Create a canvas and add images
Choose square, portrait, story, landscape or custom dimensions, then import several local images at once.
Arrange and colour-grade each layer or apply one look to the finished canvas. Original assets remain untouched; the browser renders effects at the real output size only when you export.
These facts describe the current product, not unshipped roadmap work.
Reviewed by the DataDance product and engineering team · Published · Updated
Confirm important settings before export; the processing location is always disclosed.
Choose square, portrait, story, landscape or custom dimensions, then import several local images at once.
Move, scale and rotate, then target the selected layer or the full canvas with presets, light, colour, beauty and detail controls.
Keep every parameter editable in .imging, or bake the look into full-size PNG, WebP or JPG only when exporting.
ImgIng uses a non-destructive structure similar to professional adjustment layers. Original asset bytes are never re-encoded: layer adjustments run before scale and rotation, while canvas adjustments run after every layer and the background are composited. Both are saved in the open .imging manifest for undo, redo, reopening and hand-off.
The engine estimates skin-colour confidence and mixes edge-preserving smoothing only into likely skin regions. Eyes, hair, clothing texture and transparent contours receive less smoothing. Skin brightening and rosy tone use the same confidence mask, never expanding or shrinking alpha. It is a natural retouching control, not face reshaping or identity recognition.
The interactive preview uses an adaptive resolution, while export preserves the requested canvas dimensions. High-resolution effects run in overlapping bands to control peak memory and avoid seams; layer effect rasters adapt to their actual displayed size and browser limits. Every adjustment preserves source alpha pixel for pixel, so cutouts, shadows and translucent artwork remain valid PNG or WebP.
Twelve presets cover original, natural beauty, clear skin, cream glow, Japanese fresh, warm portrait, cool white, teal and orange, vintage film, mono portrait, food colour and clear landscape. Fine controls include intensity, exposure, contrast, highlights, shadows, saturation, vibrance, temperature, tint, hue, skin smoothing, skin brightening, rosy tone, clarity, sharpening, fade, vignette and grain.
Canvas limits remain 8192 px per side, 48 megapixels and 24 image layers. Skin-aware beauty is a local colour-and-detail algorithm; it is not Photoshop and cannot promise the same aesthetic result for every complexion and lighting condition.
Updated 2026-08-20 · Live capability detection inside the tool is authoritative
These visible answers match the current product behaviour and structured data.
No. Image decoding, filter preview, full-size processing and final compositing all run in the current browser.
No. Parameters are stored separately, source asset bytes stay unchanged and pixel processing preserves alpha exactly.
Yes. The .imging manifest stores adjustments for every image layer and the full canvas alongside original assets and a filtered preview.
Each task page documents real settings, limits and format advice—not keyword-swapped duplicates.