ijkplayer和ffplay在播放ts流时起播慢的原因分析

ijkplayer和ffplay在播放ts流时起播慢的原因分析

1.前言

在优化ijkplayer起播时间的过程中,发现设置probesize和analyzeduration可以减少起播时间,在播放4k ts流时,发现需要将probesize和analyzeduration设置的很大才能找到音视频信息(通过av_dump_format)并播放,并且发现设置的很大的时候播放1080P的时候就变得很慢,发是在avformat_find_stream_info函数中很耗时,probesize和analyzeduration这两个值是用于探测数据里面的音视频流信息,一个做了大小限制一个做了时间限制,只要其中一个超出,则退出avformat_find_stream_info中的一个循环,我就在想,按照ffmpeg的逻辑,找到了音视频信息就会退出,那么probesize和analyzeduration这两个值应该和起播时间没有关系才对,为什么这两个值设置过大会导致起播时间延长呢?

2.排查过程

2.1 avformat_find_stream_info函数

先看一下avformat_find_stream_info函数中最主要的一个循环(精简了一下):

for (;;) {
        int analyzed_all_streams;
        /* check if one codec still needs to be handled */
        for (i = 0; i < ic->nb_streams; i++) {
            int fps_analyze_framecount = 20;
            
            st = ic->streams[i];
            char *errmsg = NULL; 
            if (!has_codec_parameters(st, errmsg)) {
                av_log(ic, AV_LOG_INFO, "stream_index:%d %s\n", errmsg);
                break;
        }
        analyzed_all_streams = 0;
        if (i == ic->nb_streams) {
            analyzed_all_streams = 1;
            /* NOTE: If the format has no header, then we need to read some
             * packets to get most of the streams, so we cannot stop here. */
            if (!(ic->ctx_flags & AVFMTCTX_NOHEADER)) {
                /* If we found the info for all the codecs, we can stop. */
                ret = count;
                av_log(ic, AV_LOG_DEBUG, "All info found\n");
                flush_codecs = 0;
                break;
            }
        }
        /* We did not get all the codec info, but we read too much data. */
        if (read_size >= probesize) {
            ret = count;
            av_log(ic, AV_LOG_DEBUG,
                   "Probe buffer size limit of %"PRId64" bytes reached\n", probesize);
            for (i = 0; i < ic->nb_streams; i++)
                if (!ic->streams[i]->r_frame_rate.num &&
                    ic->streams[i]->info->duration_count <= 1 &&
                    ic->streams[i]->codecpar->codec_type == AVMEDIA_TYPE_VIDEO &&
                    strcmp(ic->iformat->name, "image2"))
                    av_log(ic, AV_LOG_WARNING,
                           "Stream #%d: not enough frames to estimate rate; "
                           "consider increasing probesize\n", i);
            break;
        }

        /* NOTE: A new stream can be added there if no header in file
         * (AVFMTCTX_NOHEADER). */
        ret = read_frame_internal(ic, &pkt1);
        if (ret == AVERROR(EAGAIN))
            continue;

        if (ret < 0) {
            /* EOF or error*/
            eof_reached = 1;
            break;
        }

        pkt = &pkt1;
        st = ic->streams[pkt->stream_index];
        if (!(st->disposition & AV_DISPOSITION_ATTACHED_PIC))
            read_size += pkt->size;
        }
        if (st->codec_info_nb_frames>1) {
            int64_t t = 0;
            int64_t limit;

            if (st->time_base.den > 0)
                t = av_rescale_q(st->info->codec_info_duration, st->time_base, AV_TIME_BASE_Q);
            if (st->avg_frame_rate.num > 0)
                t = FFMAX(t, av_rescale_q(st->codec_info_nb_frames, av_inv_q(st->avg_frame_rate), AV_TIME_BASE_Q));

