问题描述
我正在创建这样的一份切割清单
I am creating a cutlist like this
file 't1.MP4'
inpoint 1.45067
outpoint 3.18133
file 't2.MP4'
inpoint 3.45067
outpoint 6.18133
...
我这样使用它
ffmpeg -y -f concat -safe 0 -i list.txt -c:v libx264 -c:a aac -af aresample=async=1 -pix_fmt yuv420p output.mp4
输出几乎总是在开始时包含静止图像，有时是视频时间的一半，有时更少，显示的是第一帧，然后是最后的视频.我注意到，inpoint和outpoint之间的距离越少，开始时的静止图像就越长.也许它也与inpoint和outpoint的数量有关.
The output almost always contains a still image at the start, sometimes half the video time, sometimes less, showing the first frame then the final video starts. I noticed, the less inpoint and outpoint are apart, the longer the still image at the start. Maybe its also connected to the amount of inpoint and outpoint.
我该如何解决?
推荐答案
由于视频将需要更早的帧进行解码，因此concat多路分配器输出的数据包超出了范围内指定的数据包.
The concat demuxer is outputting more packets than that specified in the ranges, since video will need earlier frames to decode.
尝试
ffmpeg -y -f concat -safe 0 -copyts -segment_time_metadata 1 -i list.txt -vf select=concatdec_select,setpts=N/FRAME_RATE/TB,format=yuv420p -c:v libx264 -af aselect=concatdec_select,asetpts=N/SR/TB -c:a aac output.mp4
我假设所有文件的视频和音频都具有相同的属性，例如帧率，分辨率，音频采样率，通道等.
I assume all files have videos and audios with same properties i.e. framerate, resolution, audio sampling rate, channels..etc
这篇关于Concat在视频之前创建静止图像的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持！