Horizontal video
Also known as: landscape video, widescreen video
Video wider than it is tall, the native shape of every camera and the format that everything except a phone feed expects.
Updated
What it means
Horizontal is the native shape. Every action camera, drone and cinema body records it, and a vertical clip is always something done to a horizontal frame afterwards.
It suits the subject. Landscape, a trail, a wave and the space around a jump all extend sideways, which is why action footage and this shape fit each other without any effort.
It is the counterpart of Vertical video (9:16) rather than an alternative to it. One is what cameras produce and the other is what feeds consume, and both of them exist in most workflows now.
How it works in practice
Filming horizontal keeps both options open. A wide frame crops to a tall one with the subject tracked, where the reverse recovers nothing that was never recorded in the first place.
Delivery is still mostly this shape. A session edit watched on a laptop or a television wants it, and an Aspect ratio crop produces the feed version from the same master.
Composition changes with the shape. Rule of thirds placement that works wide puts a subject off the edge of a tall crop, so framing allows for both when both of them are planned for.
What to watch out for
Phone feeds penalise it. A horizontal video in a vertical slot occupies a small strip of the screen, and the result is read as effort nobody made rather than as a format choice.
Rotation metadata gets lost. A clip recorded with the camera on its side relies on a rotation flag in the Video metadata, and a tool that ignores the flag plays the clip sideways.
Nobody wants two edits. Delivering both shapes doubles the finishing work, which is why the crop is applied to a finished sequence rather than produced as a separate edit.
How RawClip handles it
Sequences are assembled in the shape the footage was recorded in, which for action cameras and drones is this one. A vertical or square version is produced by cropping that result with the subject tracked rather than by starting the edit again from the footage.