Skip to content

Write Page Overview

The Write workspace is where comic artwork meets your narration script. You view each page alongside its detected speech bubbles and panels, assign dialogue speakers, and review action descriptions.

Write workspace showing the page image canvas on the left and the interactive script editor on the right
The Write workspace: image canvas and drawing tools on the left, script elements and speaker tagging on the right.

The screen is split into two synchronized panes with an adjustable divider:

Left Pane · Interactive Canvas

Displays the active page image with overlaid detection boxes. Includes a collapsible floating dock to switch box types (Speech vs. Art), draw rectangles or polygons, adjust zoom, and customize box styling.

Right Pane · Script Editor

Displays an ordered list of script cards matching the boxes on the page. Speech cards let you pick the character speaker; art cards let you edit the visual panel description. Selecting a box on either pane automatically scrolls and focuses the other.


The header above the script editor provides chapter context and primary actions:

  • Breadcrumbs: Shows Project Name > Current Chapter so you always know where you are.
  • Page Indicator: Displays your current position in the chapter (e.g., Page 3 / 24).
  • Characters: Opens the Character Management modal showing the project’s character count.
  • Process: Opens the unified Process Document dialog to run Detection and OCR.
  • Chapters (List icon): Opens the right-side chapter drawer.

Clicking Process opens the configurator modal where you choose your target scope and task:

Process Document dialog showing Scope and Task selection options
Select your scope and task, then click Run Process.
  • Project: Processes every chapter in your manhwa.
  • Chapter: Processes only the currently active chapter.
  • Current Page: Processes only the single page visible in the canvas.
  • Detection & OCR (Recommended): Detects panel and speech boundaries first, then transcribes text inside speech bubbles sequentially.
  • Detection Only: Runs computer vision models to detect speech balloons and art panels without transcribing text.
  • OCR Only: Transcribes text inside speech boxes that were already detected or hand-drawn.

Click the list icon in the top toolbar to open the Chapters Drawer.

Each chapter card displays:

  • Chapter title and sequential order number.
  • ART Progress Bar: Percentage of pages with detected panel boundaries.
  • OCR Progress Bar: Percentage of pages with transcribed text.
  • Clear Button (Eraser): Wipes all detection boxes and script text for that chapter while keeping the source images safe.

Click any chapter card to switch the editor to that chapter.


The Write workspace includes keyboard shortcuts for rapid navigation and tagging:

ShortcutAction
Ctrl + → / ←Next / Previous page
Ctrl + Shift + → / ←Next / Previous chapter
Ctrl + ↑ / ↓Select previous / next box on page
DeleteDelete selected box
Alt + QToggle box type (Speech vs. Art)
Alt + R / PSwitch drawing tool (Rectangle / Polygon)
Ctrl + /Quick-assign last used character to active box
Ctrl + .Open and focus character dropdown
Ctrl + ;Clear character assignment from active box

Once your boxes, dialogue text, and speaker tags are reviewed across your chapters, proceed to the Process workspace to configure audio voices, crop ratios, and generate your narration videos.