CandyWrite
HomeBlogs
CandyWrite

An independent publishing platform for essays on technology, design, and creative work. Free to read, free to write.

Explore

  • Home
  • All Blogs
  • Topics
  • Tags
  • Writers
  • Most Read

About

  • About Us
  • Help & Support
  • Contact

Contact

  • hello@candywrite.com
  • support@candywrite.com

Get Updates

© 2026 CandyWrite. All rights reserved.

AboutHelpContactPrivacy PolicyTerms of Service
  1. Home
  2. Blogs
  3. Design
  4. A practical shot list for AI product visuals
Design

A practical shot list for AI product visuals

A product image has a job: help someone understand an object and imagine using it. A short shot list makes AI-assisted visuals easier to review.

L

Written by

liangqi zhao
  • 31 October 2026
  • 2 min read
  • 0 views
0

A product image has a job: help someone understand an object and imagine using it. A short shot list makes AI-assisted visuals easier to review.

1. Start with the decision

Choose one purpose per image: show the whole product, clarify a detail, or place it in a believable setting. Note what the viewer must see at thumbnail size.

2. Lock the product facts

Write down the shape, materials, color, labels, and number of parts from an approved reference. Treat these as constraints rather than details to improvise. Use only reference images you have permission to use.

3. Make a small shot list

Plan a clean front view, a three-quarter view, a detail crop, and a contextual scene. Specify the surface, light direction, camera angle, and negative space for each. Ask for one change at a time so differences remain useful.

4. Review before adding copy

Compare edges, proportions, labels, reflections, and packaging against the source. Check whether a hand or prop obscures a key feature. Regenerate or edit inconsistencies, then add factual copy in a separate design step.

5. Save what worked

Keep the approved reference, prompt, selected frame, and review notes together. A simple record helps the next campaign stay visually consistent without pretending generated details are product facts.

This workflow was prepared by the Jynno team. Jynno is an AI creative studio for image and video workflows: https://jynno.com/

Enjoyed this article?

Show some love or pass it on to someone who'd like it.

About the author

L

liangqi zhao

@zhaoliangqi117

View profile

Author and writer at CandyWrite. Sharing knowledge, tutorials, and reflections on technology, design, and ideas.

Join the conversation

Discussion0

Live updates

Join the conversation. Sign in to leave a response or reply to comments.

Sign InCreate Account
No responses yet. Be the first to share your thoughts!

Keep reading

Related articles

More essays on the same topics, picked for you.

View all blogs
Design

13 Aug 2026

Calm Interfaces: Designing for Attention You Did Not Earn

The dominant design conversation has shifted from engagement to restraint. Reducing cognitive load is now the differentiator, and it is harder than adding features.

MMuhammad Umer
Design

12 Aug 2026

Design Systems Became Governance. Act Accordingly.

A component library is a convenience. A design system that enforces rules across human and machine-generated interfaces is infrastructure, and it needs owners with authority.

MMuhammad Umer
Design

10 Aug 2026

Accessibility Is a Rendering Requirement, Not a Review Step

Audits find issues after they ship. Building accessibility into the primitives means most of those issues never get written, which is the only approach that scales.

MMuhammad Umer
Design

9 Aug 2026

Dark Mode Is a Colour System Problem, Not a Theme Toggle

Inverting a palette produces muddy shadows, glaring surfaces, and unreadable states. Doing it properly means designing roles rather than values.

MMuhammad Umer

Enjoyed this perspective?

Join 12,000+ readers getting our Saturday morning editorial dispatch with our top essays and reading recommendations.