Prompt From Video / Scene intelligence
Build Veo prompts that describe the shot and its audio, including dialogue, ambience, music, and synchronized sound effects.
Track a red rally car through a desert bend. Keep the body sharp as the dust plume expands; tire noise, engine rise, and wind remain synchronized…
01The output stays grounded in visible evidence: shots, timing, movement, lighting, and sound.
Synchronized pictures, dialogue, and sound
A strong Veo prompt treats audio as part of the shot design, not an afterthought. Describe the visible scene first, then specify spoken words, vocal delivery, ambience, music, and timed effects. Put dialogue in quotation marks and identify who speaks. If nobody speaks, say “no dialogue” so environmental sound is not mistaken for narration.
Veo handles cinematic direction well when the prompt has a clear temporal arc. Establish the opening composition, describe the important change, and define the final beat. Lens, focus, and camera movement should support that arc. A rack focus is useful when attention passes between subjects; a locked wide shot is better when sound carries the moment.
Avoid contradictory sound requests. “Silent room” cannot also contain loud crowd ambience, and background music may compete with intelligible speech. The Veo format isolates audio cues while keeping synchronization explicit.
Write the exact line, speaker, tone, and whether lips are visible.
Connect a click, impact, footstep, or swell to the visible event that causes it.
Specify room tone, weather, distant activity, and whether music is diegetic.
State no captions, subtitles, logos, or on-screen text when they are unwanted.
Editable prompt framework
Draft the visual shot before writing the soundtrack. This template keeps dialogue and effects attached to visible causes, which makes the prompt easier to review for timing conflicts. The field names are an editing checklist; combine the completed notes into natural direction before you generate.
Template structure
Visual: [location, time, opening frame, and subject]. Action: [chronological change and final beat]. Camera: [framing, focus, and one move]. Dialogue: [speaker, exact line, and delivery] or no dialogue. Audio: [effects tied to actions], [ambience], [music or no music]. Exclude [captions, logos, or unwanted sound].
Establish location, time, framing, and the people or objects visible at the start. Specify the speaker’s position when dialogue matters. Keep the number of visual decisions realistic for the clip duration instead of describing an entire screenplay in one generation.
Describe what changes in chronological order and where the shot ends. Pair that change with one motivated camera move, focus shift, or locked composition. A clear final beat helps dialogue, reactions, and camera timing resolve in the same moment.
Put the exact spoken line in quotation marks, name the speaker, and state the delivery. Note whether the face is visible for lip sync. Keep the words short enough to be spoken naturally inside the requested duration, and say no dialogue when nobody should speak.
Tie each important sound to its source: a latch clicks as a door closes, shoes cross tile, or birds lift from a tree. Then define the quieter sound bed and music choice. Remove captions, logos, voiceover, or music explicitly when they would spoil the intended output.
Read the visual and audio instructions as a single timeline. If a line of dialogue takes most of the clip, simplify the camera move and background action. If an impact should land on a reveal, state both events together. Specific synchronization usually contributes more than a long list of unrelated sound adjectives.
For a reference video, keep observed shot order, lip movement, effects, and ambience separate from details you want to change. You might preserve a rack focus and cup clink while replacing the location and characters. The Veo prompt generator exposes those parts so the final prompt can remain coherent after the creative substitutions.
Review unwanted output as a source problem. Accidental subtitles call for a direct no-captions instruction; muddled speech may mean the line is too long for the shot; overpowering music may need a lower role or removal. Change that one audio relationship before changing the visual prompt. The goal is not more sound detail, but a mix in which every requested source has room.
Locked medium two-shot in a nearly empty chrome diner at 2 a.m. A tired server sets down coffee and says softly, “You missed the last train.” The traveler looks toward the rain-streaked window. Natural lip sync; ceramic cup clink, refrigerator hum, rain outside, no music, no captions.
A slow Steadicam move follows a child through waist-high ferns into a sunlit clearing where hundreds of birds lift at once. Wingbeats erupt exactly on the reveal, then recede into wind and insects. No dialogue; restrained orchestral note rises only after the birds launch.
Macro close-up of caramel cracking under a spoon. The shell breaks on contact with a crisp snap, followed by a soft custard sound. Warm restaurant ambience remains low, shallow focus, controlled push-in, no voiceover or text.
Reference workflow
Upload a short reference or paste a public link in the workspace above. The analyzer follows the timeline rather than guessing from a thumbnail: it records shot order, subject motion, camera behavior, light, transitions, dialogue, ambience, and music.
Select the Veo 3 result to see the same evidence written in this model’s preferred syntax. Keep the facts that define the reference, then edit creative choices such as subject, location, palette, or ending. The result remains a new production brief, not a promise of a frame-for-frame copy.
Quality control
Quote the exact words, name the speaker, and describe delivery. Keep the exchange short enough for the requested shot duration.
Yes. Identify the sound source and connect important effects to visible actions for better synchronization.
State no dialogue, no music, and only the specific room tone or environmental sound you want; request complete silence only when that is truly the goal.
Yes. You can use the on-page template to draft a Veo prompt manually or use the limited free reference analysis to extract one, then revise its visual and audio direction.
Yes. Quote the line, identify the speaker and delivery, and connect it to the visible performance and shot duration. The template also separates ambience, effects, and music.
