1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229
|
From: ihy123 <aladinandreyy@gmail.com>
Date: Thu, 14 Aug 2025 12:44:10 +0300
Subject: ip/ffmpeg: more precise seeking
av_seek_frame() and avformat_seek_file() seek to nearest "keyframe". For
codecs like, for example, ape this means that seeking will be very off
(5 seconds or more). So what we do is:
1. seek to nearest "keyframe" before the desired time,
2. discard some frames to approach the desired time.
---
ip/ffmpeg.c | 154 +++++++++++++++++++++++++++++++++++++-----------------------
1 file changed, 94 insertions(+), 60 deletions(-)
diff --git a/ip/ffmpeg.c b/ip/ffmpeg.c
index 21b9a01..ecbf005 100644
--- a/ip/ffmpeg.c
+++ b/ip/ffmpeg.c
@@ -44,6 +44,8 @@ struct ffmpeg_input {
AVPacket pkt;
int curr_pkt_size;
uint8_t *curr_pkt_buf;
+ int64_t seek_ts;
+ int64_t prev_frame_end;
int stream_index;
unsigned long curr_size;
@@ -76,6 +78,8 @@ static struct ffmpeg_input *ffmpeg_input_create(void)
return NULL;
}
input->curr_pkt_size = 0;
+ input->seek_ts = -1;
+ input->prev_frame_end = -1;
input->curr_pkt_buf = input->pkt.data;
return input;
}
@@ -314,10 +318,7 @@ static int ffmpeg_fill_buffer(struct input_plugin_data *ip_data, AVFormatContext
#else
AVFrame *frame = avcodec_alloc_frame();
#endif
- int got_frame;
while (1) {
- int len;
-
if (input->curr_pkt_size <= 0) {
#if LIBAVCODEC_VERSION_MAJOR >= 56
av_packet_unref(&input->pkt);
@@ -333,78 +334,108 @@ static int ffmpeg_fill_buffer(struct input_plugin_data *ip_data, AVFormatContext
#endif
return 0;
}
- if (input->pkt.stream_index == input->stream_index) {
- input->curr_pkt_size = input->pkt.size;
- input->curr_pkt_buf = input->pkt.data;
- input->curr_size += input->pkt.size;
- input->curr_duration += input->pkt.duration;
- }
- continue;
- }
- {
- AVPacket avpkt;
- av_new_packet(&avpkt, input->curr_pkt_size);
- memcpy(avpkt.data, input->curr_pkt_buf, input->curr_pkt_size);
+ if (input->pkt.stream_index != input->stream_index)
+ continue;
+ input->curr_pkt_size = input->pkt.size;
+ input->curr_pkt_buf = input->pkt.data;
+ input->curr_size += input->pkt.size;
+ input->curr_duration += input->pkt.duration;
+
#if LIBAVCODEC_VERSION_INT >= AV_VERSION_INT(57, 48, 101)
- int send_result = avcodec_send_packet(cc, &avpkt);
- if (send_result != 0) {
- if (send_result != AVERROR(EAGAIN)) {
- d_print("avcodec_send_packet() returned %d\n", send_result);
- char errstr[AV_ERROR_MAX_STRING_SIZE];
- if (!av_strerror(send_result, errstr, AV_ERROR_MAX_STRING_SIZE ))
- {
- d_print("av_strerror(): %s\n", errstr);
- } else {
- d_print("av_strerror(): Description for error cannot be found\n");
- }
- av_packet_unref(&avpkt);
- return -IP_ERROR_INTERNAL;
+ int send_result = avcodec_send_packet(cc, &input->pkt);
+ if (send_result != 0 && send_result != AVERROR(EAGAIN)) {
+ d_print("avcodec_send_packet() returned %d\n", send_result);
+ char errstr[AV_ERROR_MAX_STRING_SIZE];
+ if (!av_strerror(send_result, errstr, AV_ERROR_MAX_STRING_SIZE ))
+ {
+ d_print("av_strerror(): %s\n", errstr);
+ } else {
+ d_print("av_strerror(): Description for error cannot be found\n");
}
- len = 0;
- } else {
- len = input->curr_pkt_size;
+ return -IP_ERROR_INTERNAL;
}
-
- int recv_result = avcodec_receive_frame(cc, frame);
- got_frame = (recv_result == 0) ? 1 : 0;
-#else
- len = avcodec_decode_audio4(cc, frame, &got_frame, &avpkt);
-#endif
-#if LIBAVCODEC_VERSION_MAJOR >= 56
- av_packet_unref(&avpkt);
-#else
- av_free_packet(&avpkt);
#endif
}
+
+#if LIBAVCODEC_VERSION_INT >= AV_VERSION_INT(57, 48, 101)
+ int recv_result = avcodec_receive_frame(cc, frame);
+ if (recv_result < 0) {
