test(producer): add hdr-regression and hdr-hlg-regression test suites

.github/workflows/regression.yml

@@ -2,21 +2,10 @@
 
 on:
   pull_request:
-    branches: [main]
   push:
     branches:
       - main
 
-concurrency:
-  group: regression-${{ github.ref }}
-  cancel-in-progress: true
-
-# Least-privilege token: only reading code. Jobs that need more (e.g. GHA
-# cache reads/writes from docker/build-push-action with `type=gha`) elevate
-# their own permissions inline.
-permissions:
-  contents: read
-
 jobs:
   changes:
     name: Detect changes
@@ -36,132 +25,89 @@
               - "packages/engine/**"
               - "Dockerfile*"
 
-  # Build the regression Docker image once, export it as a tarball, and upload
-  # as an artifact. Each matrix shard then downloads + `docker load`s it instead
-  # of rebuilding from cache. Measured on PR #419: the Docker build step takes
-  # ~4 min per shard even with GHA cache, so 11 shards = ~44 min of redundant
-  # build time per run. This job replaces that with a single ~4 min build plus
-  # ~15s of artifact download per shard.
-  build-image:
-    name: Build regression test image
-    needs: changes
-    if: needs.changes.outputs.code == 'true'
-    runs-on: ubuntu-latest
-    timeout-minutes: 20
-    permissions:
-      contents: read
-      actions: write # docker/build-push-action `type=gha` cache reads + writes
-    steps:
-      - name: Checkout
-        uses: actions/checkout@v4
-        # No LFS needed here — Dockerfile.test only copies source + package manifests,
-        # not the golden baselines under packages/producer/tests/**/output.
-
-      - name: Set up Docker Buildx
-        uses: docker/setup-buildx-action@v3
-
-      - name: Build test image to tarball
-        uses: docker/build-push-action@v6
-        with:
-          context: .
-          file: Dockerfile.test
-          tags: hyperframes-producer:test
-          cache-from: type=gha,scope=regression-test-image
-          cache-to: type=gha,mode=max,scope=regression-test-image
-          outputs: type=docker,dest=/tmp/regression-test-image.tar
-
-      - name: Report image size
-        run: ls -lh /tmp/regression-test-image.tar
-
-      - name: Upload image artifact
-        uses: actions/upload-artifact@v4
-        with:
-          name: regression-test-image
-          path: /tmp/regression-test-image.tar
-          retention-days: 1
-          compression-level: 1
-
   regression-shards:
-    needs: [changes, build-image]
+    needs: changes
     if: needs.changes.outputs.code == 'true'
     runs-on: ubuntu-latest
-    timeout-minutes: 40
+    timeout-minutes: 60
     strategy:
       fail-fast: false
       matrix:
         include:
           - shard: fast
             args: "--sequential --exclude-tags slow,render-compat,hdr"
           - shard: render-compat
             args: "--sequential gsap-letters-render-compat css-spinner-render-compat raf-ball-render-compat iframe-render-compat"
           - shard: hdr
-            args: "--sequential hdr-pq hdr-image-only"
+            args: "--sequential hdr-regression hdr-hlg-regression"
           - shard: styles-a
             args: "style-1-prod style-2-prod style-3-prod"
           - shard: styles-b
             args: "style-4-prod style-5-prod style-6-prod"
           - shard: styles-c
             args: "style-7-prod style-8-prod style-9-prod"
           - shard: styles-d
             args: "style-10-prod style-11-prod style-12-prod"
           - shard: styles-e
             args: "style-13-prod style-15-prod style-16-prod"
           - shard: styles-f
             args: "style-17-prod style-18-prod"
           - shard: styles-g
             args: "overlay-montage-prod"
     steps:
       - name: Checkout (with LFS)
         uses: actions/checkout@v4
         with:
           lfs: true
 
       - name: Validate LFS files
         run: |
           echo "Checking golden baseline MP4s are real files (not LFS pointers)..."
           for mp4 in packages/producer/tests/*/output/output.mp4; do
             if [ -f "$mp4" ]; then
               size=$(stat --format=%s "$mp4")
               if [ "$size" -lt 1000 ]; then
                 echo "ERROR: $mp4 appears to be an LFS pointer ($size bytes)"
                 exit 1
               fi
               echo "OK: $mp4 ($size bytes)"
             fi
           done
 
-      - name: Download test image artifact
-        uses: actions/download-artifact@v4
-        with:
-          name: regression-test-image
-          path: /tmp
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
 
