New version of vfe.sh works with ffmpeg v1 and up. Overall the changes to ffmpeg increase the similarity between the ffmpeg and the libav suites, so that makes my life a little easier. I did not try for backward compatibility. If you need to use it with ffmpeg 0.6, then you need to use the 2.0 version. You can get either version from the repo on github:
The latest version of vfe.sh (2.0) allows you to select whether the script will run the transcoding commands using an 'ffmpeg' or 'avconv' command. The default is ffmpeg, and it's the kind of thing one is likely set and leave, so I have added a 'converter' value in my .vferc file. On my Mac I'm using ffmpeg, and on my Ubuntu machine avconv. With avconv I found that Ogg encoding was sensitive to the audio bitrate, and that 128 was a good value. At any rate, it is now a setting that you can adjust on the command line or in your .vferc file. Lastly I added more explicit output so that you can now see in the terminal the full form of the command that the script ran for each video type.
Updated files are on:
I just pushed a new version of the vfe.sh script to github:
This version (1.11) drops the use of ffmpeg2theora. All of the relevant functionality of this utility is available through ffmpeg. For my script the switch has these advantages:
- Fewer dependencies
- Greater range of input formats now possible
- Consistent and predictable handling of anamorphic pixel aspect ratios
Point #2 lets me work more easily with kdenlive, which I have recently begun using for video editing. Point #3 makes it easier to handle input from a wider range of cameras.