Skip to content
As Scenes sees it

AI Images: Which Method Should You Choose?

A practical guide to product accuracy, creative flexibility and total cost in AI-created environments.

Michael Rying · 23 Sep 2026 · 6 min read · Creative production

Two people look up at a simple green chair standing on a mountain peak in a stylised landscape.

Scenes was founded to place products in worlds that were previously impossible to shoot.

1. Existing packshots and photos

We place your existing product photos, at packshot quality, into AI environments. Our post-production team adjusts perspective, ambient light, contact shadows and reflections so the product sits naturally in the scene.

  • The advantage: Makes more of your existing image library without booking a new shoot. It is an efficient route to seasonal updates, new channel assets and regional variations.
  • The trade-off: The original camera angle is fixed. Strong reflections, limited resolution or awkward edges can restrict which scenes will look convincing.
  • Best for: Turning good packshots into a broader collection of lifestyle images. This is often used for social media, where speed and smaller campaigns matter.

2. In-house studio photography

When a scene needs a specific angle or quality of light, we photograph the physical product in our Copenhagen studio for that particular setting. Camera position, lens, lighting and styling are planned before the product is integrated into the AI-generated environment.

  • The advantage: This is the most direct way to capture true colour, texture, construction and fine material detail. It is typically used for products where details must be completely accurate, often accompanied by close-ups. The product owner is frequently involved to ensure every important detail is considered.
  • The trade-off: It requires physical samples, upfront planning and a shoot with the product owner present. Once the shoot is over, new viewpoints require additional photography.
  • Best for: Hero campaign images, luxury products and product details that must withstand close inspection.

3. 3D model

With a usable 3D model, we can render the product from any required viewpoint and match its perspective and lighting to the environment. This allows us to create new views without another physical shoot.

  • The advantage: This offers the greatest freedom to vary camera angles and reuse the product across many settings. Because the product owner and our 3D expert have already finalised every detail, future images often do not need the product details approved again. It is especially useful for very large, technical or unreleased products, as well as campaigns requiring many variations.
  • The trade-off: Requires existing 3D files in the right quality and format. If these are not available, we can create a 3D model from scratch; it must be created and approved by your product owner before image production begins.
  • Best for: Complex product ranges and campaigns that need multiple angles, markets or formats.

4. Fully AI-generated product and scene

Reference content - such as sketches, blueprints, photos or screenshots - guides an AI model to create the product and its surroundings together. For some projects, we also train a product-specific model.

  • The advantage: Rapid creative iteration. Teams can test the product, setting, lighting and mood together while the campaign direction is still open.
  • The trade-off: An AI-generated product may look convincing while changing details a customer expects to be accurate. Logos, stitching, hardware, proportions and colour can drift between images, creating more review and correction work.
  • Best for: Concepts, pitch materials and campaigns where the product's details can be checked and corrected against reliable references.

Essential in every method: art direction and post-production

Our experienced art director can turn a brief or hand-drawn sketch into the right scene for your brand. Composition, lighting, atmosphere and the role of the product are all considered before we choose how to produce the image.

At Scenes, an image never goes straight from a camera, 3D render or AI model to final delivery. Every finished asset passes through professional post-production.

This stage is built into both the schedule and budget. We start by analysing your brand guidelines, visual identity and best-practice examples. Our expertise across all four production methods allows us to choose the best approach for every brief and change course whenever the image demands it.

Methods comparison

Five dots indicate a strong advantage; one dot indicates a limited advantage. These are practical guides rather than measured scores. Camera flexibility means the ability to change the viewpoint after the source asset has been created.

Production methodProduct detailCamera flexibilityLighting matchSpeed to first visualReuse across scenes
Existing packshots and photos●●●●●●○○○○●●●○○●●●●●●●●●○
In-house studio photography●●●●●●●○○○●●●●○●●○○○●●●○○
3D model●●●●○●●●●●●●●●○●●●○○●●●●●
Fully AI-generated●●○○○●●●●○●●●●○●●●●○●●●●○

The 3D speed rating assumes a usable model exists. Building one from scratch adds time. Speed reflects an initial usable visual; approval and finishing can take longer, especially for a fully AI-generated product.

Production complexity and total cost

The fastest way to test an idea is not necessarily the least expensive way to deliver a finished, approved image. Total cost depends on the required volume and formats, and typically includes art direction, asset preparation, photography or 3D work, scene creation, professional post-production, revisions and final delivery. We can deliver files scaled for everything from social media posts to large, print-ready wall posters.

Production methodComplexityTotal cost profileMain cost drivers
Existing packshots and photosLow to mediumUsually lower when source images are strongImage quality, masking, reflections and correction work.
In-house studio photographyMedium to highHigher upfront; more efficient when one shoot supplies several assetsSamples, studio time, specialist crew and retouching.
3D modelMedium with an existing model; high if one must be builtInitial model cost can be high; reuse lowers the cost per additional imageGeometry, materials, rendering and the number of versions.
Fully AI-generatedLow for concepts; variable for final productionLow to explore; difficult to predict for an accurate final imageProduct drift, revisions, manual corrections or a change of method.

These are relative comparisons, not price quotes. Product complexity, source quality, accuracy requirements and image volume all affect the final estimate.

Choosing the right method

Start with the product details that must be exact, then consider the required output formats, the number of markets and channels, and the total volume of images needed.

  • When exact product detail matters most: Use Scenes studio photography or a carefully built 3D model.
  • When you have strong assets and need new contexts quickly: Start with existing packshots and photos.
  • When you are still finding the creative direction: Use fully AI-generated images to explore ideas before committing to final production.

A campaign can easily combine methods. For instance, a hero image may call for a dedicated shoot, while a 3D model supplies additional angles and existing packshots support regional or seasonal variations.

The environment can shift almost anywhere - the right production method ensures the product remains unmistakably yours.

As Scenes sees it.

By Michael Rying

Talks, interviews and articles · Discuss what this means for your brand