9.1 Image Prompt + Image Weight

Midjourney Image Manipulation Techniques

  • Introduction to Outputs with Images

    • Explore how to influence Midjourney outputs using images.

    • Key methods: Image prompting, Image weight, and Seed.

  • Image Prompting

    • Definition: Combining an image with a text prompt to generate an artistic output.

    • It sets the baseline for the influences approaching the final image.

  • Image Weight

    • Description: A parameter that adjusts the focus between text and the image in the output.

    • Importance: It controls how much influence the source image has compared to the text prompt.

    • Parameter Values:

    • Ranges from 0 to 3 (default is 1).

    • Lower values emphasize the text; higher values emphasize the image.

    • Example of usage:

    • To set image weight, include --iw at the end of the prompt with the chosen value (e.g., --iw 1.5).

    • Sensitivity to decimal values allows for precision (e.g., 1.4, 2.6).

  • Examples of Image Weight Adjustments

    • Starting with a basic prompt: "pixel art" leads to a diverse set of images.

    • Adding a personal image (self-portrait):

    • Default image weight (1): Influenced output resembling the user with pixel art style.

    • Weighting from 0 to 3 significantly alters the resemblance to the user.

    • 0: Output focuses on "pixel art" with no personal features.

    • 3: Output is very close to a realistic image of the user in pixel art format.

  • Creative Exploration with Image Prompting

    • Using Midjourney creatively:

    • Develop distinct parts of a complex idea separately.

    • Example: Close-up of a woman’s face and an intricate nineties pattern are generated separately to later combine.

    • This method produces sharper, more intense results compared to combining prompts directly.

  • Combining Elements Effectively

    • Step-by-step synthesis:

    • Begin with clean elements (e.g., facial close-up, specific patterns, textures).

    • Resulting images benefit from clear separations of detailed elements before merging them into one final piece.

  • Adding Textures and Final Outputs

    • Including textures (e.g., reflective chrome, cellophane) enhances the complexity and uniqueness of the output.

    • Adjusting parameters like image weight directly affects detail; testing different values (e.g., 0.5, 1.5) gives different artistic results.

  • Conclusion and Future Learning

    • Understanding how each part influences the whole supports creativity in art generation.

    • Upcoming discussion on Seed parameter (how to use past generations to influence current outputs).

  • Key Takeaway

    • The combination of image prompting, adjusting image weight, and using seed parameters enhances the output's uniqueness and alignment with user vision.

    • Continual experimentation with these tools leads to richer, more personalized results.