            if (   t == 0
                && st->codec_info_nb_frames>30
                && st->info->fps_first_dts != AV_NOPTS_VALUE
                && st->info->fps_last_dts  != AV_NOPTS_VALUE)
                t = FFMAX(t, av_rescale_q(st->info->fps_last_dts - st->info->fps_first_dts, st->time_base, AV_TIME_BASE_Q));

            if (analyzed_all_streams)                                limit = max_analyze_duration;
            else if (avctx->codec_type == AVMEDIA_TYPE_SUBTITLE) limit = max_subtitle_analyze_duration;
            else                                                     limit = max_stream_analyze_duration;

            if (t >= limit) {
                av_log(ic, AV_LOG_VERBOSE, "max_analyze_duration %"PRId64" reached at %"PRId64" microseconds st:%d\n",
                       limit,
                       t, pkt->stream_index);
                if (ic->flags & AVFMT_FLAG_NOBUFFER)
                    av_packet_unref(pkt);
                break;
            }
            if (pkt->duration) {
                if (avctx->codec_type == AVMEDIA_TYPE_SUBTITLE && pkt->pts != AV_NOPTS_VALUE && pkt->pts >= st->start_time) {
                    st->info->codec_info_duration = FFMIN(pkt->pts - st->start_time, st->info->codec_info_duration + pkt->duration);
                } else
                    st->info->codec_info_duration += pkt->duration;
                st->info->codec_info_duration_fields += st->parser && st->need_parsing && avctx->ticks_per_frame ==2 ? st->parser->repeat_pict + 1 : 2;
            }
        }
        
        /* If still no information, we try to open the codec and to
         * decompress the frame. We try to avoid that in most cases as
         * it takes longer and uses more memory. For MPEG-4, we need to
         * decompress for QuickTime.
         *
         * If AV_CODEC_CAP_CHANNEL_CONF is set this will force decoding of at
         * least one frame of codec data, this makes sure the codec initializes
         * the channel configuration and does not only trust the values from
         * the container. */
        try_decode_frame(ic, st, pkt,
                         (options && i < orig_nb_streams) ? &options[i] : NULL);

        if (ic->flags & AVFMT_FLAG_NOBUFFER)
            av_packet_unref(pkt);

        st->codec_info_nb_frames++;
        count++;
    }

要跳出最外层的for循环,三个条件满足一个即可:
① i == ic->nb_streams 并且 !(ic->ctx_flags & AVFMTCTX_NOHEADER)
② read_size >= probesize
③ t >= limit
2和3就是上面说的一个大小和一个时间限制,然后查看之前的起播日志,发现都是在2或者3条件满足下跳出循环的,也就是说每次都是大小超出或者时间超出才退出循环的。就感觉很奇怪,为什么每次都要超时才能退出循环呢?查看条件1下有条注释和日志输出:

analyzed_all_streams = 1;
/* NOTE: If the format has no header, then we need to read some
 * packets to get most of the streams, so we cannot stop here. */
if (!(ic->ctx_flags & AVFMTCTX_NOHEADER)) {
    /* If we found the info for all the codecs, we can stop. */
    ret = count;
    av_log(ic, AV_LOG_DEBUG, "All info found\n");
    flush_codecs = 0;
    break;
}

看这个意思就是如果流是没有header的,就会继续读更多的包,然后加日志发现,i == ic->nb_streams这个条件在解码几帧视频后满足,但是ic->ctx_flags这个值一直都为1,导致这里永远不会走到break,看下i == ic->nb_streams这个条件是如何满足的:

for (i = 0; i < ic->nb_streams; i++) {
    int fps_analyze_framecount = 20;
    
    st = ic->streams[i];
    char *errmsg = NULL; 
    if (!has_codec_parameters(st, errmsg)) {
        av_log(ic, AV_LOG_INFO, "stream_index:%d %s\n", errmsg);
        break;
}

上面的日志是我加的,原生ffmpeg没有这个日志,这里主要判断每一路流,有没有相应的解码信息,这个函数定义如下:

static int has_codec_parameters(AVStream *st, const char **errmsg_ptr)
{
    AVCodecContext *avctx = st->internal->avctx;

