vangel76andClaude Opus 5.5 39bd0efac2 Luma Key: Key Out Brighter and Darker cut hard at Threshold, as After Effects ...

After Effects' Luma Key removes everything darker (or brighter) than Threshold
and ignores Tolerance for those two key types; Tolerance only sets the band of
Key Out Similar / Dissimilar. EffectCraft ramped the matte over Tolerance, so
case 719 (a 50 % grey, Key Out Darker, Threshold 140, Tolerance 20) kept about
80 % of the layer where AE keys it out completely.

Measured in AE 2026 on a vertical grey ramp (cases 812-814): Darker keys up to
level 139.9 and keeps 140.1 (EffectCraft now the same row); Similar keys levels
120.3-160.0; with Edge Thin 1 / Edge Feather 2.5 the matte rises over rows
591-596 (EffectCraft within 0.2 of AE's alpha there). CPU and GPU.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

2026-10-19 16:15:02 +02:00

..

automation

fix(automation): avoid replaying edits after lost replies

2026-10-15 02:13:41 -06:00

av1enc

G2: improve editing reliability and Windows rendering fidelity

2026-10-06 00:00:31 -05:00

color

Blend modes: classic modes fade with opacity and HDR modes keep over-range values, as in After Effects (#494)

2026-10-10 20:51:24 +00:00

effects

Luma Key: Key Out Brighter and Darker cut hard at Threshold, as After Effects

2026-10-19 16:15:02 +02:00

engine

Format merged community PRs

2026-10-18 09:37:06 -04:00

export

Merge pull request #1753

2026-10-18 09:35:35 -04:00

expr

Merge branch 'main' into expr-property-group-1619

2026-10-16 14:28:34 +02:00

geom

Never crash: deny panic lints in the already-clean crates and apps

2026-10-04 14:52:39 -04:00

gpu

Luma Key: Key Out Brighter and Darker cut hard at Threshold, as After Effects

2026-10-19 16:15:02 +02:00

hevcenc

Untrack crates/hevcenc/target (36 MB of build output) and ignore nested target/

2026-10-06 15:02:20 -04:00

host

Merge pull request #1216

2026-10-18 09:35:32 -04:00

interchange

Never crash: reject timeline documents nested too deeply

2026-10-04 14:05:57 -04:00

keyframe

Masks: addVertex and setVertex fill in missing tangents of a loaded path (#1250)

2026-10-13 11:04:02 -06:00

lottie

Footage: follow stored footage paths only on this computer or the project's share

2026-10-16 17:43:58 +00:00

media

Land batch: fix clippy lints and the #1424 test after #1413 changed layer_menu

2026-10-18 09:42:33 -04:00

model

Merge pull request #1376

2026-10-18 09:35:33 -04:00

opusenc

G2: improve editing reliability and Windows rendering fidelity

2026-10-06 00:00:31 -05:00

path

G2: improve editing reliability and Windows rendering fidelity

2026-10-06 00:00:31 -05:00

pdf

Merge pull request #1527

2026-10-18 09:35:34 -04:00

plugin

G2: improve editing reliability and Windows rendering fidelity

2026-10-06 00:00:31 -05:00

project

#1714: test several comps, negative and malformed saved times

2026-10-17 15:35:12 +05:30

psd

PSD: the optimisation tests use the reference decoder once #1137 and #1139 land together

2026-10-13 04:47:52 -04:00

raster

Cryptomatte: the rank planes are resolved once per evaluation and the GPU path reduces them on all cores, instead of formatting and looking up channel names for every pixel in a single-threaded loop (#473)

2026-10-12 13:29:03 +02:00

render

Levels: channel controls before RGB, "Off for 32 bpc Color" unclipped in 32 bpc, as After Effects

2026-10-19 16:14:42 +02:00

script

Merge pull request #1696

2026-10-18 09:36:05 -04:00

segment

Merge pull request #374 from storytold/fix/368-tflite-budgets

2026-10-09 09:25:24 -04:00

svg

Merge pull request #1369

2026-10-18 09:35:33 -04:00

text

Land batch: fix clippy lints and the #1424 test after #1413 changed layer_menu

2026-10-18 09:42:33 -04:00

time

Format negative Audio Samples time as a signed timecode

2026-10-18 14:21:15 +02:00

track

Tracker: the 1080p timing test takes the fastest step, so a loaded FreeBSD VM doesn't fail it

2026-10-13 04:47:27 -04:00

ui-egui

Merge pull request #1317

2026-10-18 11:12:44 -04:00

vp9enc

VP9: frames never end in a byte that reads as a superframe marker (#161)

2026-10-07 14:15:33 +07:00