-      - name: Load test image
-        run: |
-          docker load -i /tmp/regression-test-image.tar
-          docker image ls hyperframes-producer:test
+      - name: Build test Docker image (cached)
+        uses: docker/build-push-action@v6
+        with:
+          context: .
+          file: Dockerfile.test
+          load: true
+          tags: hyperframes-producer:test
+          cache-from: type=gha,scope=regression-test-image
+          cache-to: type=gha,mode=min,scope=regression-test-image
 
       - name: "Run regression shard: ${{ matrix.shard }}"
         run: |
           echo "Shard: ${{ matrix.shard }}"
           echo "Args: ${{ matrix.args }}"
           docker run --rm \
             --security-opt seccomp=unconfined \
             --shm-size=4g \
             -v ${{ github.workspace }}/packages/producer/tests:/app/packages/producer/tests \
             hyperframes-producer:test \
             ${{ matrix.args }}
 
       - name: Upload failure artifacts
         if: failure()
         uses: actions/upload-artifact@v4
         with:
           name: regression-failures-${{ matrix.shard }}
           path: packages/producer/tests/*/failures/
           if-no-files-found: ignore
 
   # Summary job — matches the required check name in branch protection
   regression:
     runs-on: ubuntu-latest
     needs: [changes, regression-shards]
     if: always()

packages/engine/src/utils/ffprobe.test.ts

@@ -55,7 +55,7 @@ describe("extractVideoMetadata", () => {
   it("reads HDR PNG cICP metadata when ffprobe color fields are absent", async () => {
     const fixturePath = resolve(
       __dirname,
-      "../../../producer/tests/hdr-image-only/src/hdr-photo.png",
+      "../../../producer/tests/hdr-regression/src/hdr-photo-pq.png",
     );
 
     const metadata = await extractVideoMetadata(fixturePath);
@@ -102,7 +102,7 @@ describe("extractPngMetadataFromBuffer", () => {
 
   it("continues to parse the checked-in HDR PNG fixture", () => {
     const fixture = readFileSync(
-      resolve(__dirname, "../../../producer/tests/hdr-image-only/src/hdr-photo.png"),
+      resolve(__dirname, "../../../producer/tests/hdr-regression/src/hdr-photo-pq.png"),
     );
     expect(extractPngMetadataFromBuffer(fixture)?.colorSpace?.colorTransfer).toBe("smpte2084");
   });

packages/producer/tests/hdr-hlg-regression/README.md

@@ -0,0 +1,52 @@
+# hdr-hlg-regression
+
+Regression test that locks down end-to-end **HDR HLG (BT.2020 ARIB STD-B67)**
+video rendering. Companion to `hdr-regression` (PQ), kept as a separate suite
+so the HLG-specific encoder/metadata path stays tested in isolation.
+
+## What it covers
+
+| Window | Time         | Shape                                 | Expected |
+| ------ | ------------ | ------------------------------------- | -------- |
+| A      | 0.0 – 2.5 s  | Baseline HLG video + DOM overlay      | pass     |
+| B      | 2.5 – 5.0 s  | Wrapper opacity fade around HLG video | pass     |
+
+The test pins the contract that:
+
+- `extractVideoMetadata` reports `bt2020/arib-std-b67/limited` for the HLG
+  source (i.e. HLG is detected and not silently coerced to PQ).
+- `isHdrColorSpace` flips the orchestrator into the layered HDR path on the
+  HLG signal.
+- The HLG source is decoded into `rgb48le` and blitted under the SDR DOM
+  overlay on every frame.
+- Wrapper-opacity composition (window B) does not break HLG pass-through.
+- `hdrEncoder` writes HEVC Main10 / `yuv420p10le` / BT.2020 HLG with the
+  correct color tags (no PQ mastering display metadata for HLG).
+
+The suite is intentionally short (5 s, two windows) — it exists to detect
+regressions in the HLG-specific code path, not to enumerate every composition
+shape (those live in `hdr-regression`).
+
+## Tolerance
+
+`maxFrameFailures` is **0** here. HLG is a pure pass-through path — no known
+failures, no transcoder workarounds — and HEVC encoding against the rendered
+`rgb48le` buffer is byte-deterministic on the same fixture. Any drift is a
+real regression, not codec noise, so the budget is the strictest possible.
+
+## Fixture
+
+`src/hdr-hlg-clip.mp4` — last 5 seconds of a user-recorded HEVC HLG clip,
+remuxed (no re-encode) so the HLG color tags survive verbatim.
+
+## Running
+
+```bash
+cd packages/producer
+bun run test hdr-hlg-regression
+
+bun run test:update hdr-hlg-regression
+```
+
+In CI it runs in the `hdr` shard alongside `hdr-regression`
+(see `.github/workflows/regression.yml`).

