Image Workspace (@vpaint)

01. Image Workspace overview

Image Workspace is a browser-based AI image studio. Create a new image with natural language, transform a photo or sketch, remove or replace a selected region, and finish the result with layers, shapes, text, adjustments, and filters.

02. Start with @vpaint and a launch link

A workspace link is valid for one hour and binds to the first browser that opens it. The same link can be rejected in another browser or device, so keep the original window open while working.

03. Editor layout

Use Preview to zoom or fit the document. Layer Details provides exact X and Y positions, width, height, rotation, and opacity. With Auto select enabled, clicking an object on the canvas quickly activates its layer.

04. Create a document and import images

• Choose File → New and set the canvas size and background to start from an empty document.
• Bring in an external image from a file, directory, URL, data URL, webcam, or image search.
• Drop a file onto the canvas or press Ctrl/Cmd+V to add it as a new layer.
• AI-panel Upload image does not place the file first; it reserves that file as the input to the next Edit request.

Canvas Size defines the outer boundary of the final document; layer size defines how an individual image is placed. Change Canvas Size to add working space, and Resize to change the image's pixel dimensions.

05. AI generation and infographics

1. Describe the subject, setting, composition, style, lighting, aspect ratio, and intended use in the AI prompt.
2. For an information card, enable Infographic and specify the title, sections, figures, and display language.
3. Select Generate and watch the progress label and elapsed time. The result arrives as a new layer.

Prompts may be written in Korean, Japanese, English, or another supported language. Titles, labels, and captions inside the image keep the request language unless you say otherwise, but always zoom in and proofread generated lettering.

06. AI editing and target priority

Edit resolves its input from the current state. The priority is an image uploaded in the AI panel, then the selected image layer, a selected text/shape/drawing layer, and finally the full visible canvas.

07. Compose up to four reference images

1. Import each image you want to combine as a separate layer.
2. In Layers, hold Cmd on macOS or Ctrl on Windows and click two to four layers in the order you want them used.
3. Check the numbered badges and reference list, then describe the role of each source in the prompt.
4. Select Edit. The originals stay intact and the composition is added as a new layer.

Example: “Combine the person from reference 1, clothing from 2, and background from 3; preserve facial identity and the product logo; 4:5 advertising photo.” A regular click clears the reference set and returns to single-layer selection.

08. Inpaint: remove, replace, or remove a background

1. Select an unrotated image layer and choose Inpaint.
2. Choose an action and paint slightly beyond the object or region that should change.
3. For replacement, describe the new object plus the lighting, angle, and material that must match the surroundings.

09. Outpaint: extend the scene

1. Select an unrotated image layer and choose Outpaint.
2. Drag an edge or corner handle outward to set the left, right, top, and bottom expansion. Each side supports up to 700 px.
3. Keep Grow canvas to fit enabled when the expanded scene should remain inside the document boundary.
4. Select Apply. AI continues the scene and expands the target layer while keeping the original pixels in place.

10. Layers and nondestructive editing

Layers keep the background, subject, product, lettering, and decoration independent. Rename and reorder them, toggle visibility, and apply an effect to a duplicate so you can compare options while preserving the source.

• Use New, New from Selection, Duplicate, and Delete to separate elements.
• Use Composition and Opacity to control how layers overlap.
• Merge Down combines a local pair; Flatten Image combines the full document.
• When an image-only tool rejects vector or text content, duplicate it and convert the copy to raster.

11. Selection, drawing, and retouching tools

12. Image adjustments and effects

The Image menu covers canvas and pixel dimensions, rotation, flips, opacity, brightness, contrast, hue, saturation, luminance, automatic color adjustment, color depth, palettes, and histograms. Effect Browser helps compare filters before applying them.

• Photo cleanup: denoise → small brightness, contrast, and white-balance changes → sharpen
• Mood: vignette, grain, sepia, vintage, and Instagram filters
• Graphic treatments: mosaic, emboss, blueprint, pencil, oil, and dot screen
• Focus: Gaussian, box, and zoom blur plus tilt shift

13. Text, shapes, and precise layout

Keep text and shapes on separate layers, then set exact position, size, rotation, and opacity in Layer Details. Grid and Guides make margins and alignment consistent across cards, banners, and thumbnails.

• Letters generated inside an AI image are pixels and are hard to edit precisely. Generate the background first, then place exact copy with the Text tool.
• Add logos, prices, captions, and any other exact content as separate layers near the end.
• If rotation makes a tool unavailable, preserve the original and work on a rasterized duplicate.

14. Write a strong image request

An edit request should say both what changes and what stays. Example: “Change only the bag to deep navy. Preserve its shape, metal logo, the person's hands, and the background lighting.”

15. Undo, redo, and shortcuts

AI Edit, Inpaint, and Outpaint bundle their related image and geometry changes so one Undo restores the previous state. Duplicate a layer and use Quick Save before a major operation.

16. Export and Save to VORA

Find VORA-saved files under Settings → My AI Generated Files. If the workspace reports a file-size or storage limit, remove files you no longer need or reduce the canvas and image dimensions before saving again.

17. Recommended workflows

18. Troubleshooting and final review

• Inspect faces, hands, lettering, logos, product shapes, and mask boundaries at 100% zoom.
• Review factual accuracy, rights, personal information, and the policies of the destination where the image will be used.
• Open the final file and confirm its dimensions, ratio, transparency, color, and save location.