[FFmpeg-user] Slicing a 4K video to images and then encoding them back using the same codec

[Moritz Barsnick]

On Thu, Oct 15, 2015 at 16:14:40 +0000, Liran Sharoni wrote:

> But the video created is not the same at all ( I can see this using
> media info) :The container format profile is not Base Media/Version 2
> , the codec ID is not mp42, some data is not even displayed here like
> Overall bit rate mode and the list goes on.

You weren't clear with what you meant by "a nearly identical video".

First of all, ffmpeg's encoder/muxer knows little about the
demuxer/decoder. So YOU will have to provide the correct encoding
parameters.

Secondly, you need to define what your constraints are. Same
format/codec/profile? Same bitrate? Same codec parameters (there are
many!)?

You didn't provide the full uncut console output, as requested by
Kieran, so there is some guessing involved. And mediainfo output isn't
always useful. Please (also) provide the output of ffprobe or "ffmpeg
-i" for both files, and the correct command and the full console output
for the conversion process.

Most importantly: How did you manage to create a 2-second output from a
3-minute input???? Your command line, the console output, and the file
analysis (in this case: mediainfo output) must match, otherwise you are
causing a lot of confusion!

> * also I couldn’t use the drawtext filter to display the numbers from
> 1 ( only from 0 , tried using expr but it only showed me "n+1}" on
> the video)

I think you used expressions incorrectly. And I needed more backslashes
than you (only the "text=" is relevant to you):
  -vf "drawtext=fontsize=200:fontfile=c\\:\\\\Windows\\\\Fonts\\\\consola.ttf:box=1::text=%{eif\\:n+1\\:d}"
Works for me.

Cheers,
Moritz