+ input->curr_pkt_size = 0;
+ continue;
+ }
+#else
+ int got_frame;
+ int len = avcodec_decode_audio4(cc, frame, &got_frame, &input->pkt);
if (len < 0) {
/* this is often reached when seeking, not sure why */
input->curr_pkt_size = 0;
continue;
}
- input->curr_pkt_size -= len;
- input->curr_pkt_buf += len;
- if (got_frame) {
- int res = swr_convert(swr,
- &output->buffer,
- frame->nb_samples,
- (const uint8_t **)frame->extended_data,
- frame->nb_samples);
- if (res < 0)
- res = 0;
- output->buffer_pos = output->buffer;
+ if (!got_frame)
+ continue;
+#endif
+
+ int64_t frame_ts = -1;
+ if (frame->pts)
+ frame_ts = frame->pts;
+ else if (frame->pkt_pts)
+ frame_ts = frame->pkt_pts;
+ else if (frame->pkt_dts)
+ frame_ts = frame->pkt_dts;
+
+ const uint8_t **in = (const uint8_t **)frame->extended_data;
+ int in_count = frame->nb_samples;
+ if (input->seek_ts > 0 && (frame_ts >= 0 || input->prev_frame_end >= 0)) {
+ struct ffmpeg_private *priv = ip_data->private;
+ AVStream *st = priv->input_context->streams[priv->input->stream_index];
+ if (frame_ts >= 0)
+ frame_ts = av_rescale_q(frame_ts, st->time_base, AV_TIME_BASE_Q);
+ else
+ frame_ts = input->prev_frame_end;
+ int64_t frame_dur = av_rescale(frame->nb_samples, AV_TIME_BASE, sf_get_rate(ip_data->sf));
+ int64_t frame_end = frame_ts + frame_dur;
+ input->prev_frame_end = frame_end;
+ d_print("seek_ts: %ld, frame_ts: %ld, frame_end: %ld\n", input->seek_ts, frame_ts, frame_end);
+ if (frame_end <= input->seek_ts)
+ continue;
+
+ /* skip part of this frame */
+ int64_t skip_samples = av_rescale(input->seek_ts - frame_ts, sf_get_rate(ip_data->sf), AV_TIME_BASE);
+ in_count -= skip_samples;
+ if (av_sample_fmt_is_planar(frame->format)) {
+ for (int i = 0; i < cc->channels; i++) {
+ in[i] += skip_samples * sf_get_sample_size(ip_data->sf);
+ }
+ } else {
+ *in += skip_samples * cc->channels * sf_get_sample_size(ip_data->sf);
+ }
+
+ input->seek_ts = -1;
+ input->prev_frame_end = -1;
+ }
+
+ int res = swr_convert(swr,
+ &output->buffer,
+ frame->nb_samples,
+ in,
+ in_count);
+ if (res < 0)
+ res = 0;
+
+ output->buffer_pos = output->buffer;
#if LIBAVCODEC_VERSION_MAJOR >= 60
- output->buffer_used_len = res * cc->ch_layout.nb_channels * sf_get_sample_size(ip_data->sf);
+ output->buffer_used_len = res * cc->ch_layout.nb_channels * sf_get_sample_size(ip_data->sf);
#else
- output->buffer_used_len = res * cc->channels * sf_get_sample_size(ip_data->sf);
+ output->buffer_used_len = res * cc->channels * sf_get_sample_size(ip_data->sf);
#endif
+
#if LIBAVCODEC_VERSION_MAJOR >= 56
- av_frame_free(&frame);
+ av_frame_free(&frame);
#else
- avcodec_free_frame(&frame);
+ avcodec_free_frame(&frame);
#endif
- return output->buffer_used_len;
- }
+ return output->buffer_used_len;
}
/* This should never get here. */
return -IP_ERROR_INTERNAL;
@@ -437,13 +468,16 @@ static int ffmpeg_seek(struct input_plugin_data *ip_data, double offset)
AVStream *st = priv->input_context->streams[priv->input->stream_index];
int ret;
- int64_t pts = av_rescale_q(offset * AV_TIME_BASE, AV_TIME_BASE_Q, st->time_base);
+ priv->input->seek_ts = offset * AV_TIME_BASE;
+ priv->input->prev_frame_end = -1;
+ int64_t ts = av_rescale(offset, st->time_base.den, st->time_base.num);
avcodec_flush_buffers(priv->codec_context);
/* Force reading a new packet in next ffmpeg_fill_buffer(). */
priv->input->curr_pkt_size = 0;
- ret = av_seek_frame(priv->input_context, priv->input->stream_index, pts, 0);
+ ret = avformat_seek_file(priv->input_context,
+ priv->input->stream_index, 0, ts, ts, 0);
if (ret < 0) {
return -IP_ERROR_FUNCTION_NOT_SUPPORTED;
|