packages/producer/tests/hdr-hlg-regression/meta.json

@@ -0,0 +1,14 @@
+{
+  "name": "hdr-hlg-regression",
+  "description": "Regression test for HDR HLG (BT.2020 ARIB STD-B67) video pass-through. Two windows: (A) baseline HLG video + SDR DOM overlay, (B) wrapper-opacity fade applied to an HLG video. Verifies that an HLG HEVC source drives the layered HDR pipeline end-to-end (extractVideoMetadata reports hlg, ffmpegFrameSource decodes correctly, hdrEncoder writes HEVC Main10 / yuv420p10le / BT.2020 HLG with the appropriate mastering metadata) and that opacity composition does not break HLG signal pass-through.",
+  "tags": ["regression", "hdr"],
+  "minPsnr": 28,
+  "maxFrameFailures": 0,
+  "minAudioCorrelation": 0,
+  "maxAudioLagWindows": 1,
+  "renderConfig": {
+    "fps": 30,
+    "workers": 1,
+    "hdr": true
+  }
+}

packages/producer/tests/hdr-hlg-regression/src/index.html

@@ -0,0 +1,112 @@
+<!doctype html>
+<html lang="en">
+  <head>
+    <meta charset="UTF-8" />
+    <title>HDR HLG Regression Suite</title>
+    <script src="https://cdn.jsdelivr.net/npm/gsap@3.14.2/dist/gsap.min.js"></script>
+    <style>
+      html,
+      body {
+        margin: 0;
+        padding: 0;
+        background: #000;
+      }
+
+      #main {
+        position: relative;
+        width: 1920px;
+        height: 1080px;
+        overflow: hidden;
+        background: #000;
+      }
+
+      .hdr-video {
+        position: absolute;
+        inset: 0;
+        width: 100%;
+        height: 100%;
+        object-fit: cover;
+        display: block;
+      }
+
+      .label {
+        position: absolute;
+        padding: 14px 22px;
+        font-family: -apple-system, BlinkMacSystemFont, "Segoe UI", sans-serif;
+        font-size: 36px;
+        font-weight: 700;
+        color: #ffffff;
+        background: rgba(0, 0, 0, 0.6);
+        border-radius: 8px;
+        letter-spacing: 0.04em;
+        z-index: 100;
+      }
+
+      .label-tl {
+        top: 48px;
+        left: 64px;
+      }
+
+      /* Window B: opacity-tweenable wrapper around the HLG video */
+      #window-b-wrapper {
+        position: absolute;
+        inset: 0;
+        opacity: 1;
+      }
+    </style>
+  </head>
+  <body>
+    <div
+      id="main"
+      data-composition-id="hdr-hlg-regression"
+      data-start="0"
+      data-duration="5"
+      data-width="1920"
+      data-height="1080"
+    >
+      <!-- Window A · Static baseline HLG · 0.0–2.5s -->
+      <video
+        id="wa-video"
+        class="clip hdr-video"
+        data-start="0"
+        data-duration="2.5"
+        data-track-index="0"
+        src="hdr-hlg-clip.mp4"
+        muted
+        playsinline
+      ></video>
+      <div class="label label-tl clip" data-start="0" data-duration="2.5">
+        A · HLG baseline + DOM overlay
+      </div>
+
+      <!-- Window B · Wrapper opacity fade on HLG · 2.5–5.0s -->
+      <div id="window-b-wrapper">
+        <video
+          id="wb-video"
+          class="clip hdr-video"
+          data-start="2.5"
+          data-duration="2.5"
+          data-track-index="0"
+          src="hdr-hlg-clip.mp4"
+          muted
+          playsinline
+        ></video>
+      </div>
+      <div class="label label-tl clip" data-start="2.5" data-duration="2.5">
+        B · HLG wrapper opacity fade
+      </div>
+    </div>
+
+    <script>
+      window.__timelines = window.__timelines || {};
+
+      const tl = gsap.timeline({ paused: true });
+
+      // Window B · wrapper opacity 1 → 0.15 → 1 inside the 2.5s window
+      tl.to("#window-b-wrapper", { opacity: 0.15, duration: 1.0, ease: "power2.inOut" }, 2.75);
+      tl.to("#window-b-wrapper", { opacity: 1.0, duration: 1.0, ease: "power2.inOut" }, 3.75);
+
+      window.__timelines["hdr-hlg-regression"] = tl;
+    </script>
+  </body>
+</html>