On 15/07/15 11:56, Chris Angelico wrote: > On Wed, Jul 15, 2015 at 7:42 PM, Luca Barbato <[email protected]> wrote: >> On 15/07/15 05:23, Chris Angelico wrote: >>> I have a video of a talking head, which I wish to replace at certain >>> points with still images (slides, if you like), for X seconds. The >>> 'overlay' filter seems to be most of what I want, and I can ensure >>> that all the images are the same size as the video itself (which means >>> they completely cover the image). But after poking around in the man >>> page and on the internet, the best I can come up with is this: >> >> probably you can take the two inputs, one for the image with continuous >> loop and the other for the video and use overlay and select filters (see >> https://libav.org/documentation/avconv.html#select) >> >> The input image would be a constant stream and you use select to decide >> when to insert it and for how long. > > Sounds interesting. I'd have to have one for each image - the example > I posted was placing one still image at one location, but the purpose > was to expand that to N images, each at its correct location.
That would require N inputs, you gave me a nice idea on how to extend the select filter concept. I hope to have time to get something done in this regard soon. >> Could you please give me the output of the mpegts avconv with -v debug ? >> that part sounds quite fishy. > > Oh? You mean this? The line that returns an error =) lu _______________________________________________ libav-tools mailing list [email protected] https://lists.libav.org/mailman/listinfo/libav-tools
