT.
diff --git a/libavcodec/avcodec.h b/libavcodec/avcodec.h
index 79558786ee..a31d413179 100644
--- a/libavcodec/avcodec.h
+++ b/libavcodec/avcodec.h
@@ -249,6 +249,13 @@ typedef struct RcOverride{
* @ref AV_CODEC_CAP_ENCODER_REORDERED_OPAQUE capability flag.
*/
#define AV_CODEC_FLAG_COPY_OPAQUE (1 << 7)
+/**
+ * Signal to the encoder that the values of AVFrame.duration are valid and
+ * should be used (typically for transferring them to output packets).
+ *
+ * If this flag is not set, frame durations are ignored.
+ */
+#define AV_CODEC_FLAG_FRAME_DURATION (1 << 8)
/**
* Use internal 2pass ratecontrol in first pass mode.
*/
diff --git a/libavcodec/encode.c b/libavcodec/encode.c
index e4270b6c34..00b7a7dde8 100644
--- a/libavcodec/encode.c
+++ b/libavcodec/encode.c
@@ -215,10 +215,13 @@ int ff_encode_encode_cb(AVCodecContext *avctx, AVPacket *avpkt,
if (avpkt->pts == AV_NOPTS_VALUE)
avpkt->pts = frame->pts;
- if (avctx->codec->type == AVMEDIA_TYPE_AUDIO) {
- if (!avpkt->duration)
+ if (!avpkt->duration) {
+ if (frame->duration)
+ avpkt->duration = frame->duration;
+ else if (avctx->codec->type == AVMEDIA_TYPE_AUDIO) {
avpkt->duration = ff_samples_to_time_base(avctx,
frame->nb_samples);
+ }
}
avctx->reordered_opaque = frame->reordered_opaque;
@@ -454,6 +457,13 @@ FF_ENABLE_DEPRECATION_WARNINGS
return ret;
}
+ // unset frame duration unless AV_CODEC_FLAG_FRAME_DURATION is set,
+ // since otherwise we cannot be sure that whatever value it has is in the
+ // right timebase, so we would produce an incorrect value, which is worse
+ // than none at all
+ if (!(avctx->flags & AV_CODEC_FLAG_FRAME_DURATION))
+ dst->duration = 0;
+
return 0;
}
diff --git a/libavcodec/options_table.h b/libavcodec/options_table.h
index ad41dc7bed..011314db96 100644
--- a/libavcodec/options_table.h
+++ b/libavcodec/options_table.h
@@ -59,6 +59,7 @@ static const AVOption avcodec_options[] = {
{"qscale", "use fixed qscale", 0, AV_OPT_TYPE_CONST, {.i64 = AV_CODEC_FLAG_QSCALE }, INT_MIN, INT_MAX, 0, "flags"},
{"recon_frame", "export reconstructed frames", 0, AV_OPT_TYPE_CONST, {.i64 = AV_CODEC_FLAG_RECON_FRAME}, .unit = "flags"},
{"copy_opaque", "propagate opaque values", 0, AV_OPT_TYPE_CONST, {.i64 = AV_CODEC_FLAG_COPY_OPAQUE}, .unit = "flags"},
+{"frame_duration", "use frame durations", 0, AV_OPT_TYPE_CONST, {.i64 = AV_CODEC_FLAG_FRAME_DURATION}, .unit = "flags"},
{"pass1", "use internal 2-pass ratecontrol in first pass mode", 0, AV_OPT_TYPE_CONST, {.i64 = AV_CODEC_FLAG_PASS1 }, INT_MIN, INT_MAX, 0, "flags"},
{"pass2", "use internal 2-pass ratecontrol in second pass mode", 0, AV_OPT_TYPE_CONST, {.i64 = AV_CODEC_FLAG_PASS2 }, INT_MIN, INT_MAX, 0, "flags"},
{"gray", "only decode/encode grayscale", 0, AV_OPT_TYPE_CONST, {.i64 = AV_CODEC_FLAG_GRAY }, INT_MIN, INT_MAX, V|E|D, "flags"},
--
2.35.1
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".
