FFmpeg Video Processing Master

Problem

Video processing often fails after the command succeeds: output is unplayable on the target platform, concatenation produces black frames, subtitles render as mojibake, audio drifts, or compressed files still exceed size limits. Listing FFmpeg flags is not enough; the workflow needs to confirm input format, target platform, quality priority, and compatibility constraints before encoding.

How It Works

The skill structures FFmpeg video work as a closed loop: diagnose the input container, codec, resolution, frame rate, and audio parameters; select a toolchain such as libx264 + aac for broad distribution, libx265 for lower-bitrate archiving, or h264_nvenc/h264_qsv for hardware acceleration; design filter chains for scale, crop, transpose, setpts/atempo, drawtext, gblur, and unsharp; and control audio/video timing with -vsync cfr, -async 1, and -copyts. Finally, it verifies that the output plays correctly, matches expected duration/resolution/frame rate, stays within size constraints, and produces a processing report.

Scope and Caveats

It fits transcoding, clipping, merging, compression, subtitle burn-in, color correction, GIF creation, and batch processing. It does not replace checks for licensed or sensitive material, HDR workflows, multi-track audio, or platform-specific delivery rules. For VFR sources, mismatched codecs, or inaccurate keyframe cuts, avoid blind -c copy; re-encode or normalize parameters first.