Five input references and sources
woman · man · cafe · camera · dog20/20 clips collected · 10/10 paired comparisons ready. Five references · long prompts · 10 seconds · 1344×768 · Hyperflow full, 8 steps · Euler · fixed pair seeds.
These ten pairs show no consistent visual advantage from the extra retention block or the appropriate mode markers. Explicit wardrobe, color, daylight and environment changes happen under both marker choices; selective lighting fails in both. One notable exception favors A: the environment test has one dog in A and two in B. This supports the practical importance of prose, not a claim that markers are ignored. Results are single-seed observations with this Hyperflow setup.
Study 1: fully_preserved retention block versus omitting that entire block. Study 2: appropriate preservation markers versus fully_preserved, with identical instructions.
One seed per scenario, five scenarios per claim; one shared five-image reference set; exploratory rather than a population-level conclusion. Marker-B deliberately conflicts with explicit requested changes. No deterministic repeat control was added. Different outputs alone do not establish a quality advantage.
Native 243-frame (10.125s) renders are preserved in each originals folder. Viewing copies are trimmed to 240 frames (10s), with no speed change. Side-by-side downloads use A audio; the paired players let you switch.
woman · man · cafe · camera · dogWritten findings · Method and limitations · Experiment metadata
The two referenced adults exchange the referenced camera inside the referenced cafe while the puppy watches; identity, wardrobe, environment, and prop appearance remain unchanged.
Native originals (243 frames): A · B
Download side-by-side (audio A) · Full contact sheet
Denser timeline: A at 2 fps · B at 2 fps
Exact prompt A · Exact prompt B
Both versions preserve the recognizable people, dark clothing, café fixtures, golden dog and black camera while completing the handoff and final placement. Framing and hand positions differ, but neither shows a clear reference-retention advantage in the inspected frames. Both deviate from the requested standing adults / puppy-on-floor arrangement.
A: fully_preserved retention block; B: identical prompt with entire block omitted
The woman photographs the referenced puppy as the man quietly attracts its attention, using the same referenced identities, clothes, camera, and cafe.
Native originals (243 frames): A · B
Download side-by-side (audio A) · Full contact sheet
Denser timeline: A at 2 fps · B at 2 fps
Exact prompt A · Exact prompt B
Both versions show the woman crouching, raising the camera to photograph the dog, then lowering it to inspect the display. Clothing and café remain similar. Both put the dog on the tabletop instead of its requested floor mat; retaining the block does not fix that instruction failure.
A: fully_preserved retention block; B: identical prompt with entire block omitted
A slow camera orbit reveals the referenced people, puppy, cafe and handheld camera from changing viewpoints without changing their identities or wardrobe.
Native originals (243 frames): A · B
Download side-by-side (audio A) · Full contact sheet
Denser timeline: A at 2 fps · B at 2 fps
Exact prompt A · Exact prompt B
Both versions change viewpoint substantially while retaining the people, clothing, café and camera. The fully_preserved block does not freeze the shot. Camera handling, dog position and foreground occlusions differ, but there is no clear overall retention winner from the inspected frames.
A: fully_preserved retention block; B: identical prompt with entire block omitted
The man briefly crosses in front of the woman and camera, then reveals them unchanged in the referenced cafe while the puppy follows a small movement.
Native originals (243 frames): A · B
Download side-by-side (audio A) · Full contact sheet
Denser timeline: A at 2 fps · B at 2 fps
Exact prompt A · Exact prompt B
Both show the man moving from the foreground toward the woman, followed by a shared look at the camera. Their appearances remain coherent in the sampled sequence. The crossing is not a clean, prolonged face occlusion, so this pair provides only weak evidence about recovery from occlusion.
A: fully_preserved retention block; B: identical prompt with entire block omitted
The referenced adults exchange two short lines about taking a picture while the puppy settles beside their table in the same cafe.
