Planes prepared for the GPU carry their depth of field (padding, radius map, levels) and the
bokeh_boost / bokeh_prefix / bokeh_gather kernels blur them where they are drawn, with the CPU's
spans; GPU-vs-CPU agreement tests for iris shapes, highlights, fringe, Fast Rectangle, tilted
planes and a 2D layer with a 3D track matte.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- Transform: Use Composition's Shutter Angle / Shutter Angle, sampling the transform across
the shutter via EffectHost::params_at (EffectEnv::shutter carries the comp shutter when
motion blur is on); the GPU path defers to the CPU when blurring.
- Glow: Glow Operation (blend modes) and Glow Colors > Arbitrary Map.
- Cartoon: Advanced > Edge Enhancement (shock-filter sharpening / softening).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- Wave Type: Polygon, Image Contours (source layer channel threshold, pre-blur, tolerance,
contour index, source centre) and Mask, rippling outward from the outline (exact signed
distance transform, util::signed_distance).
- Parameters Are Set At Birth via the new EffectHost::params_at; Render Quality; Reflection
off the layer edges; Stroke Profile.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Display Color Space setting joins the live list; View3D::Default maps to view.3d.default for
the Assign Shortcut submenu; stub tests use the stubs that remain.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Reconciled with main's keyframe labels (same Keyframe::label field; keys.setLabel and
keys.selectLabelGroup from main are kept, M14.5's duplicates removed; Edit ▸ Label on
selected keys delegates to the label helper), proxies (RenderOpts::proxy next to
nested_switches/draft_shadows), offloaded renders (Work carries nested_switches), wasm
audio pump (preview mix rate), variable font axes and mask feather points.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- card3d: shared Camera System (Camera Position with rotations, X/Y position and
Transform Order; Corner Pins with focal length; Comp Camera via EffectHost::comp_scene),
Lighting (Distant / Point / First Comp Light, colour, position, depth, ambient) and
Material (diffuse, specular, highlight sharpness).
- Shatter: Custom Shatter Map (+ White Tiles Fixed), Gradient group, Textures group
(front/side/back modes and layers, colour, opacity) and extrusion side rendering.
- Card Wipe: true 3D cards, Back Layer as a layer parameter (old None/Self popup migrated),
Flip Direction, Position / Rotation Jitter.
- Renderer supplies the comp camera and first light to effects.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Layer > Camera > Create Stereo 3D Rig (left/right eye comps with expression-linked eye
cameras, Stereo 3D Controls effect, stereo comp through 3D Glasses), Create Orbit Null,
Create Cameras from 3D Model; Layer > Light > Create Lights from 3D Model
(KHR_lights_punctual), Control Light with Camera, Create Environment Light Background
Layer (equirectangular sky drawn through the camera). glTF cameras and punctual lights
are parsed by effectcraft-model.
Advanced 3D: per-sample motion blur over the comp shutter, and blend modes / track
mattes / Preserve Transparency composited through the 2D path in depth order.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Text: text.animatorFontAxes (replaces the stub) lists the layer font's
variation axes (skrifa) or adds an animator property that offsets one axis
(weighted by the selectors, from the axis default); the renderer redraws each
character's outline at its design-space position (cached per location). The
Timeline's Animate: menu lists the font's axes.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Text: vertical type turns Roman characters on their side (CJK stays
upright) unless Standard Vertical Roman Alignment is on; Tate-Chu-Yoko sets
selected characters horizontally in a one-em cell of the column (layer.setText
tateChuYoko / verticalRomanUpright, Character panel toggles). A Left-to-Right
paragraph is now forced LTR instead of auto-detected. The editing caret follows
animators and text on a path (render::text::caret_maps). Lottie export writes
per-character style runs as index-unit range animators.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
3D-1: Classic 3D depth of field renders the camera's Iris Shape (Fast
Rectangle box or Triangle…Decagon polygon), Iris Rotation, Roundness, Aspect
Ratio and Diffraction Fringe as a span-kernel gather blur, with Highlight
Gain/Threshold/Saturation boosting highlights first; tilted layers blur
progressively by per-pixel camera depth. Layer > Camera > Link Focus Distance
to Point of Interest / to Layer (expressions) and Set Focus Distance to Layer
replace the stubs.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Shapes (SHP-2): Stroke Taper (pixels/percent start/end length, width, ease) and
Wave (amount, pixels/cycles, wavelength, phase) rendered as a variable-width
outline; Dash 2/Gap 2/Dash 3/Gap 3 (shape.dashes.add/remove, timeline +/-
buttons); radial gradient Highlight Length/Angle as a focal gradient.
