In this guide, you will learn the complete process of creating this type of video using an AI-generated character image and a video generation tool. Instead of simply giving you a prompt, this tutorial explains what to do at every step, why each step matters, how to maintain facial consistency and how to fix common AI generation problems.
What You Will Create
The final video will be a 10-second vertical 9:16 cinematic sequence with an authentic 1980s Indian atmosphere. The story is divided into four small scenes so that the video feels more like a planned cinematic sequence instead of a random AI animation.
- 1–2 seconds: Riding a classic vintage motorcycle through an old Indian market.
- 2–5 seconds: Stopping at a roadside chai stall and drinking tea.
- 5–8 seconds: Listening to a portable radio while reading a vintage Hindi newspaper.
- 8–10 seconds: Riding toward a beautiful golden sunset.
The most important part of this workflow is that the same facial identity should remain consistent throughout the image and video.
What You Need Before Starting
Before generating anything, keep a clear reference photograph ready. A front-facing or naturally visible face generally works better because the AI needs enough facial information to understand the character's identity.
You will need:
- A clear reference photo.
- An AI image generator such as ChatGPT.
- An AI video generator such as Google Flow.
- A stable internet connection.
- A few minutes to review and regenerate the result if necessary.
Step 1 — Create the 1980s Character Image
The first step is to create a strong character image that will later be used as the visual reference for the video. This step is important because the video generator will use the generated image to understand the character, clothing, motorcycle and overall visual environment.
1. Upload Your Reference Photo
Open ChatGPT and upload a clear image of the person whose facial identity you want to use. The reference image should mainly help the AI understand the face. The original clothes and hairstyle do not need to be copied.
2. Copy the Image Prompt
Use the button below to copy the complete image prompt. Paste it into ChatGPT after uploading your reference image.
Image Prompt
Use the uploaded reference image only for the character’s facial identity. Preserve the exact face, facial structure, skin tone, eyes, eyebrows, nose, lips, jawline and distinctive facial features with maximum accuracy. Do not copy the reference image’s original outfit or hairstyle.Create an ultra-realistic cinematic portrait of the same character riding a classic vintage motorcycle through an authentic 1980s Indian market during golden hour. The character wears authentic 1980s Indian retro fashion, naturally styled for the era, with period-appropriate clothing and hairstyle. The character is seated naturally on a beautifully detailed classic vintage motorcycle, hands realistically positioned on the handlebars, confident but natural expression. The environment features a nostalgic Indian marketplace with old roadside shops, vintage signboards, bicycles, classic scooters, traditional architecture, glass orange beverage bottles, dusty roads and subtle background activity. Everything should feel genuinely from the 1980s, with no modern objects. Capture the scene from a dynamic low-angle cinematic perspective, showing the motorcycle prominently in the foreground while the character remains clearly visible. Warm golden sunlight streams through the market, creating natural lens flares, atmospheric haze, soft volumetric rays and beautiful rim lighting around the subject. Photorealistic 35mm film aesthetic, premium Indian commercial cinematography, realistic skin texture, natural facial details, shallow depth of field, cinematic HDR, warm amber and golden color grading, subtle film grain, soft halation, realistic shadows, natural motion-ready composition, highly detailed motorcycle chrome and textures. Composition: vertical 9:16, full-body character and motorcycle visible, cinematic depth, balanced framing, realistic perspective, premium movie still. Negative prompt: face alteration, identity change, face drift, modern clothing, modern motorcycle, smartphone, modern vehicles, futuristic objects, distorted hands, extra fingers, unrealistic anatomy, plastic skin, excessive beauty retouching, unwanted text, subtitles, logo, watermark.
3. Generate the Image
After pasting the prompt, generate the image. Give the AI some time to process the reference and prompt. When the result appears, do not immediately move to the video generation stage.
First check the face, hands, motorcycle, clothing and background. The character should look natural and the environment should feel like an old Indian market rather than a modern location with a vintage filter.
4. Check These Things Before Continuing
- Is the face similar to your reference?
- Is the character fully visible?
- Does the motorcycle look realistic?
- Does the clothing feel appropriate for the 1980s?
- Are there any smartphones or modern vehicles?
- Are the hands and fingers natural?
- Is the image vertical 9:16?
If the image has a major problem, regenerate it before moving to the video stage. A good starting image can make the next step much easier.
Step 2 — Convert the Image Into a Video
Once your character image is ready, the next step is to animate it. For this tutorial, the video prompt is designed as a complete 10-second sequence with four different moments.
The timeline is important because AI video generators sometimes ignore complicated instructions. Dividing the video into clearly defined time periods helps the generator understand what should happen during each part.
1. Open Your Video Generator
Open your preferred AI video generator. You can use Google Flow for this workflow.
Upload the 1980s motorcycle image that you created in Step 1. Then add the video prompt below.
🎬 80s Video Prompt
Step 3 — Understand the Video Timeline
The prompt is not simply asking the AI to make a random retro video. It gives the video a small storyline. Understanding this timeline will help you identify problems when reviewing the generated result.
1–2 Seconds: Motorcycle Riding
The video begins with movement. The character rides through a vintage Indian marketplace while the camera follows from a cinematic angle. This establishes the location and immediately creates visual energy.