#define FAIL(errmsg) do {                                         \
        if (errmsg_ptr)                                           \
            *errmsg_ptr = errmsg;                                 \
        return 0;                                                 \
    } while (0)

    if (   avctx->codec_id == AV_CODEC_ID_NONE
        && avctx->codec_type != AVMEDIA_TYPE_DATA)
        FAIL("unknown codec");
    switch (avctx->codec_type) {
    case AVMEDIA_TYPE_AUDIO:
        if (!avctx->frame_size && determinable_frame_size(avctx))
            FAIL("unspecified frame size");
        if (st->info->found_decoder >= 0 &&
            avctx->sample_fmt == AV_SAMPLE_FMT_NONE)
            FAIL("unspecified sample format");
        if (!avctx->sample_rate)
            FAIL("unspecified sample rate");
        if (!avctx->channels)
            FAIL("unspecified number of channels");
        if (st->info->found_decoder >= 0 && !st->nb_decoded_frames && avctx->codec_id == AV_CODEC_ID_DTS)
            FAIL("no decodable DTS frames");
        break;
    case AVMEDIA_TYPE_VIDEO:
        if (!avctx->width)
            FAIL("unspecified size");
        if (st->info->found_decoder >= 0 && avctx->pix_fmt == AV_PIX_FMT_NONE)
            FAIL("unspecified pixel format");
        if (st->codecpar->codec_id == AV_CODEC_ID_RV30 || st->codecpar->codec_id == AV_CODEC_ID_RV40)
            if (!st->sample_aspect_ratio.num && !st->codecpar->sample_aspect_ratio.num && !st->codec_info_nb_frames)
                FAIL("no frame in rv30/40 and no sar");
        break;
    case AVMEDIA_TYPE_SUBTITLE:
        if (avctx->codec_id == AV_CODEC_ID_HDMV_PGS_SUBTITLE && !avctx->width)
            FAIL("unspecified size");
        break;
    case AVMEDIA_TYPE_DATA:
        if (avctx->codec_id == AV_CODEC_ID_NONE) return 1;
    }

    return 1;
}

可以看到该函数就是检查音视频的一些参数,视频检查宽高,像素格式,音频检查帧大小,采样率采样格式等,只要有一项不满足,就返回0,就跳出该for循环,i就不会等于ic->nb_streams,就会继续解码音视频帧直到获取到上面这些参数,i就等于ic->nb_streams了,这里的含义清楚了,只要音视频的以上参数有了就会跳出循环,但是ic->ctx_flags一直为1,所以需要找到为什么ic->ctx_flags一直为1即可。

2.1 mpegts.c文件分析

在该文件中搜索ic->ctx_flags,发现有两个地方对其赋值,一个是mpegts_read_header中将其赋值为1:

if (s->iformat == &ff_mpegts_demuxer) {
        /* normal demux */

        /* first do a scan to get all the services */
        seek_back(s, pb, pos);

        mpegts_open_section_filter(ts, SDT_PID, sdt_cb, ts, 1);

        mpegts_open_section_filter(ts, PAT_PID, pat_cb, ts, 1);

        handle_packets(ts, probesize / ts->raw_packet_size);
        /* if could not find service, enable auto_guess */

        ts->auto_guess = 1;

        av_log(ts->stream, AV_LOG_TRACE, "tuning done\n");

        s->ctx_flags |= AVFMTCTX_NOHEADER;
}

该函数会在avformat_open_input中被调用(在找到相应的AVInpuFormat后):

 if (!(s->flags&AVFMT_FLAG_PRIV_OPT) && s->iformat->read_header)
     if ((ret = s->iformat->read_header(s)) < 0)
         goto fail;

