On 2/19/2020 10:20 AM, Anton Khirnov wrote:
> Quoting James Almer (2020-02-18 17:13:35)
>> +static int dash_parse_prft(DASHContext *c, AVPacket *pkt)
>> +{
>> + OutputStream *os = &c->streams[pkt->stream_index];
>> + int side_data_size;
>> + AVProducerReferenceTime *prft;
>> +
>> + if (!c->write_prft)
>> + return 0;
>> +
>> + prft = (AVProducerReferenceTime *)av_packet_get_side_data(pkt,
>> AV_PKT_DATA_PRFT, &side_data_size);
>> + if (!prft || side_data_size != sizeof(AVProducerReferenceTime) ||
>> prft->flags) {
>> + // No encoder generated AVProducerReferenceTime side data. Instead
>> of letting the mov muxer
>> + // generate a capture-time one for the first packet, do it here so
>> we can also use it for the
>> + // manifest.
>> + prft = (AVProducerReferenceTime *)av_packet_new_side_data(pkt,
>> AV_PKT_DATA_PRFT,
>> +
>> sizeof(AVProducerReferenceTime));
>> + if (!prft)
>> + return AVERROR(ENOMEM);
>> + prft->wallclock = av_gettime();
>> + prft->flags = 24;
>
> It's pretty unclear what does this magic number mean.
> The flags field in AVProducerReferenceTime is also undocumented.
It's the flags field in the isobmff spec for the prft box. I can add a
comment mentioning that.
0 and 24 are the only values accepted by the dash-ll spec. The former
signals the time a frame was feed to an encoder, and the latter signals
an arbitrary but consistent time (in this case, the time a packet was
feed to a packetizer).
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".
