What is an acceptable way I can give you a sample to reproduce? You mentioned build scripts are not supported, would you like me to give you the binaries that I compiled from both the times or should I give you sample android apps for both the cases ?
On Tue, Feb 2, 2016 at 11:02 AM, Carl Eugen Hoyos <[email protected]> wrote: > Srikanth Kommineni <srikanth0569 <at> gmail.com> writes: > > > > > NDK using the guarding project for android, the result is > > > > now a 15 seconds video consistently takes about 1 minute > > > > 30 seconds to process. > > > > > > How can I reproduce this? > > > > > > [...] > > > > > > > *The following is my original build script * > > > > > > Unfortunately, build scripts are not supported here: > > > Please test with sane configure commands, if > > > possible identical ones > > > > > > > Testing with same configure commands stills results in > > slower processing speeds. > > Please provide the used configure command. > > > Does that mean the latest just takes more time to process ? > > You will have to answer above question: > How can I reproduce the regression? > > FFmpeg of course should never get slower, only faster. > In the past, some so-called "cleanups" made FFmpeg > slower for no gain, but this was by a much smaller > amount than what you reported. > > Carl Eugen > > _______________________________________________ > ffmpeg-user mailing list > [email protected] > http://ffmpeg.org/mailman/listinfo/ffmpeg-user > -- Srikanth Kommineni _______________________________________________ ffmpeg-user mailing list [email protected] http://ffmpeg.org/mailman/listinfo/ffmpeg-user