Masks (MSK-1): mask feather points on the Mask Path (mask.featherPoint.*,
Mask Feather tool add/drag), per-point signed radius and tension, rendered
from the signed distance to the path; carried through vertex edits.
Motion blur (LYR-10): Adaptive Sample Limit comp setting; 2D layers take one
sample per pixel of shutter travel between Samples Per Frame and the limit.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Items (footage and comps) carry a proxy rendered in their place at the source's size when
its switch and Render Settings > Proxy Use allow it; Create Proxy queues a half-resolution
render whose post-render action attaches it. Interpret Footage gains Separate Fields
(upper/lower first, fields become frames at twice the rate), Guess alpha with matte colour,
explicit matte colour, Invert Alpha, Interpret As Linear Light; non-square footage, solids
and precomps are stretched by their pixel aspect in the comp. The four proxy stubs are gone.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
sampleImage renders the sampled layer through the evaluating renderer's footage source
(EvalCtx::footage), cached per thread and frame. File > Import reads .json/.csv/.tsv as data
footage kept in the project. expr.errors collects failing expressions for the error bar;
expr.languageMenu lists the categorised Expression Language menu. Marker keys expose
protectedRegion and cue points; colorDepth, timeToFeetAndFrames, comp.activeCamera.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Comps expose controls (properties typed as text/color/slider/checkbox/point/angle/dropdown,
Media Replacement, groups, comments); precomp layers get an Essential Properties group whose
edits override per instance (Push to Comp / Revert); templates export/import as an open ZIP
(manifest.json + project subset + media + poster). Stretched precomps keep protected regions
at their original speed.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Extruded text keeps per-character 3D: each glyph's outline is extruded in its own space and
placed by its animated matrix; shape layers extrude their fill outlines (gradients use their
mid colour). Shadow-map bias scales with the texel footprint at the receiver and the PCF
reach (no acne on curved casters), identically on CPU and GPU. Docs: Advanced 3D in
architecture.md, parity table updated (3D, Text).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Aux-channel model (effectcraft_raster::channels3d: depth, object/material
ID, normals, UV, coverage, Cryptomatte ranks + manifests, MurmurHash3);
EffectHost::aux; multi-layer OpenEXR channels via the exr crate
(FootageSource::aux); the compositor's depth / layer-ID / Cryptomatte pass
(Renderer::comp_aux) so the effects work on precomps of 3D comps.
3D Channel Extract, Cryptomatte, Depth Matte, Depth of Field, EXtractoR,
Fog 3D, ID Matte, IDentifier.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Advanced 3D comps render their 3D runs as one scene of world-space triangles: model and
primitive layers (meshes, skinning, animation clips by layer time), extruded text/shape
layers, and textured cards for every other layer with their Material Options. Shading is
physically based in linear light: Cook-Torrance GGX + Lambert for parallel/spot/point lights
(AE falloff and cone), ambient lights, image-based light from an Environment light's layer
(equirectangular; cosine irradiance map + roughness mip chain + split-sum BRDF fit), and
PCF shadow maps (spot/parallel/point cube faces). 2x2 SSAA, depth of field as a gather blur
by circle of confusion, back to the working encoding.
The software rasteriser (depth buffer, top-left rule, perspective-correct, opaque visibility
then sorted transparent blending) is the reference and CI path; effectcraft-gpu implements
Accelerator::raster_3d with a wgpu render pipeline (advanced3d.wgsl) and matches it within
tolerance. Environment layers are hidden from both renderers.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
New crate effectcraft-model (L2): glTF 2.0 (.gltf/.glb, data URIs, external buffers and
images, PBR metallic-roughness with base/MR/normal/occlusion/emissive textures, node
hierarchy, skins, step/linear/cubic animations) and OBJ/MTL import, parametric primitives,
extruded/bevelled outline meshes and polygon triangulation with holes.
