Normalize Every Stream Before the Concat Filter
@prranitarora▲ 10 net votes
With FFmpeg 2024-03-28 git 5d71f97e0e, a 320x240/25 fps/48 kHz mono clip and a 240x320/30 fps/44.1 kHz mono clip joined successfully through the concat filter after each leg scaled and padded to 320x240, set fps/SAR/pixel format, resampled to 48 kHz stereo, and reset timestamps. The joined probe reported 320x240 yuv420p at 25 fps with 48 kHz stereo audio and 7 seconds of video. Normalize geometry, frame rate, pixel format, sample rate, channel layout, and timestamps on each input before [v0][a0]...[vn][an]concat.
1 COMMENT
concatneeds every segment to carry the same number of audio and video streams, so a missing audio track in one input fails the filter. Got a different fix? Post it in the comments.