还有一个是在handle_packet函数中对其赋值:

if (ts->stream->ctx_flags & AVFMTCTX_NOHEADER && ts->scan_all_pmts <= 0) {
    int i;
    for (i = 0; i < ts->nb_prg; i++) {
        if (!ts->prg[i].pmt_found)
            break;
    }
    if (i == ts->nb_prg && ts->nb_prg > 0) {
        int types = 0;
        for (i = 0; i < ts->stream->nb_streams; i++) {
            AVStream *st = ts->stream->streams[i];
            if (st->codecpar->codec_type >= 0)
                types |= 1<<st->codecpar->codec_type;
        }
        if ((types & (1<<AVMEDIA_TYPE_AUDIO) && types & (1<<AVMEDIA_TYPE_VIDEO)) || pos > 100000) {
            av_log(ts->stream, AV_LOG_DEBUG, "All programs have pmt, headers found\n");
            ts->stream->ctx_flags &= ~AVFMTCTX_NOHEADER;
        }
    }
}

于是添加打印发现ts->scan_all_pmts为1,永远不会走进来,最终发现在ijkplayer中的read_thread函数中有一段代码(查看原生ffplay中也有该语句):

if (!av_dict_get(ffp->format_opts, "scan_all_pmts", NULL, AV_DICT_MATCH_CASE)) {
    av_dict_set(&ffp->format_opts, "scan_all_pmts", "1", AV_DICT_DONT_OVERWRITE);
    scan_all_pmts_set = 1;
}

这里将scan_all_pmts设置为了1,最终在avformat_open_input中设置到了mpegts里面MpegTSContext中的scan_all_pmts 变量中:

/* Allocate private data. */
if (s->iformat->priv_data_size > 0) {
    if (!(s->priv_data = av_mallocz(s->iformat->priv_data_size))) {
        ret = AVERROR(ENOMEM);
        goto fail;
    }
    if (s->iformat->priv_class) {
        *(const AVClass **) s->priv_data = s->iformat->priv_class;
        av_opt_set_defaults(s->priv_data);
        if ((ret = av_opt_set_dict(s->priv_data, &tmp)) < 0)
            goto fail;
    }
}

上面的av_opt_set_dict会将tmp字典中的"scan_all_pmts"对应字段的值设置到MpegTSContext中去,这是ffmpeg的AVClass结构和功能,很nice。
上面说到了scan_all_pmts这个值为1,导致ic->ctx_flags一直为1,一直没有跳出avformat_find_stream_info函数,导致起播变慢,将其改为0,起播时间果然变快了,但是为什么ijkplayer和ffplay要将其设置为1呢,分析了下mpegts.c,发现如果scan_all_pmts为1,会分析并合并ts流中所有的pmt,但是在直播中,ts会一直传输,并且其中隔段时间就会有pat和pmt,所以一直不会分析完,导致起播变慢。
如果scan_all_pmts为0的话,上面就会检查pat中的pmt是否都收到,并且音视频格式都有的话,就会赋值ic->ctx_flags为0,最终就会退出avformat_find_stream_info中的循环了。
经过以上分析,猜测一下代码作者加入scan_all_pmts变量的想法应该跟以前的dvb那种ts流有关,那个时候ts流中有很多节目表(pmt),这个时候设置scan_all_pmts为1,就可以找到这个流中的所有节目信息,而现在网络播放中,ts流一般就一个节目,所以需要设置scan_all_pmts为0,找到一个pat后,解析其中的pmt,后面收到该pmt后就退出了。

结论:
将scan_all_pmts设置为0后,在mpegts.c里面的pat_cb回调中,会检查pat中pmt表中的数量,并记录每个pmt对应的pid,在pmt_cb回调中,会设置对应Program中的pmt_found字段,在handle_packet中会检查如果pmt的pmt_found都置1了,并且stream中音视频已经找到,就会将ic->ctx_flags置为1,从而影响到avformat_find_stream_info中的循环。

  • 2
    点赞
  • 5
    收藏
    觉得还不错? 一键收藏
  • 0
    评论

“相关推荐”对你有帮助么?

  • 非常没帮助
  • 没帮助
  • 一般
  • 有帮助
  • 非常有帮助
提交
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值