Lock brand visuals with Midjourney Style References

The --sref parameter works by pointing Midjourney at a reference image (either a URL or a numeric style code), telling the model to borrow that image's aesthetic while generating something entirely new. You append --sref [image URL] --sw [weight] to your prompt, where --sw controls how strongly the reference influences the output (0-1000 range; default is 100). Here's the professional workflow: Create or screenshot a mood board of your brand visual direction—lighting style, color palette, texture, composition feel. Upload it to a private URL (most teams use Notion, Figma, or a private Slack shared drive). Then include that URL in every prompt with --sref. For example: "Corporate team meeting, diverse group, modern office, 35mm film grain, warm lighting --sref https://your-url.com/brand-mood --sw 250 --v 8.2". Every image you generate from that point keeps the same visual DNA—same color temperature, same texture quality, same overall feel. The weight parameter (--sw) is critical: use 100-200 for light influence (prompt leads, reference tints), 300-500 for medium influence (balanced), and 500+ when you want the reference to dominate the style. For B2B marketing where authenticity matters, add descriptors like "photorealistic, candid, slight film grain" to prevent over-processed, synthetic-looking imagery that alienates professional audiences. Common mistake: uploading a reference once and expecting consistency across dozens of generations. The technique only works if you include --sref in every single prompt. Version it into your standard prompt template so you never forget. This turns Midjourney from a one-off image generator into a brand asset factory where every visual stays on-brand.