Course Objective & Scope
The real challenge in image generation is not getting one good image but getting the result an art director wants consistently and repeatably. This course systematically covers how to steer a model with reference images where a text prompt is not enough.
Participants apply Midjourney's --sref, --oref and weight parameters, structure and style references in Adobe Firefly, edit-oriented direction with GPT image and Gemini, and ControlNet and IP-Adapter workflows with open models. Each participant builds a documented prompt and reference library for their own projects.
Who Should Attend?
- Designers with basic experience in image generation tools
- Art directors and creative directors
- Illustrators and concept artists
- Prerequisite: basic experience with Midjourney, Firefly or a similar tool
Key Takeaways
- Write prompts composed of subject, style, composition, lighting and lens components
- Produce consistent image series with style and character references
- Control pose and composition with image-to-image and ControlNet
- Fine-tune results by adjusting reference weights
- Document a prompt and reference library the team can share
Curriculum
Anatomy of an Image Prompt
- Subject, setting, style, lighting, lens and framing components
- Negative guidance and parameters
Style Control with Style References
- Midjourney
--srefand style weights - Using Firefly style reference
Character and Object Consistency
- Keeping the same character across scenes with omni/character references
- Object consistency in product visuals
Composition and Pose Control
- Image-to-image, structure reference and sketch-based generation
- Open-model workflows with ControlNet and IP-Adapter
Prompt and Reference Library
- Documenting successful combinations
- Sharing and versioning within the team
Other Courses in This Category
-
Producing campaign, social media and web visuals with Midjourney, Adobe Firefly, GPT image and Gemini image models, covering formats, series consistency, basic editing and upscaling. Teams deliver a publish-ready visual content set.
- Live Course
- 5 modules
-
Hands-on training in producing A/B test, localisation and format variants from existing images using Photoshop Generative Fill, inpainting/outpainting, background replacement, prompt-based editing (GPT image, Gemini) and upscaling.
- Live Course
- 5 modules
-
Training on translating brand guidelines into Midjourney moodboard/style references, Adobe Firefly Custom Models and LoRA workflows. Design teams build a repeatable visual system that preserves brand colour, lighting and aesthetic rules.
- Live Course
- 5 modules
-
Legible typography, grid and negative-space composition, and fine style control in generative image models (Ideogram, Flux, GPT image, Midjourney, ControlNet). Designers produce key visuals, posters and packaging with full compositional control.
- Live Course
- 4 modules
-
Product design training spanning user research synthesis, information architecture, screen design with Figma AI/Figma Make and v0, and usability review. Product designers deliver design-system-compliant screen flows and a UX artefact pack.
- Live Course
- 5 modules
-
Moving from concept to interactive, data-driven prototypes with Figma Make, v0, Lovable and Bolt. Product and design teams build multiple design variants in hours for stakeholder alignment and usability testing.
- Live Course
- 5 modules
Get in touch
