How to Build an AI Node Canvas Workflow for Images and Videos
Producing a series of product images and videos from scratch can mean repeating the same prompts, edits, and exports. A node canvas connects those steps in one graph, so you can reuse a starting idea and develop multiple creative variations.
Plan the creative series
Choose one product, one audience, and one campaign objective before opening the canvas. Define the outputs you need, such as three product images and two short video concepts. Pick an aspect ratio for each placement: 1:1 for square posts, 4:5 for portrait feeds, 9:16 for vertical video, or 16:9 for widescreen video.
Collect a product reference you have permission to use. Note the details that must remain consistent, including its shape, color, label, and packaging. If accurate product text matters, plan to add or correct it during final editing rather than relying on generated lettering.
Build the node workflow
- Start with the brief. Open Oranzy Canvas and create a text prompt describing the product, audience, setting, lighting, and desired aspect ratio. If you are new to Oranzy, sign up for 100 bonus tokens before running the workflow.
- Generate and select images. Connect the prompt to an image generator and provide your product reference where appropriate. Run a small test, then inspect the results for product shape, colors, and composition. Select a strong image before creating further variations.
- Prepare image branches. Connect the selected image to background removal if you need a clean cutout, and to upscaling when you need more detail for export. Create separate prompt branches for alternate scenes or aspect ratios. Check the cutout edges and upscaled image before using them downstream.
- Create video variations. Connect a selected image to a video model and describe the motion: for example, a slow camera push-in while the product stays still. Generate a short test, check for distortion, and repeat with a different motion or scene for the second concept. Export the approved images and clips for final editing.
Use prompts that are easy to vary
Keep the product description fixed and change one creative variable at a time. This makes it easier to see whether a new background, lighting setup, or camera movement improves the result.
Image prompt: “Create a 4:5 product image of a matte blue water bottle on a pale stone surface. Soft side lighting, minimal background, bottle centered with clear space above for a headline. Preserve the reference bottle's shape and color.”
Video prompt: “Animate the selected product image with a slow camera push-in. Keep the bottle stationary and its shape consistent. Soft highlights move subtly across the surface. Vertical 9:16 composition.”
For a second image, change the setting to a gym bench while keeping the product description. For a second video, test a gentle camera orbit instead of a push-in. Review each variation before generating the rest of the series.
Avoid common mistakes and wasted generations
- Changing too much at once: Keep product details and framing stable while testing one new scene or lighting treatment.
- Using the wrong source ratio: Create or select an image composed for the intended video ratio. Cropping a square image into 9:16 can cut off the product or leave too little room for text.
- Animating an inaccurate image: Fix obvious shape, color, or label problems before sending an image to video. Motion can make defects more noticeable.
- Upscaling every draft: Select the strongest candidates first, then upscale only the images needed for delivery.
- Overloading the motion prompt: Start with one camera movement and restrained subject motion. Add complexity only after the test clip looks stable.
Review and export the series
Check every asset at its intended display size. Confirm that the product is recognizable, cutout edges are clean, and important details are not cropped. Watch each clip from start to finish for flicker or changes in product shape.
Export approved assets in the ratios needed for their placements. Add final headlines, logos, subtitles, and calls to action in your editing workflow, then check the finished files on both desktop and mobile.
FAQ
What does an AI node canvas workflow do?
It connects prompts, image generation, background removal, upscaling, and video generation in a graph. You can use selected outputs as inputs for later steps and create variations from the same starting idea.
Which aspect ratio should I use for product creatives?
Use 1:1 for square posts, 4:5 for portrait feed images, 9:16 for vertical video, and 16:9 for widescreen video. Check the requirements of the placement before generating.
How can I reduce wasted generations?
Test a small number of images first, select an accurate product image, and change one variable at a time. Upscale and animate only the candidates that pass your visual checks.
Can I turn one product image into multiple videos?
Yes. Use the selected image as the starting point for separate video generations, changing the motion prompt for each concept. Review each clip for product distortion and flicker.