Advanced Prompt Engineering

Lesson 1: Multi-Prompt Syntax and Concept Separation

Lesson Objectives

By the end of this lesson, students should be able to:

  • Use :: to separate prompt concepts for independent treatment
  • Apply multi-prompt syntax to avoid unintended concept blending
  • Understand when multi-prompt is more effective than single-prompt
  • Test multi-prompt results vs. single-prompt for specific subjects

Lesson Content

What multi-prompt syntax does.

The :: separator in Midjourney divides a prompt into separate concept segments that the model treats independently rather than blending together. Standard prompt: "hot dog" (generates the food). Multi-prompt: "hot :: dog" (generates a hot dog breed or a warm/hot dog – the concepts are separated). This separation is powerful for: subjects where two words have unintended combined meaning; complex multi-element images where specific concepts should not blend; and deliberate concept emphasis through weighting.

Preventing unintended blending.

Many compound concepts have unintended interpretations when written together: "space ship" might be interpreted as a spacecraft or a ship in space; "black bird" might produce a dark-colored bird or a bird called a blackbird specifically. Multi-prompt separation allows specifying the actual intent: "space :: ship" explicitly separates the location (space) from the subject (ship) – preventing the compound-word shortcut interpretation.

When to use multi-prompt vs. single prompt.

Multi-prompt adds complexity and should be used when it solves a specific issue: unintended compound-word interpretation, complex multi-element scenes where specific separation is needed, or deliberate concept weighting (addressed in the next lesson). For straightforward descriptive prompts without compound issues, single-prompt is simpler and equally effective.

Testing multi-prompt vs. single-prompt.

For subjects where compound interpretation might be an issue, test both: a single-prompt and a multi-prompt version. Compare results to determine whether the :: separation produces meaningfully different or better output for the specific subject. Not every subject benefits from multi-prompt – test before committing to it.

Enroll to continue this lesson.

The preview above shows the lesson objectives and opening lesson content. Enroll to view the full lesson, complete the practice work, and take the lesson quiz.

Log In / Create Account
Back to Course