[FFmpeg-user] Efficiently doing thousands of edits?
dopelabs at dubstep.fm
Thu Apr 11 14:27:06 EEST 2019
cropping out the problem quarter to a separate file seems like it could be faster... less pixels to black detect.
if you know the start and stop times/frames, and also keep in mind that:
OPTIONS FOR FILTERS WITH SEVERAL INPUTS
Some filters with several inputs support a common set of options. These options can only be set by name, not with the short notation.
The action to take when EOF is encountered on the secondary input; it accepts one of the following values:
Repeat the last frame (the default).
End both streams.
Pass the main input through.
If set to 1, force the output to terminate when the shortest input terminates. Default value is 0.
If set to 1, force the filter to extend the last frame of secondary streams until the end of the primary stream. A value of 0 disables this behavior. Default value is 1.
setting eof_action=repeat might just do what your looking for...
though it would be nice to know which filters 'some filters' are...
also the tpad filter looks like it would be a good contender as well..
you can pad either before or after the input stream, and use the clone mode which would be clones of the first or last frame respectively.
tbh, after seeing the yt video, i personally would probably just overlay the bottom left with solid black once it starts to flicker as i, for the most part, consider that unwatchable anyways =]
if this is an ongoing problem, maybe its time for a new camera? =]
> On Apr 11, 2019, at 3:01 28AM, Bouke <bouke at editb.nl> wrote:
>> On 11 Apr 2019, at 04:44, John Hawkinson <jhawk at alum.mit.edu> wrote:
>> Michael Shaffer <mikeshaffer at gmail.com> wrote on Wed, 10 Apr 2019
>> at 19:40:36 -0400 in <CAMrzi1s55GuHbXoWW+r6XibbDKdtsHa=x04vf5DO4fhUqUh4Jg at mail.gmail.com>:
>>> I'm pretty sure you could use Python and OpenCV to create a solution..
>> "Just because you can, doesn't mean you should."
>> My original post explained how to determine where the runs of black are, using ffmpeg (libavfilter)'s "blackdetect." The only piece left is applying the edits in a practical fashion.
>> Rolling your own with opencv would be a whole lot more work. I'm not entirely sure that I agree with Carl Eugen that ffmpeg isn't a video editing tool, but if it's not, Python certainly isn't. Yes, you can make it work. But you're going to spend a lot more time doing it than if higher level tools were used.
>> The goal here is for higher-level tools, not lower-level ones.
>> Or I might just end up using ffmpeg to split it into 7,000 files and then concatenating them. That's hardly the worst thing in the world.
> Why split and cat? You are not doing editing, you want to mask (little) pieces, not shorten / extend / reshuffle, right?
> If you have all the in/out point, why not do a (png or alike) overlay on those points?
> Or, a subtitle overlay with a strange custom font (one big black rectangle char).
>> jhawk at alum.mit.edu
>> John Hawkinson
>> 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".
> 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