Tighten screenshot comparison coverage

[matthargett]

Summary

• tighten visual comparison budgets for GUI3D SpherePanel, GUI Slate, and GUI Near Menu
• add Standard/PBR texture-repetition coverage with native-compatible RGBA references
• use a synchronous GUI Slate playground without test-owned readiness or render-loop warmup
• keep the GUI gradient cases quarantined with precise bgfx Canvas limitation reasons

Why

The default 2.5% screenshot budget permits roughly 6,000 differing pixels in a 600x400 image. That is enough to hide missing or materially underdrawn controls and compact material differences. The tighter per-test budgets preserve normal anti-aliasing/color tolerance while making those regressions actionable.

The three GUI gradient cases are not tolerance problems on bgfx. Forced Metal/bgfx runs show missing Canvas semantics for gradient stroke styles, two-circle radial geometry, and endpoint-space linear mapping, so they remain excluded rather than being marked passing with loose thresholds.

Dependencies

GUI Slate requires BabylonJS/Babylon.js#18709 to be present in BabylonNative before this PR merges. That fix coerces fractional native dynamic-texture dimensions before allocation.

This PR does not depend on #1762. The replacement Slate playground is synchronous and its configured 60 validation frames produce the same capture with and without the readiness pump.

Validation

• config JSON parse and git diff --check
• GUI Slate, Metal/bgfx: 1,708 of 240,000 pixels (0.712%) at threshold 25
• GUI Slate, NativeWebGPU: 0 pixels over threshold 25
• GUI Slate A/B with and without Preserve animation frames during screenshot readiness renders #1762: byte-identical captures
• Texture Repetition Standard, NativeWebGPU: 2 pixels over threshold 25
• Texture Repetition PBR, NativeWebGPU: 761 pixels (0.317%) over threshold 25

[Copilot]

Pull request overview

Note

Copilot was unable to run its full agentic suite in this review.

Updates the Playground visual regression configuration to better handle pixel-diff variability and to add new texture repetition baseline tests.

Changes:

• Added per-test pixel comparison tuning via threshold and/or errorRatio for several existing tests.
• Added two new “Texture Repetition” playground entries with reference images and engine exclusions.

[bkaradzic-microsoft]

Thanks for tightening the screenshot tolerances — the per-test threshold/errorRatio approach is the right direction. I ran the changed/added tests locally against a Playground build (D3D11 + OpenGL) and hit blocking issues on the re-enabled and new tests. Two are structural and independent of Babylon version; a few others I'd like you to confirm pass in your CI, since my local build may differ from your base.

Blocking:

1. New Texture Repetition reference images are 24-bit RGB, but every existing reference is 32-bit RGBA. The harness compare() requires equal decoded-buffer lengths, so both new tests throw an uncaught Reference data length (720000) must match render data length (960000) and exit 1 — they can never pass on Babylon Native. Please re-export both PNGs as RGBA (ideally captured from Native's own output rather than the web playground).
2. Un-excluding OpenGL for GUI Near Menu re-introduces a hard crash. On the OpenGL build it hits BGFX FATAL: Failed to compile shader ... '=' : cannot convert from 'highp float' to 'bool' (renderer_gl.cpp:6956) — exactly the failure the reason you removed described. This aborts the entire OpenGL test run. The OpenGL exclusion should stay (e.g. excludedGraphicsApis: ["OpenGL"]) until the GUI fragment-shader codegen issue is fixed.

Please verify (may be environment-specific on my end):

3. Re-enabled GUI Gradient Radial / Linear with transparency don't crash/hang here (good — that reason is stale), but they diff ~60k/~102k pixels, far above errorRatio: 1.0 (~1%). GUI Gradient Linear throws TypeError: A string was expected, and GUI Slate fails to evaluate with The data size does not match width, height, and format. Can you confirm these actually pass in your run?

Minor: excludedEngines and dependsOn on the new entries aren't implemented by validation_native.js (only excludedGraphicsApis, onlyVisual, excludeFromAutomaticTesting are honored) — they'll be silently ignored, giving a false sense of gating.

[bkaradzic-microsoft]

Heads-up on CI: this branch is currently ~38 commits behind master (last pushed 2026-06-03), and there's no GitHub Actions check-suite on the head commit — so there's no pending workflow run for a maintainer to approve, which is why CI isn't showing up here.

Could you please rebase onto master and push? That synchronize event will create a fresh CI run that we can approve, and it'll also pick up the current workflow set. Thanks!

[matthargett]

Rebased onto current master (fcfda740) and force-pushed. I also removed the unsupported excludedEngines / dependsOn fields from the two native test entries. The two GUI verification threads remain open while I reconcile the reported runtime failures.