[FFmpeg-user] ffmpeg serving to Avisynth

Frank Tetzel s1445051 at mail.zih.tu-dresden.de
Mon Sep 14 21:39:34 CEST 2015

> > What processing do you want to do with ffmpeg?
> At a minimum I foresee concatenating input files with it. My current
> project has 350+ video files. Avisynth cannot work with more than 25
> - 35 without crashing. I have not found an Avisynth mechanism for
> concatenating files that preserves audio.

There are multiple ways to concatenate files depending on the input
codecs and what other processing you want to do with it [4]. Not sure
if it handles hundreds of input files well enough.

> > And why do you want to
> > send it over tcp, if that's what TCPSource reads (not an avisynth
> > user)?
> To avoid intermediate storage. Workspace for this project is 2TB.
> Each additional version of the project is currently ~700GB. Some form
> of inter-process communication is required to avoid intermediate
> storage. TCPSource() seems the only type of built-in IPC input
> Avisynth supports.

I don't know which data layout they expect in TCPSource and if it is in
any way compatible with the tcp output protocol in ffmpeg, or any other
protocol. I know this was your question in the first place but i can't
help you there. You could play around and just try to connect [1][2].

There's also some avisynth support in ffmpeg [3]. As i never used it i
don't know about its capabilities.

What are you doing after processing with avisynth? Do you pipe it back
into ffmpeg for encoding? Can't you use built-in filters [5] instead of
an avisynth script?

[1] http://avisynth.nl/index.php/TCPServer
[2] http://ffmpeg.org/ffmpeg-protocols.html#tcp
[3] http://www.ffmpeg.org/faq.html#How-can-I-read-DirectShow-files_003f
[4] http://trac.ffmpeg.org/wiki/Concatenate
[5] http://ffmpeg.org/ffmpeg-filters.html#Description

More information about the ffmpeg-user mailing list