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.
TemplateTested 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.
Created with Made · Reviewed for rights, safety, and workflow accuracy.
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.
- 01Fixed subject
- The object, person, or scene that must remain recognizable across the set.
- 02Composition rule
- A camera family, surface, and framing boundary shared by every image.
- 03One variation axis
- The single property allowed to change, such as angle, light warmth, or distance.
- 04Constraints
- 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
- 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.
Continue from this workflow
Example
Coffee Detail Study with 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
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
An overhead commercial still life arranging leather accessories, sunglasses, watches, and denim on white marble with crisp natural detail.