BUILT FOR ALIGNMENT
Five surfaces.
One job done well.
What the app does and how it gets there, pulled from the code and written for the people who will actually use it. No marketing fog.
01 · Alignment
Twenty-eight anchors.
Real geometry.
Pick the point you want every photo to share. Lign locks pupils, mouth corners, shoulders, animal eyes, or any of twenty-eight detected landmarks across the whole set. The rotation, scale, and translation math runs on the iPhone.
- SMART
When the set is mixed.
Picks the optimal shared angle so no single photo gets rotated too far. The default, and the right answer most of the time.
- CLEAN EDGES
When you hate empty corners.
Fits an inscribed rectangle inside the rotated content so the export has no blank wedges. Best for tight portraits and posted social cuts.
- LEVEL
When level is the point.
Rotates every photo to a perfect zero degrees. Best for clinical portfolios and architectural progress where the horizon line matters.
02 · Object Anchor (Beta)
Tap a thing. Lign aligns to it.
When the subject is an object instead of a face, tap it on the reference photo. Vision isolates the object, learns what it looks like, and finds the same thing in every other shot.
- WORKS FOR
Art, plants, builds, restoration, tattoos.
Anything Vision can isolate as a foreground instance is fair game. Sourdough crumb, a sapling, a model kit, a sleeve in progress.
- HOW
Feature-print matching across the set.
Lign generates a feature print of the tapped object on the reference photo, then locates the closest match in every other frame. The matched bounding box drives the alignment.
- NEURAL ENGINE
Requires a real iPhone.
Object Anchor uses the Neural Engine for instance segmentation. The Simulator does not ship that pipeline; bring the photos onto a phone and it just works.
03 · Manual editor
Tap two points. Override Vision.
When auto-detection picks the wrong face, or when there is no face at all, drop two points on each photo. Lign aligns to those instead.
- PLACE
Three placement shapes.
Tap a pair, tap a single point, or drag a rectangular region. The shape you pick decides which alignment math runs.
- OVERRIDE
Manual takes priority.
Place a manual landmark on a photo and it overrides whatever Vision detected for that shot. The rest of the set keeps its auto-detected anchors.
- PERSIST
Saved with the project.
Manual landmarks persist with the project, so re-exporting a year later uses the same anchors you placed the first time.
04 · Export
Photo or video, ready for socials.
Every export renders from the aligned set with the math pre-applied. 9:16 for Reels, 1:1 for posts, 4:5 for the Instagram feed. Aspect swap with one tap.
- RATIOS
Five aspect ratios.
Original, Square 1:1, Portrait 9:16, Portrait 4:5, Landscape 16:9. Switch ratios from the same aligned set; no re-analysis.
- FIT
Fit or Fill, landmark-aware.
Fit letterboxes the photo. Fill crops to the frame with the anchor still centred, so nothing important falls off the edge.
- WATERMARK
Visible mark on Free, off on Pro.
Free exports carry a small rotating-corner watermark. Pro removes the visible mark. IPTC and EXIF provenance metadata is embedded either way, for authenticity.
- SAVE
H.264 video, direct to Photos.
Slideshow video encoded with AVAssetWriter, saved with add-only authorisation. Lign never reads the rest of your library; it only adds to it.
05 · Privacy
100% on device.
Photos stay yours.
Apple Vision runs on the iPhone. There is no Lign server. There is no upload step. The depth of the technical proof lives on its own page.
LIGN