Basic image manipulation: Upscale, vary, zoom, and pan

Once you’ve generated an image in Midjourney, your creative process doesn’t stop there. Midjourney provides essential image manipulation tools—upscale, vary, zoom out, and pan—to help you refine, expand, or explore new directions with your image. These operations are simple to apply but carry significant creative potential.

In this section, we’ll first introduce what each of these four features does. Then, we’ll walk through how to use them on both the Midjourney web interface and the Discord interface, noting differences where they exist. For easier learning, we’ll also indicate where screenshots should be included.

What are upscale, vary, zoom out, and pan?

Before diving into each UI, it’s helpful to understand the purpose of each operation:

Upscale

Upscaling enhances your chosen image by increasing its resolution, texture quality, and overall detail. You begin with a 4-image grid preview and can upscale one of them for a more refined version.

Vary

Vary creates new image options based on one of your existing results. You can choose between subtle variations (minor changes in color, form, or detail) or strong variations (broader composition or style shifts). This helps you explore alternatives without writing a new prompt.

Zoom out

Zooming out adds more visual space around the existing image content while keeping the original subject in the center. You can choose preset zoom levels (1.5x or 2x), allowing you to frame the subject in a wider context.

Pan

Panning extends the canvas in a specific direction—left, right, up, or down. The newly generated content builds upon the original scene as if you're moving a virtual camera to reveal more of the environment.

These functions are available across both Midjourney platforms, though the interface to access them differs slightly.

Using upscale, vary, zoom out, and pan on the web interface

After the initial grid is generated, you can refine any of the four images to move closer to your desired result. Simply hover over an image to reveal quick action buttons, or click on the image to open a detailed view with more regeneration options.

The most commonly used tools are:

Vary

• Subtle: Makes small changes while keeping the core composition and structure.
• Strong: Introduces more noticeable differences in shape, layout, or details.
  
  

Upscale

• Subtle: Enhances resolution with minimal stylistic changes.
• Creative: Adds more artistic flair, sometimes altering finer details or lighting for visual impact.

These options let you control how much variation or enhancement you want to apply when refining an image.

If you click into the image, you’ll see additional actions such as:

• Pan: Extends the image in a specific direction (up, down, left, or right).
• Zoom: Adds framing space around the image for an expanded view.
• More Options: Includes rerun, edit, and prompt reuse actions.

The image below is the result of panning upward—you can see more of the sky and flying vehicles above the original frame, expanding the vertical scope of the scene.

This is the result of a 2x zoom-out—you can see a significantly wider view of the cityscape, including more surrounding buildings, flying vehicles, and the extended river leading into the horizon.

These tools let you regenerate and explore variations based on the original image—perfect for fine-tuning results or exploring creative branches without starting over.

Using upscale, vary, zoom out, and pan on Discord

Midjourney’s Discord interface provides interactive buttons to manipulate generated images. While Discord leans toward a more command-driven experience, many image operations are accessible with a simple click once your image is generated.

Understanding the image grid and button layout

When you submit a prompt in Discord, Midjourney returns a 2x2 image grid, labeled with U1–U4 and V1–V4 buttons beneath it.

Here’s how the image numbering works:

• Image 1: Top-left
• Image 2: Top-right
• Image 3: Bottom-left
• Image 4: Bottom-right
  
  

The buttons below the grid correspond to these positions:

• U1–U4: “Upscale” the selected image to a high-resolution version.
• V1–V4: “Vary” the selected image to generate four alternatives.

Understanding post-upscale options: Further refinement, variation, and canvas expansion

After upscaling an image in Discord, Midjourney presents an extended set of options for additional editing. These tools help you either refine the current image, generate new variations, or expand the scene.

Upscale options

• Upscale (Subtle): Enhances the image with finer details and sharper resolution while keeping the composition and styling consistent. Useful when you're nearly satisfied but want a cleaner final output.
• Upscale (Creative)
  Offers a more imaginative upscale that may slightly reinterpret elements of the image—such as adjusting lighting, form, or texture. A good choice if you're open to modest artistic shifts during enhancement.
  
  
• Vary (Subtle): Creates four new versions with minimal changes in details or composition. Ideal for fine-tuning without deviating far from the original look.
• Vary (Strong): Produces more noticeably different interpretations, potentially altering layout, perspective, or style. Useful when you want to explore new directions while keeping the same concept.
• Vary (Region): This advanced feature allows you to select and regenerate a specific part of the image without affecting the rest. It’s particularly helpful when you want to change localized areas—such as fixing a face, redesigning an outfit, or modifying a background element.

Vary options

Example: Removing a targeted flying car using Vary (Region)

Using Vary (Region) in Discord, you can easily remove specific elements from your image—like one flying car—without affecting the rest of the scene. Select the area using the selection tool. Then, type a simple prompt: "remove this flying car” or “clear sky with sunset clouds."

The flying car is gone. The area now blends naturally into the sky.

Zoom and pan functions

• Zoom Out 1.5x / 2x: These options add new visual space around the current image, maintaining the main subject's position while extending the canvas. The 1.5x and 2x choices offer different levels of expansion. Midjourney fills in the surroundings to match the original style.
• Custom Zoom: Provides manual control over zoom level and aspect ratio. You can adjust how much to zoom out and reshape the image dimensions (e.g., for widescreen or vertical compositions).
• Pan (Left, Right, Up, Down): Lets you move the canvas in a specific direction, revealing new content along the edges. The original image stays anchored while Midjourney generates a seamless extension in the chosen direction. This is helpful for building out landscapes or repositioning compositions gradually.
• Favorite: Marks the image for quick access later on the Midjourney web platform.
• Web: Opens the image on the Midjourney Web interface. This is useful for downloading, organizing, or further editing using Web-based tools like Remix or image history.

Additional options

---

Understanding how to use upscale, vary, zoom out, and pan gives you more than just refinement options—it gives you creative control. These tools let you reshape the output without rewriting your prompt or starting over from scratch. The web UI makes this process highly visual and intuitive, while the Discord interface gives fast access for experienced users who enjoy a command-based flow.

Once you’re comfortable with these basics, you’ll be ready to explore more advanced image editing and re-prompting techniques, which we’ll cover in the next section.