您必须生成一个4秒钟的视频，该视频具有与现有视频(包括时基)的参数匹配的虚拟音频，然后将concat多路分配器与streamcopy一起使用.

You'll have to generate a 4 second video with dummy audio matching the parameters of the existing video, including timebase, and then use the concat demuxer with streamcopy.

对于Q中显示的示例文件:

For the sample files shown in Q:

步骤1 生成文字视频

ffmpeg -f lavfi -r 30 -i color=black:1280x720 -f lavfi -i anullsrc -vf "drawtext="fontfile=/path/to/font.ttf:fontcolor=FFFFFF:fontsize=50:text='Your text':x=(main_w-text_w)/2:y=(main_h-text_h)/2",fade=t=in:st=0:d=1,fade=t=out:st=3:d=1" -c:v libx264 -b:v 1000k -pix_fmt yuv420p -video_track_timescale 15360 -c:a aac -ar 48000 -ac 2 -sample_fmt fltp -t 4 intro.mp4

对于WebM，将-c:v libx264替换为-c:v libvpx，将-c:a aac替换为-c:a libvorbis，将intro.mp4替换为intro.webm.您可能会删除-video_track_timescale 15360，因为WebM倾向于使用单个时标，这已经见过.

For WebM, replace -c:v libx264 with -c:v libvpx, -c:a aac with -c:a libvorbis and intro.mp4 with intro.webm. You may remove the -video_track_timescale 15360 since WebMs tend to use a single timescale, that I've seen.

步骤2 ，准备concat文件，例如list.txt

Step 2 Prepare concat file, say, list.txt

file 'intro.mp4'
file 'input.mp4'

第3步 Concat

ffmpeg -f concat -i list.txt -c copy -fflags +genpts joined.mp4

这里重要的变量是视频大小1280x720，帧速率-r 30，-pix_fmt yuv420p，采样率-ar 48000，格式-sample_fmt fltp，频道布局-ac 2，当然还有编解码器.

The variables important here are video size 1280x720, frame rate -r 30, -pix_fmt yuv420p, sample rate -ar 48000, format -sample_fmt fltp, channel layout -ac 2 and of course, codecs.