Verified against Midjourney · 2026-07-27
Hold one visual style across a whole campaign with --sref
A style-reference workflow using Midjourney's --sref and --sw parameters to keep color grading, lighting mood, and rendering texture consistent across a multi-asset social or ad campaign, so ten different scenes still read as one shoot.
The prompt
Ready to copy — highlighted parts are example details you can swap.
STYLE ANCHOR One Midjourney-generated (or carefully chosen existing) image whose color grade, lighting mood, and overall rendering texture is exactly what the whole campaign should look like — not necessarily featuring the same subject as any individual asset, since --sref transfers style, not content. Its URL goes in https://cdn.midjourney.com/def456-style-anchor.png. WHAT THIS SPECIFIC ASSET SHOWS a pair of hands tying the laces of a running shoe on a park bench at dawn WHY THIS ASSET NEEDS TO MATCH THE OTHERS one of twelve social ad assets for a running-shoe launch, all needing to read as one cohesive shoot when scrolled past quickly in a feed STYLE WEIGHT --sw 200 — this controls how strongly --sref's color, light, and texture override the model's own default aesthetic for this subject. Start around 100-250 for a firm but not overpowering match; push toward 500-1000 only if lower values are still letting Midjourney's default look bleed through on a subject very different from the one in the reference image. WHAT SHOULD NOT DRIFT BETWEEN ASSETS warm dawn color temperature, visible film grain, slightly desaturated shadows, soft contrast — no punchy commercial HDR look on any asset — name these explicitly here even though --sref is doing the heavy lifting, since calling out exactly what must not change (grain amount, color temperature, contrast level) gives you something concrete to check each new asset against rather than a vague "does it feel consistent" judgment call. NEGATIVE ELEMENTS --no text, logos, watermark, competitor branding PARAMETERS --sref https://cdn.midjourney.com/def456-style-anchor.png --sw 200 --ar 4:5 --v 7 OUTPUT a pair of hands tying the laces of a running shoe on a park bench at dawn --sref https://cdn.midjourney.com/def456-style-anchor.png --sw 200 --no text, logos, watermark, competitor branding --ar 4:5 --v 7 REUSE ACROSS THE FULL ASSET SET Every asset in this campaign should reuse the identical https://cdn.midjourney.com/def456-style-anchor.png and a consistent 200 — the subject line changes per asset, the style anchor does not. If an asset's subject is dramatically more or less detailed than the reference image's original subject (a tight product close-up versus a wide lifestyle scene), that asset alone may need a slightly higher --sw to keep the style from thinning out across a busier or emptier frame; note that adjustment explicitly if you make it, so the team building the next asset knows why one --sw value differs from the rest.
Customize
Optional — swap in your own details for the highlighted parts above.
Why this works
--sref conditions a generation on the color grade, lighting character, and rendering texture of a reference image while leaving the subject matter entirely to the text prompt — this is the mechanical reason it works for a campaign of visually different scenes: a close product shot and a wide lifestyle shot can share the exact same --sref value and come out feeling like they were shot by the same photographer on the same roll of film, even though the reference image itself might show neither scene. Without --sref, each asset's color grade and mood is instead determined independently by Midjourney's own default aesthetic reacting to that specific subject and text, which is why a campaign built from ten separately-worded prompts with no shared style anchor routinely comes back looking like ten different photographers' work, however carefully each individual prompt was written. --sw is the dial that solves the real failure mode of a style reference thinning out: a subject very different in composition or busyness from the reference image's original subject can pull the generation back toward Midjourney's default look even with --sref applied, especially at low --sw values, because the model is still balancing the reference style against its own strong aesthetic priors for that kind of scene. Raising --sw specifically for an outlier asset — a much busier or much emptier frame than the rest of the set — is the correct, targeted fix, rather than rewriting that one asset's text prompt to try to talk the model out of its own defaults. Naming the specific traits that must not drift — color temperature, grain, contrast — turns "does this feel consistent" from a subjective, easy-to-miss judgment call into a short checklist a reviewer can actually verify asset by asset. This matters in practice because style drift across a large asset set is rarely dramatic in any single image; it is usually a slow accumulation of slightly warmer color, slightly less grain, slightly punchier contrast from one asset to the next, each individually unnoticeable and collectively obvious once all twelve assets sit next to each other in a campaign deck. A named checklist catches that drift asset by asset, before the full set is assembled and the problem becomes expensive to fix.
What you get back
A set of campaign images with different subjects and compositions that all share the same warm dawn color temperature, grain level, and contrast curve from the style anchor — recognizable as one shoot when placed side by side in an ad set.
Verified against
Midjourney v7 · 2026-07-27
Changelog
- 2026-07-27 — Initial publish, verified against Midjourney v7 --sref/--sw for multi-asset campaign consistency.
Need this built into your business?
If a prompt isn't enough — what Scult builds, built and maintained for you — that's Scult's day job.
EXPLORE WHAT SCULT BUILDS
