What Is an Over-the-Shoulder Shot?

An over-the-shoulder shot is a camera angle that positions the lens behind and slightly to the side of one character's shoulder, looking toward the face of another character they're speaking to or reacting to.

The framing does two things at once. It anchors the viewer inside the conversation by keeping one character partially in frame, and it shows the other character's face clearly enough to read their expression. You always know whose perspective you're in and who they're looking at. It's one of the most used setups in narrative film, dialogue-heavy drama, interview content, and character-driven advertising.

How an over-the-shoulder shot works

The camera sits just behind the near character's ear, close enough that their shoulder and the side of their head occupy the bottom corner or edge of the frame. The far character, the one being addressed, fills roughly the center of the frame at medium or medium close-up distance.

A few variables control the feel:

Distance between characters. A longer lens compresses the space between the two people and reads as intimate or confrontational. More separation feels neutral or observational.

Camera height. Slightly above the near character's eyeline gives the far character authority. Below reverses it. Eye-level is the neutral default.

How much of the near character is in frame. A sliver of shoulder in the corner barely registers. A third of the frame filled with their back creates confinement, useful for interrogation or tense negotiation scenes.

In scripted filmmaking, OTS shots are paired with their reverse: the camera flips to put the other character's shoulder in the foreground. Editors cut between the two angles to build rhythm in a dialogue scene. In AI video, generate both as separate clips and cut between them, or prompt for a locked-off OTS and let the dialogue carry the motion.

When you use an over-the-shoulder shot

Dialogue scenes. The OTS is the default shot for any scene where two characters are talking. It maintains spatial coherence (the viewer always knows who is speaking to whom), and it preserves the emotional register of both characters without cutting away to a single.

Character POV and reaction. When you want the viewer to experience what a character sees, an OTS is a clean visual shorthand. You're not fully inside the character's head (that's a true POV shot), but you're close enough to feel their presence while still seeing the subject of their attention.

Interview and testimonial content. Brand videos and documentary work use the OTS to frame an interviewer or host listening to a subject. A blurred shoulder in the corner signals "this is a real conversation" rather than a talking head.

Power dynamics. The near character's size relative to the far character communicates authority. Placing a spokesperson in full face while a listener's shoulder anchors the edge of the frame signals credibility without voiceover.

Examples on 8frame

Veo 3.1 dialogue OTS. Veo 3.1 handles OTS framing reliably when the prompt names the setup explicitly. A prompt like "over-the-shoulder shot, woman in foreground facing away from camera, man in background facing her, medium close-up, soft interior light, shallow depth of field, 8 seconds" produces a stable locked-off frame with the near shoulder slightly out of focus and the far character's face in sharp focus. The model maintains that depth relationship without drifting if the clip is kept under 10 seconds. For the full prompt structure and motion language that works with Veo 3.1, see Veo 3 prompt guide.

Higgsfield Soul 2 character storytelling. Higgsfield Soul 2 handles OTS framing with strong face fidelity on the far subject. A prompt like "over-the-shoulder shot from behind a man in a dark coat, facing a woman lit by a single window, intense emotional exchange, film grain, desaturated palette, 6 seconds" gives you a scene that reads as authored. The far character's face detail holds across the clip, which matters when expression is the performance. See Higgsfield Soul 2 prompts for character stories for the prompt patterns that produce the most consistent output.

Reverse OTS pair. To build a full dialogue sequence, generate two clips with matching lighting and camera height: one with Character A's shoulder in the foreground, one with Character B's. Veo 3.1 maintains enough spatial consistency between paired OTS clips that the cut reads as motivated rather than assembled.

Related concepts

• Veo 3 prompt guide covers how to write framing and depth-of-field language for Veo 3.1, including OTS, POV, and medium close-up setups.
• Higgsfield Soul 2 prompts for character stories shows the full prompt system for generating emotionally consistent character-driven clips.

---

Want to generate an OTS scene? Open the canvas on 8frame and prompt Veo 3.1 or Higgsfield Soul 2 with the angle named directly.