---
title: "Controlled AI Image Variation Review Template | Made"
description: "Create a small visual series by holding the subject and composition steady, changing one lighting or framing variable, and reviewing the results side by side."
url: https://www.made.app/templates/controlled-image-variation-review
---

[Home](https://www.made.app/) / [Templates](https://www.made.app/templates) / Controlled AI Image Variation Review Template

Template Tested August 13, 2026

# Controlled AI Image Variation Review Template

Create a small visual series by holding the subject and composition steady, changing one lighting or framing variable, and reviewing the results side by side.

[Use this template](https://www.made.app/create?prompt=Close-up+product+study+with+crisp+texture+and+controlled+highlights.&model=flux-1.1-pro)

Created with Made · Reviewed for rights, safety, and workflow accuracy.

Product-detail study filled with a coffee subject and controlled highlights.

## Make the comparison readable

A variation set is useful only when the images share enough structure to compare. The template fixes the subject, surface, camera family, and constraint language before one selected variable changes.

## Record what moved

The coffee and portrait studies demonstrate two practical variation axes: a sequence of close product views and a sequence of natural-light treatments. Each set keeps its main subject recognizable across the comparison.

Required inputs

## Define one constant set and one change.

Made can prefill the tested Flux 1.1 Pro starting prompt. Replace the subject and comparison variable, then generate a small set before deciding which direction to refine.

**01 Fixed subject**

The object, person, or scene that must remain recognizable across the set.

**02 Composition rule**

A camera family, surface, and framing boundary shared by every image.

**03 One variation axis**

The single property allowed to change, such as angle, light warmth, or distance.

**04 Constraints**

Elements that must not appear or drift, such as logos, extra objects, or typography.

Tested workflow

## Use the exact starting point.

Prompt

Close-up product study with crisp texture and controlled highlights.

## Test record

**Model**

[Flux 1.1 Pro](https://www.made.app/models/flux-1.1-pro)

**Tool**

[AI Image Generator](https://www.made.app/tools/ai-image-generator)

**Settings**

aspectRatio: 16:9 · outputFormat: webp · variationCount: 4

**Tested**

August 13, 2026

Flux 1.1 Pro is the published Made model used for the recorded coffee-detail and natural-light comparison studies.

## Run the workflow

1. **Write the fixed visual facts**
   Describe the subject, environment, camera family, and constraints that every image must share.
2. **Choose one variation axis**
   Select only lighting, angle, crop, distance, or action stage for the first comparison.
3. **Generate a compact set**
   Create four results with the same model and output settings.
4. **Compare before refining**
   Review drift, detail, and the intended variable side by side, then refine one selected direction.

## Why it works

Fixing most visual facts turns a group of generations into a reviewable comparison instead of four unrelated concepts.

### Adjustment tips

- • Repeat the fixed subject and composition in every variation rather than relying on the previous image as implied context.
- • Judge the full set at the same display size before choosing a direction.

### Limitations

- • Text-to-image outputs can still change subject geometry, identity, or materials even when the prompt asks for consistency.
- • A montage documents the selected results, not every failed or discarded generation from an exploratory session.

Original Made media

## See what the instructions produced.

Portrait study filled with one adult subject and progressively warmer window light.

## Continue from this workflow

Example

### [Coffee Detail Study with Flux 1.1 Pro](https://www.made.app/examples/coffee-detail-study-flux-1-1-pro)

A four-image coffee series comparing the pour, surface detail, latte art, and finished cup under warm controlled highlights.

Example

### [Golden-Hour Window-Light Study with Flux 1.1 Pro](https://www.made.app/examples/golden-hour-window-light-study-flux-1-1-pro)

Four editorial portrait results compare neutral daylight with progressively warmer window light while keeping the room, clothing, and pose closely aligned.

Example

### [Editorial Accessories Flat Lay with Flux 1.1 Pro](https://www.made.app/examples/editorial-accessories-flat-lay-flux-1-1-pro)

An overhead commercial still life arranging leather accessories, sunglasses, watches, and denim on white marble with crisp natural detail.

[Use AI Image Generator](https://www.made.app/tools/ai-image-generator) [Explore Flux 1.1 Pro](https://www.made.app/models/flux-1.1-pro) [Rights or takedown request](mailto:rights@made.app?subject=Public%20content%20rights%20request)

---

Markdown version of https://www.made.app/templates/controlled-image-variation-review. Every public Made page has one at its URL plus .md (https://www.made.app/templates/controlled-image-variation-review.md). Site summary for AI assistants: https://www.made.app/llms.txt