Project: FootageKind::Model, LayerSource::Model/Primitive, Layer::environment, Geometry and
Material Options builders, Environment light Source. Media probes and loads models
(FootageSource::model). Engine: layer.newModel, layer.new3dPrimitive/newPrimitive,
comp.renderer, layer.environment, material.set/reset/duplicateAssign/revealSource (stubs
replaced); File > Import and layer.addItem pick up models. Hand-written fixtures with
attribution sidecars.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Built-in minimal OCIO config (ACES2065-1, ACEScg, ACEScct, sRGB, Rec.709,
Rec.2020, Display P3, linear variants, XYZ) from published primaries,
transfer functions and Bradford adaptation; CDL v1.2 maths; .cube/.3dl/.csp
and .cc/.ccc/.cdl file transforms with nearest/trilinear/tetrahedral
interpolation and inverse; displays/views/looks.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Keeps the GPU's public placement() API (motion-blur sample count now
shared via mb_samples) and the footprint-limited quantisation; the CPU
fallback helpers draw_layer_cpu / draw_collapsed_cpu pass blank = false.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Auto backend keeps comps with 3D layers on the CPU compositor (3D runs are
still CPU-composited; the GPU round-trip made them slower).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Region-of-interest frames render on the CPU path; the viewer keeps main's
frame drawing (Show Channel, exposure, ROI) and reads GPU-composited frames
back. Keeping frames as native egui-wgpu textures stays implemented behind
RenderSource::gpu_display (off).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
8/16 bpc no longer re-quantises the whole comp after every layer: 2D layers
report the canvas region they can change (transformed buffer bounds plus the
resampling margin; stencil modes, motion blur, perspective, styles,
adjustment and 3D runs fall back to the full canvas), and only that region
is quantised. A quantised buffer placed pixel-exactly (integer translation,
Normal, 100%) over a blank canvas needs nothing; invisible layers need
nothing; collapsed precomps quantise per nested layer. The redundant final
comp pass is skipped unless the comp is re-encoded. Solids are filled with a
pre-quantised colour and skip the source pass unless masked. Row-parallel
vectorisable kernel. Debug builds verify every pixel outside the region is
already quantised (whole workspace test suite run with 8 bpc forced: no
violations).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
New L3 crate effectcraft-gpu: a wgpu compute backend for the per-frame
compositing walk (layer transforms with the CPU's nearest/bilinear/bicubic
sampling and minification pre-filter, motion-blur accumulation, track
mattes, Preserve Transparency, layer style passes and knockout, all 38
blend modes as a WGSL port of color::blend, 8/16 bpc clamp+quantise,
colour-space conversions). 3D runs, adjustment and wireframe layers run on
the CPU between GPU steps. GPU effects with the CPU semantics: Gaussian,
Fast Box and Directional Blur, Glow, Levels, Curves, Hue/Saturation, Tint,
Fill, Gradient Ramp, Fractal Noise, Drop Shadow, Brightness & Contrast,
Exposure, Invert, Transform (registry gpu badge from effects::GPU_EFFECTS).
render: Backend (Cpu/Gpu/Auto) in RenderOpts, the Accelerator trait and
Renderer::accel, GPU effect chains in the effect stack, hooks for the GPU
walk. Project setting gpu_acceleration (Mercury GPU Acceleration / Software
Only), render.backend command, file.projectSettings renderer. Session and
Render Queue exports carry the accelerator. CLI --gpu and bench --gpu.
Tests compare GPU vs CPU renders (blend modes x bit depths, transforms,
masks, mattes, motion blur, colour management, styles, collapse,
adjustment fallback, every GPU effect) and skip without an adapter.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
TextDoc gains character style runs (font, size, colours, tracking, leading,
baseline shift, h/v scale, tsume, faux styles, caps, super/subscript, kerning
metrics/optical/manual, ligatures) and per-paragraph settings (justification
with last-line modes, indents, space before/after, direction, composer,
hanging punctuation), serde-compatible with single-style documents. Layout
handles mixed runs on a line (shared baseline, largest leading), every-line
composer, editing geometry (carets, hit testing, selection quads) and basic
vertical type. Renderer colours per run. Expressions get the sourceText style
API (style, getStyleAt, createStyle, setters with character ranges).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
