Character Consistency in Midjourney: Fixed Parameters for Brand Mascots

Character consistency in Midjourney for brand mascots requires combining the character reference (--cref) parameter with consistent weight settings (--cw) and style reference (--sref) to lock visual traits across multiple generations. Core technique: Start by generating your ideal mascot, then use that image's URL with --cref [image URL] --cw 100 to maintain facial features, body proportions, and distinctive characteristics. The character weight parameter ranges from 0-100, where 100 provides maximum consistency for brand applications. Pair this with a style reference to lock artistic treatment and color palette. Practical workflow: Professional brand designers typically create a reference sheet with 3-5 approved mascot poses first, then use these as rotating character references for different marketing materials. This ensures your mascot maintains recognizable features whether appearing in social media posts, product packaging, or advertising campaigns. Important consideration: Character reference works best with human and humanoid characters. Abstract or heavily stylized mascots may require additional prompt engineering and multiple test generations to achieve commercial-grade consistency.

Essential parameter combination: Use --cref for character consistency, --sref for visual style, --cw for character weight control, and --sw for style weight adjustment. These four parameters work together to lock your mascot's appearance. Character parameters: The --cref [URL] parameter accepts your reference image URL, while --cw [0-100] controls how strictly Midjourney follows that reference. For brand mascots requiring high consistency, use --cw 85-100. Lower values (40-60) allow more variation while preserving core identity, useful for showing your mascot in diverse scenarios. Style locking parameters: Add --sref [URL] with an approved style image and --sw [0-1000] to maintain consistent artistic treatment. Brand mascots typically use --sw 500-700 to balance style adherence with prompt flexibility. Include --seed [number] to replicate specific successful generations exactly. Advanced consistency: Combine multiple character references using --cref [URL1] [URL2] with different weights to blend traits from approved mascot variations. This technique helps when your brand needs seasonal variants or different expressions while maintaining core visual identity.

Yes, this is precisely what locked parameters for brand mascots in Midjourney enable—maintaining facial features and color palette while varying poses, actions, and contexts through prompt adjustments. The technique: Lock your mascot's face with --cref [mascot URL] --cw 100, then describe different actions in your prompt: "jumping excitedly," "sitting at desk working," "waving hello." The character reference preserves facial structure, eye shape, and distinctive features while allowing the pose to change based on your text prompt. Color consistency strategy: Include specific color codes in both your prompt and style reference. For example, if your mascot has signature brand colors, add "--sref [color palette image] --sw 600" alongside your character reference. Describe colors explicitly: "wearing bright blue #0066FF shirt" rather than just "wearing blue shirt." Real-world application: Marketing teams successfully create entire mascot asset libraries this way—one reference image generates hundreds of contextual variations for different campaigns, seasons, and platforms. Each variation maintains brand recognition while adapting to specific communication needs. The character weight at 90-100 keeps the face nearly identical, while pose variations come entirely from prompt descriptions and camera angle specifications like "full body shot," "close-up portrait," or "action pose from side angle."

Professional workflow: Brands establish a mascot reference library containing 5-10 approved character images from different angles (front, side, three-quarter view), then systematically reference these with fixed parameters across all campaign materials. Campaign implementation: Creative teams create a standardized prompt template that includes their locked parameters: base prompt structure, --cref URL pointing to their reference library, --cw 95 for strict consistency, --sref for brand style guide, and --ar for platform-specific aspect ratios. This template ensures anyone on the team generates on-brand mascot content. Multi-platform consistency: The same parameter set generates mascot variants for different channels—square images for Instagram (--ar 1:1), vertical for Stories (--ar 9:16), horizontal for YouTube thumbnails (--ar 16:9)—while maintaining identical character appearance. This solves the traditional problem of mascot inconsistency across different designers and agencies. Quality control system: Leading brands maintain a parameter documentation sheet listing approved --cref URLs, required --cw values, brand-specific --sref combinations, and forbidden prompt terms that could break consistency. Tools like Aimensa can help teams organize and standardize these parameters across projects. Version control: When mascots need updates (seasonal outfits, product tie-ins), teams generate new reference images using the original parameters plus modifications, test consistency across 20-30 variations, then update their reference library only after approval.