next prev parent reply other threads:[~2022-11-27 17:09 UTC|newest]
Thread overview: 37+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-11-27 17:03 [FFmpeg-devel] [PATCH 01/30] lavc/libx264: factor out setting up the input frame Anton Khirnov
2022-11-27 17:03 ` [FFmpeg-devel] [PATCH 02/30] lavc/libx264: reindent after previous commit Anton Khirnov
2022-11-27 17:03 ` [FFmpeg-devel] [PATCH 03/30] lavc/libx264: use a local variable for input frame in setup_frame() Anton Khirnov
2022-11-27 17:03 ` [FFmpeg-devel] [PATCH 04/30] lavc/libx264: factor out setting up ROI Anton Khirnov
2022-11-27 17:03 ` [FFmpeg-devel] [PATCH 05/30] lavc/libx264: reindent after previous commit Anton Khirnov
2022-11-27 17:03 ` [FFmpeg-devel] [PATCH 06/30] lavc/libx264: unify cleanup in setup_frame() Anton Khirnov
2022-11-27 17:03 ` [FFmpeg-devel] [PATCH 07/30] lavc/libx264: do not ignore memory allocation errors Anton Khirnov
2022-11-27 17:03 ` [FFmpeg-devel] [PATCH 08/30] lavc/libx264: reindent after previous commit Anton Khirnov
2022-11-27 17:03 ` [FFmpeg-devel] [PATCH 09/30] lavc/libx264: reorder control flow in setup_roi() to reduce nesting depth Anton Khirnov
2022-11-27 17:03 ` [FFmpeg-devel] [PATCH 10/30] lavc/libx264: reindent after previous commit Anton Khirnov
2022-11-27 17:03 ` [FFmpeg-devel] [PATCH 11/30] lavc/libx264: use a local variable to shorten code Anton Khirnov
2022-11-27 17:03 ` [FFmpeg-devel] [PATCH 12/30] lavc/libx264: print an error on invalid opaque pointer Anton Khirnov
2022-11-27 17:03 ` [FFmpeg-devel] [PATCH 13/30] lavc/libx264: zero reordered opaque on alloc Anton Khirnov
2022-11-27 17:03 ` [FFmpeg-devel] [PATCH 14/30] lavc/libx264: do not leave an invalid array size on alloc error Anton Khirnov
2022-11-27 17:03 ` [FFmpeg-devel] [PATCH 15/30] lavc/libx265: restructure handling reordered_opaque Anton Khirnov
2022-11-27 17:03 ` [FFmpeg-devel] [PATCH 16/30] lavc: add a codec flag for propagating opaque from frames to packets Anton Khirnov
2022-11-27 17:03 ` [FFmpeg-devel] [PATCH 17/30] lavc: support AV_CODEC_CAP_ENCODER_REORDERED_OPAQUE in all no-delay encoders Anton Khirnov
2022-11-27 17:43 ` Michael Niedermayer
2022-11-27 19:43 ` Anton Khirnov
2022-11-27 20:00 ` James Almer
2022-11-27 17:03 ` Anton Khirnov [this message]
2022-11-27 17:03 ` [FFmpeg-devel] [PATCH 19/30] lavc/librav1e: handle frame durations and AV_CODEC_CAP_ENCODER_REORDERED_OPAQUE Anton Khirnov
2022-11-27 17:03 ` [FFmpeg-devel] [PATCH 20/30] lavc/nvenc: " Anton Khirnov
2022-11-27 17:03 ` [FFmpeg-devel] [PATCH 21/30] lavc/adxenc: rescale packet duration according to timebase Anton Khirnov
2022-11-27 17:03 ` [FFmpeg-devel] [PATCH 22/30] lavc/adxenc: support AV_CODEC_CAP_ENCODER_REORDERED_OPAQUE Anton Khirnov
2023-01-04 16:26 ` Andreas Rheinhardt
2022-11-27 17:03 ` [FFmpeg-devel] [PATCH 23/30] lavc/ffv1enc: handle frame durations and AV_CODEC_CAP_ENCODER_REORDERED_OPAQUE Anton Khirnov
2023-01-04 17:27 ` Andreas Rheinhardt
2022-11-27 17:03 ` [FFmpeg-devel] [PATCH 24/30] lavc/pngenc: " Anton Khirnov
2022-11-27 17:03 ` [FFmpeg-devel] [PATCH 25/30] lavc/pngenc: stop setting dts unnecessarily for APNG Anton Khirnov
2022-11-27 17:03 ` [FFmpeg-devel] [PATCH 26/30] lavc/libtheoraenc: handle frame durations and AV_CODEC_CAP_ENCODER_REORDERED_OPAQUE Anton Khirnov
2022-11-27 17:03 ` [FFmpeg-devel] [PATCH 27/30] lavc/libtheoraenc: stop setting dts unnecessarily Anton Khirnov
2022-11-27 17:03 ` [FFmpeg-devel] [PATCH 28/30] lavc/libx264: pass through frame durations to encoded packets Anton Khirnov
2022-11-27 17:03 ` [FFmpeg-devel] [PATCH 29/30] lavc/libx265: " Anton Khirnov
2022-11-27 17:03 ` [FFmpeg-devel] [PATCH 30/30] lavc/libaomenc: " Anton Khirnov
2022-11-27 20:19 ` James Almer
2023-01-04 16:15 ` Anton Khirnov
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20221127170351.11477-18-anton@khirnov.net \
--to=anton@khirnov.net \
--cc=ffmpeg-devel@ffmpeg.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Git Inbox Mirror of the ffmpeg-devel mailing list - see https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
This inbox may be cloned and mirrored by anyone:
git clone --mirror https://master.gitmailbox.com/ffmpegdev/0 ffmpegdev/git/0.git
# If you have public-inbox 1.1+ installed, you may
# initialize and index your mirror using the following commands:
public-inbox-init -V2 ffmpegdev ffmpegdev/ https://master.gitmailbox.com/ffmpegdev \
ffmpegdev@gitmailbox.com
public-inbox-index ffmpegdev
Example config snippet for mirrors.
AGPL code for this site: git clone https://public-inbox.org/public-inbox.git