seedance-car-vehicle

Cars and vehicles at speed · Cars and Vehicles Speed Clips

The machine has to stay one machine while the camera does all the work. Lock the vehicle part by part, then fill the runtime with a numbered cut list that moves the camera every second.

This Skill carries one prompt structure distilled from 17 human-verified Seedance cases on goodcase.ai. It is a sibling of seedance-prompt-library (all templates in one Skill); install this one when you only want this kind of clip. Do not invent structure from general video-generation knowledge: follow the structure below and ground the draft in one anchor case from references/cases.md.

Use when

• Motorcycle and car commercials, mountain-road speed runs, chase and stunt sequences, and vehicle transformation clips.
• Chinese: Motorcycle and car commercials, mountain-road speeding, chase and stunt segments, and vehicle-transformation clips.

Workflow

1. Collect inputs. Ask only for what is missing: subject or product, setting, duration, aspect ratio, whether reference images / audio exist, and which language the final prompt should be in.
2. Pick one anchor case. Read references/cases.md, choose the case closest to the request, and name it in the reply. One anchor, never an average of several.
3. Fill the structure block by block. Every block below must be present in the final prompt; an empty block is where prompts go vague.
   1. Opening line: runtime, aspect ratio, frame rate, and the exact number of cuts
   2. Vehicle lock: model or type, colour, and the moving parts that have to behave
   3. Rider or driver lock: build, gear, helmet, closed with one consistency sentence
   4. Road and weather: the surface, what lines both sides of it, the light
   5. A ratio line stating how much of the film is camera motion and how much is scenery
   6. The numbered cut list, one line per second, each naming a camera position and what streaks past
   7. Tail: visual style, then a negative list of the ways a vehicle specifically breaks
4. Apply the guidance and check the pitfalls (next two sections) before returning anything.
5. Return the finished prompt as one copy-pasteable block, the anchor case link, and a three-line checklist of what to verify in the generated video.

Guidance (from the cases)

• Describe the vehicle by its parts, not by its badge. The Karakoram commercial names realistic suspension movement, wheel rotation, chain movement, engine vibration, then requires the proportions to hold for the whole clip.
• Declare the cut count before writing the list. The mountain-road motorcycle case opens with exactly 16 distinct cuts, total runtime ≈ 16–17 seconds and then runs CUT 01 through CUT 16, one second each.
• Put the speed in the camera position and in what gets thrown past it. The same case writes camera drops even lower, almost road-level and grass and fence posts racing past.
• State a ratio at the top and obey it below. The mountain-road case declares 90 % pure kinetic camera motion and 10 % environmental beauty, and not one of its sixteen cuts stops to admire the view.
• Write a negative list of vehicle-specific failures. Karakoram excludes no duplicated motorcycle components, no unrealistic wheel geometry, no floating motorcycle, and the motorcycle-to-dragon case adds that the two must clearly be the same entity.

Pitfalls

• Naming the vehicle and the road and leaving the rest to the model. Colour and stance then drift every cut; the mountain-road case writes Preserve the exact bike color, rider silhouette, road markings.
• Putting two camera positions inside one cut. A second only holds one position, and asking for more makes the model cut in the middle of the shot.
• Writing a transformation as an edit. The motorcycle-to-dragon case demands No cuts or jumps and spells the change out piece by piece: wheels become clawed limbs, frame expands into an armoured body.
• Holding a macro shot on the badge or the instrument cluster. Generated lettering comes out wrong; aim the close-ups at tyre contact, suspension compression and the exhaust instead.

Language

Reply in the user's language. The prompt itself can stay in English when that is what the user's Seedance workflow expects; ask if unclear. The Chinese version of this structure follows.

Chinese: Cars and Vehicles Speed Clips

The machine must be the same machine from start to finish; what delivers the output is the camera. First lock the car by components, then use a numbered shot list to fill the duration, changing camera position every second.

Structure:

1. Opening line: duration, aspect ratio, frame rate, and the total number of shots
2. Vehicle lock: model, color, and the parts that must move correctly
3. Rider or driver lock: build, gear, helmet, closing with a consistency line
4. Road surface and weather: pavement, what is on both sides of the road, lighting
5. A ratio line stating how much of the clip is camera movement and how much is scenery
6. Numbered shot list, one line per second, each naming the camera position and what is being thrown past it
7. Closing: visual style, plus a negative list specifically about how the car could break

Key points:

• Write the car by components, not by badge. The Karakoram ad names realistic suspension movement, wheel rotation, chain movement, engine vibration, then requires proportions to stay consistent throughout.
• Declare the total shot count first, then write the list. The mountain-road motorcycle one opens with exactly 16 distinct cuts, total runtime ≈ 16–17 seconds, followed by CUT 01 through CUT 16, one line per second.
• Write speed into the camera position and what is being thrown past it. In the same line, write camera drops even lower, almost road-level, then write grass and fence posts racing past。
• Give a ratio line at the start and execute accordingly below. The mountain-road one declares 90 % pure kinetic camera motion and 10 % environmental beauty, so none of the sixteen shots stops to look at the scenery.
• The negative list must name how the car could break. The Karakoram one excludes no duplicated motorcycle components, no unrealistic wheel geometry, no floating motorcycle, and the motorcycle-transforms-into-dragon one additionally requires that before and after the transformation it must be clearly the same entity.

Common pitfalls:

• Writing only the car name and the road, leaving the rest to the model. Color and pose will change from shot to shot; the mountain-road one specifically writes Preserve the exact bike color, rider silhouette, road markings。
• Writing two camera positions in one shot. Only one camera position fits in a second; writing more makes the model cut on its own mid-shot.
• Writing the transformation as a cut. The motorcycle-transforms-into-dragon one requires No cuts or jumps, and spells out piece by piece the wheels turning into clawed limbs and the frame stretching into an armored torso.
• Close-ups resting on the car badge or dashboard text. Generated text will always be crooked; change close-ups to structural parts like tire contact patches, suspension compression, and exhaust.

Copy-ready lead-in / Copy-ready lead-in

For users who would rather paste into a chat model than run this Skill, hand them this line above the structure:

I want a vehicle speed clip. [The vehicle is a white dual-sport motorcycle on a foggy gravel mountain road at dawn.] [The rider is a young man in a worn leather jacket, full-face helmet on the whole time.] Using the prompt template below, rewrite it into one ready-to-use Seedance video prompt for me:

I want to make a vehicle speed short film, [the vehicle is a white off-road motorcycle, the road condition is a gravel mountain road foggy in the early morning], [the rider is a guy in an old leather jacket, wearing a full-face helmet the whole time]. Please use the prompt template below to help me rewrite it into a Seedance video prompt that can be used directly:

Notes

• references/cases.md is generated from data/ in LearnPrompt/awesome-seedance and refreshed daily; do not hand-edit an installed copy.
• Prompts in the reference file belong to their creators (linked on every card). Transfer the structure; do not present a close copy as original work.
• For the full library across every model, install the goodcase Skill from LearnPrompt/goodcase-lite.