Character weight scale explained: The --cw parameter ranges from 0 (minimal reference influence) to 100 (maximum consistency), with different values serving specific brand needs. Maximum consistency (--cw 90-100): Use for primary brand mascot applications where facial features, proportions, and recognizability must remain identical—product packaging, logo variations, official merchandise, and hero campaign images. At --cw 100, Midjourney prioritizes matching your reference image's character traits above nearly all other prompt instructions. Balanced approach (--cw 60-80): Ideal for content marketing, social media posts, and educational materials where you want clear mascot recognition but need flexibility for different scenarios, lighting conditions, or interactive contexts. This range maintains distinctive features (eye shape, face structure, signature accessories) while allowing natural environmental integration. Creative flexibility (--cw 30-50): Useful for concept exploration, seasonal variations, or when your mascot needs to adapt significantly to different art styles or thematic campaigns. The character remains recognizable but permits substantial artistic interpretation. Testing recommendation: Generate the same prompt with --cw values at 50, 70, 90, and 100 to establish your brand's consistency threshold. Most commercial applications find --cw 85-95 provides the optimal balance—strict enough for brand guidelines yet flexible enough for diverse creative execution.

Yes, multi-reference combination significantly improves mascot stability by teaching Midjourney your character's appearance from multiple perspectives and contexts simultaneously. Syntax for multiple references: Use --cref [URL1] [URL2] [URL3] to include 2-4 reference images of your mascot. Midjourney analyzes all references to extract consistent character features while averaging out individual image anomalies. This technique works exceptionally well when your references show the mascot from different angles or in different lighting. Weighted multi-reference: Control individual reference influence using --cref [URL1]::2 [URL2]::1 where the numbers represent relative weight. If one reference image perfectly captures your mascot's face while another shows ideal body proportions, weight the facial reference higher (::3 or ::4) to prioritize those features. Strategic reference selection: Choose references that together cover front view, side profile, and three-quarter angle. Include one reference showing your mascot's signature expression or gesture. This comprehensive reference set creates a "3D understanding" that improves consistency across novel poses and scenarios. Practical limitation: Using more than 4-5 references can dilute consistency rather than improve it, as Midjourney may struggle to identify the most important character traits. Start with 2-3 carefully selected references showing your mascot clearly in different contexts, then add more only if consistency improves through testing.

Conflicting prompt instructions: The most common error is writing prompts that contradict your reference image's established traits. If your mascot has blue fur but you prompt "red fur," even --cw 100 creates confusion. Always align prompt descriptions with reference characteristics. Low-quality reference images: Using blurry, poorly lit, or low-resolution references dramatically reduces consistency. Your character reference should be clear, well-lit, and at least 1024px on the longest side. Midjourney cannot maintain consistency for details it cannot clearly see in the reference. Excessive style mixing: Applying incompatible style references alongside character references causes instability. If your mascot is cartoon-style but you add a photorealistic --sref, the system struggles to reconcile both instructions. Keep style references compatible with your mascot's established artistic treatment. Neglecting aspect ratio effects: Extreme aspect ratios (--ar 21:9 or --ar 4:5) can distort character proportions even with high --cw values. Test your mascot across all needed aspect ratios and adjust character weight accordingly—sometimes --cw 100 works for square images but --cw 85 provides better results for ultrawide formats. Ignoring version differences: Different Midjourney model versions (v6, niji, etc.) interpret character references differently. Document which model version your reference library uses and maintain consistency. Switching from standard Midjourney to Niji mode mid-campaign can significantly alter your mascot's appearance despite identical parameters.

Reference sheet essentials: A professional mascot reference sheet includes your character from multiple angles (front, side, back, three-quarter views), showing signature expressions, color specifications, and proportion guidelines—all generated with identical base parameters. Creation process: Start with your best mascot generation, then create systematic variations using that image as --cref with --cw 100. Generate "mascot facing forward, neutral expression, white background," then "mascot facing left profile," then "mascot three-quarter view." Use consistent lighting and simple backgrounds to isolate character features clearly. Documentation to include: Alongside your reference images, record the exact parameters used (--cref URLs, --cw value, --sref if used, --seed numbers for perfect reproductions), approved prompt language, hex codes for brand colors, and any crucial prompt terms that ensure consistency. This documentation becomes your "mascot brand guidelines." Testing your reference sheet: Validate effectiveness by having team members generate mascot content using only your reference sheet and documentation. If results vary significantly between users, your references need refinement or your parameters require adjustment. Consistency across different users proves your reference system works. Maintenance schedule: Review and update your reference sheet quarterly or when launching major campaigns. As Midjourney updates its models, occasionally regenerate your reference images with current versions to ensure ongoing compatibility and consistency.