2–5 Seconds: Drinking Chai
The motorcycle scene changes into a roadside chai stall. The close-up of the tea and natural steam helps make the scene feel more realistic. The rack focus then brings attention back to the character.
5–8 Seconds: Radio and Newspaper
This part establishes the everyday 1980s atmosphere. Instead of using modern technology, the character interacts with a portable radio and a vintage Hindi newspaper. These details are important because they communicate the time period without needing to add explanatory text.
8–10 Seconds: Sunset Ending
The final shot changes to a wider rear tracking view. The character rides away toward the sunset, creating a natural ending instead of stopping the video suddenly.
Step 4 — Generate and Review the Video
After adding the image and video prompt, generate the video. Once it is ready, watch the entire clip from beginning to end.
Do not judge the result only from the first few seconds. Check whether the face remains consistent after the camera angle changes, whether the clothing stays the same and whether the motorcycle and environment remain visually stable.
How to Get Better Face Consistency
Facial consistency is one of the most important parts of an AI character video. Since the video contains multiple camera angles, the AI may sometimes slightly change facial features.
To improve consistency, use one strong character image rather than changing the reference image during the process. Keep the identity instructions in the video prompt and avoid unnecessary changes to the character.
If the generated video changes the face too much, regenerate the video using the same character image and prompt. A stronger reference image can also help.
How to Make the 1980s Look More Authentic
A convincing retro video requires more than applying an old-film color filter. The objects, architecture, clothing and everyday activities should also support the period.
That is why this prompt specifically mentions vintage motorcycles, roadside chai stalls, portable radios, Hindi newspapers, bicycles, old storefronts, traditional architecture and period-appropriate clothing.
At the same time, modern objects should be avoided. A smartphone, LED screen, modern car or contemporary building can immediately break the retro illusion.
Common Problems and How to Fix Them
Problem 1 — The Face Changes
If the face changes between shots, use the same generated character image as the reference and regenerate the video. Avoid changing the character reference between attempts.
Problem 2 — Modern Objects Appear
Modern objects can sometimes appear automatically in an AI-generated environment. The negative instructions specifically tell the generator to avoid smartphones, modern vehicles, LED screens and futuristic objects. If they still appear, regenerate the scene.
Problem 3 — Hands Look Incorrect
Hands can be difficult for generative AI, especially when the character is holding a tea glass or motorcycle handlebars. If the hands look unnatural, regenerate the image or video rather than using a visibly distorted result.
Problem 4 — Clothing Changes
Keep the clothing description consistent. The video prompt specifically asks the AI to maintain the same face, hairstyle and outfit throughout the video.
Problem 5 — The AI Ignores the Timeline
If the generator does not follow the four scenes correctly, regenerate the video. Keeping each scene clearly separated by time ranges makes the instructions easier to interpret.
Tips for a More Cinematic Result
- Use a high-quality reference image.
- Keep the character clearly visible in the source image.
- Use a vertical 9:16 composition for Shorts and Reels.
- Do not add unnecessary modern objects to the prompt.
- Review the generated image before creating the video.
- Regenerate when there are major identity or anatomy problems.
- Use cinematic camera movements instead of asking for random movement.
- Keep the timeline simple and clearly divided.
Complete Workflow in Short
The complete process is simple once you understand the order. First, select a clear reference photo. Upload it to ChatGPT and use the image prompt to create the 1980s motorcycle scene. Check the generated image carefully. After that, open an AI video generator such as Flow, upload the generated image and paste the 10-second video prompt.
The video should then follow the planned sequence: motorcycle riding, drinking chai, listening to the radio while reading a newspaper, and finally riding toward the sunset.
This workflow gives you more control than simply asking an AI tool to “make an 80s video,” because the character, environment, camera movement, timeline and visual style are all described separately.
Frequently Asked Questions
Can I use any clear reference photo?
Yes. A clear image with the face visible is recommended. Avoid extremely blurry or heavily edited photographs because they may make identity consistency more difficult.
Can I use this prompt for YouTube Shorts?
Yes. The prompt is designed in vertical 9:16 format, which is suitable for platforms that use vertical short-form videos.
Why should I create an image before generating the video?
The generated image gives the video model a stronger visual reference for the character, clothing, motorcycle and environment. This can make the final video more consistent than starting from text alone.
Can I change the motorcycle or location?
Yes. You can modify the visual description while keeping the character identity instructions consistent. For example, you can change the market environment or choose another period-appropriate motorcycle.
Why does the result sometimes look too modern?
AI models may automatically introduce modern objects because they are common in their training data. That's why the prompt specifically describes 1980s objects and includes negative instructions for modern items.
Final Result
With this workflow, you can transform a simple reference photograph into a cinematic 1980s-inspired Indian story. The key is not just the prompt itself, but the complete process: create a strong character image first, review it carefully, use that image as the video reference, follow a clear timeline and regenerate whenever important visual details are incorrect.
The combination of vintage Indian locations, classic motorcycle shots, roadside chai, a portable radio, newspaper reading and a golden sunset creates a nostalgic visual story that works especially well in short-form vertical video.
If you are experimenting with AI video editing, this method also gives you a useful foundation for creating your own retro stories by changing the location, activity, camera movement and
