Fold

Kaleidoscopic and fractal-symmetric compositions from photographs. Five transform modes, intelligent focal point discovery, and XMP provenance that records every parameter.

Screenshot placeholder Main workspace showing kaleidoscope preview with source image, transform controls, mode selector, and Discover button. Suggest capturing at 1200x750px with retina resolution.

Symmetry from chaos

Kaleidoscope
Radial symmetry with configurable segments
Mandala
Rotational symmetry with alternating mirrors
Bilateral
Simple left-right mirror
4-Way Mirror
Quadrant-based symmetry
Hall of Mirrors
Recursive depth with vanishing point

Find the beauty

Discover

Composite saliency (attention, edge, pose, skin) automatically finds interesting focal points in your source.

Source Pre-Processing

Flip, mirror, rotate, and quad-exposure effects before the geometric transform.

Live Preview

Real-time Metal rendering as you adjust center point, rotation, zoom, and segments.

Metal-Accelerated

GPU rendering for instant feedback at full resolution.

XMP Provenance

Every export records mode, segments, center, rotation, and all transform parameters in fold: namespace.

Edit in Photo Developer

Cmd+Option+E renders 16-bit TIFF and opens in PD for color grading and finishing.

Connected via XMP

Fold works with Photo Developer and Photo Archive Lite through XMP sidecars. Transform provenance follows the image through the entire workflow.

  • Export to Photo Developer via photodeveloper://open?file= URL scheme
  • PD preserves fold: attributes on all subsequent writes
  • Photo Archive Lite detects Fold-processed files automatically
  • "Fold" appears in PAP's Processed By filter
  • Provenance keywords injected on Digg Photos upload
Screenshot placeholder Diagram showing workflow: PAP selects photo → DKS transforms → PD grades → PAP catalogs result. XMP sidecar with fold: namespace fields.
"In every chaos there is a cosmos, in every disorder a secret order."

— Carl Jung

Fold

A focused kaleidoscope tool built on Metal with saliency-driven focal point discovery. Five transform modes, real-time preview, and full XMP provenance.