[FFmpeg-user] SCTE-35 passthrough in transcoding process

Alessandro Molon alex.molon at vision247.com
Fri Jun 12 17:36:56 EEST 2020

HI All,

Devin I would be really interested in your patches,
SCTE-35 is a topic that I'm really interested to .


-----Original Message-----
From: ffmpeg-user <ffmpeg-user-bounces at ffmpeg.org> On Behalf Of Devin Heitmueller
Sent: 21 January 2020 14:21
To: FFmpeg user questions <ffmpeg-user at ffmpeg.org>
Subject: Re: [FFmpeg-user] SCTE-35 passthrough in transcoding process

> That’s what we were seeing from our searching as well. I do know that there are vendors that output cable channels after decryption with a cable card, with SCTE-35 that are using FFMPEG in the backend code. Silicon Dust with the now discontinued PRIME units for an example. Does anyone know how this is being facilitated?

With some custom code it isn't hard to pass through SCTE-35 triggers, largely dependent on whether you are reusing the stream timing from the original incoming stream.  If your output stream as the same timestamps as the incoming stream (i.e. using -copy_ts), then you can essentially just copy the packet to the output unmodified without having to parse the SCTE-35 payload.  However if you're retiming the stream then you need to parse the packet, adjust the pts_offset field, and then create a new packet.

And of course you need to modify the MPEG TS mux to adjust the PMT and create the table sections, since SCTE-35 is done through tables as opposed to through an elementary stream containing PES packets.

I've got patches which do this but they aren't upstream.


Devin J. Heitmueller - Kernel Labs
ffmpeg-user mailing list
ffmpeg-user at ffmpeg.org

To unsubscribe, visit link above, or email ffmpeg-user-request at ffmpeg.org with subject "unsubscribe".

More information about the ffmpeg-user mailing list