Native originals (243 frames): A · B
Download side-by-side (audio A) · Full contact sheet
Denser timeline: A at 2 fps · B at 2 fps
Exact prompt A · Exact prompt B
Both retain the referenced people and café through a very similar conversational composition. The woman raises the camera near the end in both. Both keep the dog on furniture instead of the floor mat. Spoken wording, voice quality and lip sync have not been evaluated; both audio tracks remain available for listening.
A: fully_preserved retention block; B: identical prompt with entire block omitted
The same woman and man meet in the referenced cafe with explicitly changed clothing while retaining their faces, hair and the referenced camera and puppy.
Native originals (243 frames): A · B
Download side-by-side (audio A) · Full contact sheet
Denser timeline: A at 2 fps · B at 2 fps
Exact prompt A · Exact prompt B
A (partially_preserved) and B (fully_preserved) both give the woman the yellow rain jacket and the man the teal pullover, with recognizable faces and hair. Fully_preserved does not prevent the explicit wardrobe replacement in this pair. There are staging differences, but no clear benefit from the appropriate marker on the tested clothing task.
A: appropriate partially_preserved/attribute_transfer/weak_reference markers; B: fully_preserved substituted for those markers only; all explanatory prose identical
The adults inspect a red-bodied version of the referenced camera; its shape and black lens stay the same while its casing color changes.
Native originals (243 frames): A · B
Download side-by-side (audio A) · Full contact sheet
Denser timeline: A at 2 fps · B at 2 fps
Exact prompt A · Exact prompt B
Both variants produce a red camera body with a black lens and carry that color through the handoff to the closing view. The fully_preserved marker does not block the explicitly requested recoloring. The overall camera shape is retained, but this is not a test of exact product geometry or branding.
A: appropriate partially_preserved/attribute_transfer/weak_reference markers; B: fully_preserved substituted for those markers only; all explanatory prose identical
The reference cafe is relit as a bright morning interior while its recognizable architecture, the people, the camera and the puppy remain consistent.
Native originals (243 frames): A · B
Download side-by-side (audio A) · Full contact sheet
Denser timeline: A at 2 fps · B at 2 fps
Exact prompt A · Exact prompt B
Both versions replace the moody café exposure with bright daylight while retaining its recognizable architecture and fixtures; the CAFE sign appears unlit in both. Both also generate two cameras despite the instruction to keep one. The appropriate retention marker does not resolve that duplication.
A: appropriate partially_preserved/attribute_transfer/weak_reference markers; B: fully_preserved substituted for those markers only; all explanatory prose identical
The blue side-lighting from the woman's reference portrait is transferred to the man's face while both people keep their individual identities in the cafe.
Native originals (243 frames): A · B
Download side-by-side (audio A) · Full contact sheet
Denser timeline: A at 2 fps · B at 2 fps
Exact prompt A · Exact prompt B
Both versions put blue light on the man, but both also leave strong blue light on the woman, contrary to the explicit request to keep her warm-neutral. Attribute_transfer does not provide a clear advantage over fully_preserved for this selective lighting assignment in the inspected frames.
A: appropriate partially_preserved/attribute_transfer/weak_reference markers; B: fully_preserved substituted for those markers only; all explanatory prose identical
The cafe image supplies only a warm welcoming atmosphere for a newly designed bright botanical lounge, while the referenced people, camera and puppy remain recognizable.
Native originals (243 frames): A · B
Download side-by-side (audio A) · Full contact sheet
Denser timeline: A at 2 fps · B at 2 fps
Exact prompt A · Exact prompt B
Both A (weak_reference) and B (fully_preserved) replace the café fixtures with the explicitly requested white arches, pale stone and green plants. The mode marker does not block that substantial environment change. A has one dog, whereas B generates two dogs throughout the sampled sequence. A is preferable for the requested subject count in this pair; without repeat controls or more seeds, that does not establish a general causal benefit of weak_reference. Both put the dog(s) on furniture rather than the requested floor mat.
A: appropriate partially_preserved/attribute_transfer/weak_reference markers; B: fully_preserved substituted for those markers only; all explanatory prose identical
