Video Trimming Guide: How to Cut Videos Without Re-encoding
Stream Copy vs. Re-encoding
Stream copy trims a video by copying the existing encoded data without re-encoding — fast (seconds, not minutes) and lossless. The limitation: cuts must align with keyframes (I-frames). Re-encoding is necessary for frame-accurate cuts or when changing the container format. For most use cases, stream copy is the right choice.
Keyframe Alignment
Videos are encoded as keyframes (I-frames) and predicted frames (P and B-frames). You can only cut precisely at keyframes without re-encoding. Typical keyframe intervals: 2 seconds for online video, 1 second for screen recordings, 10+ seconds for some formats. If your trim point isn't at a keyframe, the video is rounded to the nearest one.
When Re-encoding Is Necessary
Frame-accurate cuts (sub-keyframe precision). Removing a watermark that's part of the video. Changing aspect ratio or resolution. Converting to a different codec. Merging multiple clips with different settings. In these cases, quality loss from re-encoding is unavoidable — minimize it by using high-quality settings.