Crop, deskew, caption — without leaving the catalog.
A canvas-based image editor built right into the file detail view. Crop to preset archival bounds, deskew, run AI object detection, generate an archival caption, and export — without round-tripping to Photoshop.
- Tools
- 8
- Crop presets
- 6
- Filters
- 5
Six preset bounds, all visible at once.
Thumbnail (1:1), Detail Banner (16:9), Card Preview (9:16), Social Share (16:9), Catalog Print (free), and Full Frame (free) — every preset shows on the canvas at the same time, each with corner and edge handles. Auto-save per config, so you never lose a crop.
- Drag to draw, move, or resize each bound
- Aspect ratios enforced per preset
- All bounds visible simultaneously — see every output at once
- Auto-save per crop config — pick up where you left off
Adjustments and rotation, slider-driven.
Brightness, contrast, saturation, sharpness — all slider-controlled with a per-slider reset. Free rotation slider plus 90° CW / CCW buttons. Manual deskew with an angle slider, or one-click Auto-Detect.
- Brightness · Contrast · Saturation · Sharpness sliders (-100 → +100)
- Free-angle rotation + 90° CW / CCW snap
- Horizontal and vertical flip
- Manual deskew slider (-45° → +45°) with Auto-Detect
- Filters: None · Grayscale · Sepia · Invert · Auto-Levels
Before / after compare with a draggable split.
Toggle compare mode and a draggable split divider appears on the canvas — drag left/right to reveal more or less of the original. Catch every change before you commit.
- Draggable split divider over the canvas
- Undo / redo stack with Ctrl+Z and Ctrl+Shift+Z
- Zoom 25% – 400% with scroll wheel · Fit · 100% buttons
- Dark checkerboard canvas background — see transparency clearly
AI object detection and archival caption.
Run AI detection to surface labelled bounding boxes — Document, Handwritten Text, Official Seal, Photograph, Signature, Page Number — each with a confidence score. Generate an archival caption in three formats: Brief, Detailed, or Archival (MARC-style).
- Per-detection visibility toggle and confidence score
- Three caption formats: Brief · Detailed · Archival
- Copy caption to clipboard for use anywhere
- All AI output goes through track-changes — accept, edit, reject
Lossless export, three formats.
Export as PNG (lossless), JPEG (92%), or WebP — straight from the canvas. The original asset stays untouched on object storage; exports become new file children of the parent item.
- PNG · JPEG · WebP export targets
- Original asset preserved — exports become children of the item
- Loading state while the source image streams in
- Back to Item Detail navigation always one click away
Ready to see it on your collection?
Load a few records. Run the AI. Review and publish — before lunch.