Copying keeps what came in
When copying, Corneta forwards the video it receives without recreating its frames. That cuts CPU and GPU work, but it doesn't change resolution, FPS, video bitrate or framing.
Audio can be converted to the output format, or processed by the normalizer when it is turned on. The video mode alone doesn't decide whether the audio gets transcoded.
Transcoding creates a new output
Transcoding decodes the video and produces a separate live output. It's the way to go when:
- a platform accepts less bitrate;
- a destination needs a different resolution or FPS;
- a vertical stream requires a 9:16 crop;
- the incoming codec doesn't work on the platform.
The decision in three questions
- Do all of them accept exactly the signal that came in? Use Straight up.
- Do only one or two need adjusting? Use Smart.
- Do you want an independent preset per platform, and can your machine handle it? Use All out.
Don't pick All out just because the name sounds better. If three outputs already accept the same 1080p60, transcoding three times can burn GPU without any visible difference.
See the names and the estimate in Help: which quality mode to choose.