Lightbox & Collections
Palaeographic argument is built by juxtaposition: laying one hand beside another, aligning a doubtful letter-form against a securely dated exemplar, and returning to the same arrangement over weeks of work. Archetype’s lightbox is a freeform workspace for exactly this. A researcher gathers images and annotated letter-regions from anywhere in an instance, arranges them on an infinite canvas, adjusts and compares them, annotates them, and saves the result — either privately in the browser or as a citable link a colleague or student can open. It is the platform’s answer to the light table and the photocopy folder, rebuilt for zoomable IIIF imagery.
Collections and persistent worksets
Section titled “Collections and persistent worksets”Two independent save tiers serve two audiences, deliberately.
Collections are named sets of items (manuscript images and individual glyph-regions) that any visitor assembles while using Search & Browse. They persist in the browser (localStorage) with no account required, so an anonymous reader can build a working set across a session and carry it into the lightbox.
Worksets are the durable, shareable tier. A signed-in user saves the entire lightbox state — every workspace, image position, transform, and embedded collection snapshot — to the server as an opaque JSON payload. Each workset receives an unguessable UUID (public_id) and is Private by default; only when its owner sets visibility to Public does the citable link resolve for anyone. This makes a workset a stable scholarly object: a figure in an article, a seminar handout, or a teaching collection that a class opens read-only without disturbing their own work.
The lightbox canvas itself supports multiple workspaces, per-image z-ordering, and full undo/redo history. Everything is persisted client-side in IndexedDB (via Dexie) — images, workspaces, annotations, cropped regions, and sticky notes each in their own store — so a browser-local session survives reloads even without a server round-trip.
Comparing hands side by side
Section titled “Comparing hands side by side”Because Archetype models each annotated letter as a region of its source image (see Data model), the lightbox can load a single glyph cropped straight from the folio via IIIF region requests — the raw material for comparing hands letter by letter.
Comparison mode takes two selected images and offers two views:
- Side-by-side, with independent or locked (synchronised) zoom and pan — drag or scroll one panel and the other tracks it, so two hands stay registered at the same scale.
- Overlay, stacking the two images with an opacity slider to test whether two forms are the work of one scribe.
Each image carries non-destructive transforms — brightness, contrast, saturation, grayscale, invert, rotation, and horizontal/vertical flip — for reading faded ink or normalising lighting across different photographic campaigns. A dedicated region comparison view does the same for cropped excerpts saved from the crop tool. To compare whole manuscripts page by page instead, the separate Compare view opens up to four of them side by side in Mirador.
Annotate, note, and share
Section titled “Annotate, note, and share”Directly on any lightbox image a researcher can draw rectangular or freehand annotations with labels and colours, add sticky notes anywhere on the canvas for working commentary, and crop out regions to save as reusable comparison excerpts.
The workspace exports in several forms for publication and archiving:
| Format | Use |
|---|---|
| One captioned image per page, for print or circulation | |
| Image | Download the source images individually |
| JSON | Full workspace data, including annotation shapes |
| TEI XML | A <facsimile> of <surface> elements with IIIF <graphic> URLs and per-annotation <zone> coordinates |
| A browser-native, captioned print layout |
The TEI export is the interoperable path: rectangular annotations become zones with ulx/uly/lrx/lry bounds, so a lightbox arrangement carries into a TEI-based editorial workflow rather than staying locked in the tool.
The lightbox’s annotation layer is a personal working overlay, distinct from the platform’s curatorial W3C Web Annotation data. Use worksets for citable scholarly output and the annotation apparatus for the authoritative descriptive record.