Source: https://restorebunny.com/guides/when-not-to-generate-product-evidence/

# When to Keep Product Evidence Photos Unretouched

Choose real captures for condition, labels, serial details and construction instead of relying on a generated scene.

Product photography2 min readUpdated 18 October 2026

## A setting for your product

Create a scene from a real reference, compare the details and export a listing kit.

[Open Product Photo Studio](<https://restorebunny.com/signup/?from=seo-content&source=%2Fguides%2Fwhen-not-to-generate-product-evidence%2F&view=product>)1 credit per scene · Account required[See credit packs](<https://restorebunny.com/pricing/>)

Quick answer

## The short version

Use real photographs when the image is evidence of the exact item. Condition, serial markings, package text and construction details should not depend on a generative model’s interpretation. A styled scene may serve a separate presentation purpose, but it should not replace the photographs a buyer needs to inspect.

## A practical workflow

Mark evidence views in the shot list before editing starts. Keep original files and make only appropriate delivery copies. Review platform rules for used or damaged goods and avoid transformations that conceal wear. If an image is too unclear to show a meaningful detail, improve the capture instead of asking AI to fill it in. Separate factual inspection views from decorative settings in both naming and gallery order.

## Worked planning example

Planning example: A used camera has a scratch near its hot shoe. A scene edit removes the mark. The seller rejects it as condition evidence and retains the real close-up. The overall gallery can still use a suitable presentation image if it accurately represents the item and follows the destination’s rules.

## Check before you finish

1. Identify evidence views before editing.
2. Preserve real condition and identifying details.
3. Reshoot unclear facts instead of reconstructing them.
4. Check the marketplace’s current picture rules.

## Choose the next step for your task

Product Photo Studio creates a new setting from one product reference for one credit per generation. Review shape, labels, color, quantity and condition before commercial use. The listing kit adds alternate image shapes and a review manifest without an extra export charge. It does not certify marketplace acceptance. Use a real capture when exact evidence of the item is required.

## FAQs

What should I check before finishing this task?

Identify evidence views before editing. Preserve real condition and identifying details. Reshoot unclear facts instead of reconstructing them. Check the marketplace’s current picture rules.

Are these examples documented customer results?

No. They are hypothetical planning examples and editorial decision frameworks. They do not establish a sales increase, a successful restoration or a printer’s approval.

## Sources

Sources for technical facts and documented capabilities. Worked scenarios are hypothetical planning examples.

- eBay: [Picture policy](<https://www.ebay.com/help/policies/listing-policies/picture-policy?id=4370>)

## Related

- [Open the related free tool](<https://restorebunny.com/tools/product-photography-shot-list/>)
- [Explore Product Photo Studio](<https://restorebunny.com/features/product-photo-studio/>)
- [Correct a Product Reference’s Color Cast Carefully](<https://restorebunny.com/guides/correct-a-product-reference-color-cast/>)
- [Keep Product Scenes Consistent Without Extra Props](<https://restorebunny.com/guides/keep-product-scenes-uncluttered/>)

## Ready to prepare your product scene?

Create a reviewed product setting and export a listing kit. One credit per scene.

[Open Product Photo Studio](<https://restorebunny.com/signup/?from=seo-content&source=%2Fguides%2Fwhen-not-to-generate-product-evidence%2F&view=product>)[View Pricing](<https://restorebunny.com/pricing/>)

RestoreBunny

Photo workflow guides and capability comparisons

Guides follow our [editorial policy](<https://restorebunny.com/editorial-policy/>) for sourcing, product claims, AI limitations, updates, and corrections.
