I have been experimenting with creating a video from a sequence of images.
When I use the suggested ffmpeg method:
ffmpeg -f image2 -i image%d.jpg video.mpg
The video is not really as good as I hoped it would be :/
For example with VDUB, If I export avi from the same image sequence its perfect qulality, however the file size can be huge if theres a lot of jpegs.
In the past I have used x264 gui front-ends such as staxrip and the video produced from an uncompressed AVI is very exceptional and the compression is very very good, relitively tiny output files (mp4).
So what is the best way to compress the image sequence so that there is very high quality? Surely there is something better than ffmpeg? is it possible to use the x264 from an image sequence as you would with ffmpeg, and get higher quality? FYI i will be excecuting the task from within a c#.net project using startprocess();
The quality of that command's output is bad for a few reasons:
Try something like:
ffmpeg -f image2 -i image%d.jpg -vcodec mpeg4 -b 800k video.avi
for mpeg 4 video or:
ffmpeg -f image2 -i image%d.jpg -vcodec libx264 -b 800k video.avi
for H.264 video (You will have to have libx264 installed for this to work). You can play around with the bitrate because it depends on the size of your frames what bitrate you will need. Also, running ffmpeg -formats
will display all of the output formats and codecs if you want to experiment more.
See the ffmpeg documentation for even more options.