fix: fix blown out water specular

[doodlum]

Fix specular brightness being overblown because specular mult in vanilla records can be over 2000

Summary by CodeRabbit

• Bug Fixes
  • Refined water shading calculations to correct sun reflection intensity and how deep-water color attenuates with depth. Results include more consistent linear lighting, improved normalization of sunlight contribution, and richer, more physically plausible deep-water coloration across viewing conditions.

[coderabbitai]

📝 Walkthrough

Walkthrough

GetSunColor in the Water shader was modified: SunDir.w is now applied to SunColor before scaling, DeepColor.w is replaced by (1.0 - exp(-DeepColor.w)) in the composition, and llDirLightMult remains applied to the final directional-light term. (39 words)

Changes

Cohort / File(s)	Summary
Water Shader Lighting package/Shaders/Water.hlsl	Modified GetSunColor: directional light input changed from SunColor.xyz / max(llDirLightMult, 1e-5) to (SunColor.xyz * SunDir.w) / max(llDirLightMult, 1e-5); replaced direct DeepColor.w multiplier with (1.0 - exp(-DeepColor.w)); final expression multiplies by llDirLightMult after the directional term.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~10 minutes

Suggested reviewers

• jiayev
• alandtse

Poem

🐰💦 I hopped into shaders at twilight's bend,
I tucked SunDir.w where the sunbeams blend,
Deep color softened by an exponential friend,
Waves now shimmer — rabbit-tested, code-penned.

🚥 Pre-merge checks | ✅ 3

✅ Passed checks (3 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title 'fix: fix blown out water specular' directly describes the main change—addressing overblown water specular brightness by tonemapping the calculation, which aligns with the shader modifications and PR objectives.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment
• Commit unit tests in branch water-brightnes-fix

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[github-actions]

Using provided base ref: 299717a
Using base ref: 299717a
Base commit date: 2026-02-09T19:39:49Z (Monday, February 09, 2026 07:39 PM)
No actionable suggestions for changed features.

[github-actions]

✅ A pre-release build is available for this PR:
Download

[coderabbitai]

Actionable comments posted: 1

🤖 Fix all issues with AI agents

In `@package/Shaders/Water.hlsl`:
- Line 1099: The tonemap expression using (1.0 - exp(-DeepColor.w)) compresses
moderate DeepColor.w values and dims specular; replace it with a scaled variant
such as (1.0 - exp(-DeepColor.w / SpecularTonemapK)) where SpecularTonemapK is a
tunable constant (e.g., 8–16) so normal water (0–2 range) preserves previous
brightness while still clamping extremes, and add the missing space before the
final *; update the expression involving reflectionMul, Color::DirectionalLight,
DeepColor.w and llDirLightMult to use the new scaled function/constant and
ensure formatting includes a space before the last multiplication.

🧹 Nitpick comments (1)

package/Shaders/Water.hlsl (1)

1087-1101: PR title and references.

Per the coding guidelines for conventional commits, the current title "fix: fix blown out water specular" has a redundant "fix" in the description. A more descriptive title would be:

fix(water): tonemap specular to prevent overblown highlights

If there's a related issue tracking this bug, consider adding a Fixes #NNN`` reference to the PR description.

[coderabbitai]

⚠️ Potential issue | 🟡 Minor

Tonemap may unintentionally dim specular for moderate DeepColor.w values.

1.0 - exp(-x) asymptotes to 1.0, which fixes the blown-out case (x > 2000), but it also compresses moderate values: at DeepColor.w = 1.0 the result is ~0.632 (a ~37% reduction vs. the previous direct multiply), and at DeepColor.w = 0.5 it's ~0.393 (a ~21% reduction). If vanilla records commonly use values in the 0–2 range for normal water, this would noticeably reduce their specular brightness compared to before.

Was this dimming for "normal" records evaluated, or is the intent purely to tame the extreme outliers? A scaled variant like 1.0 - exp(-DeepColor.w / k) with an appropriate k could shift the knee point to preserve existing behavior for typical values while still clamping the extremes.

Also, minor formatting nit — missing space before the last *:

-	return reflectionMul * Color::DirectionalLight((SunColor.xyz * SunDir.w) / max(llDirLightMult, 1e-5), SharedData::linearLightingSettings.isDirLightLinear) * (1.0 - exp(-DeepColor.w))* llDirLightMult;
+	return reflectionMul * Color::DirectionalLight((SunColor.xyz * SunDir.w) / max(llDirLightMult, 1e-5), SharedData::linearLightingSettings.isDirLightLinear) * (1.0 - exp(-DeepColor.w)) * llDirLightMult;

📝 Committable suggestion

‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.

Suggested change

	return reflectionMul * Color::DirectionalLight((SunColor.xyz * SunDir.w) / max(llDirLightMult, 1e-5), SharedData::linearLightingSettings.isDirLightLinear) * (1.0 - exp(-DeepColor.w))* llDirLightMult;
	return reflectionMul * Color::DirectionalLight((SunColor.xyz * SunDir.w) / max(llDirLightMult, 1e-5), SharedData::linearLightingSettings.isDirLightLinear) * (1.0 - exp(-DeepColor.w)) * llDirLightMult;

🤖 Prompt for AI Agents

In `@package/Shaders/Water.hlsl` at line 1099, The tonemap expression using (1.0 -
exp(-DeepColor.w)) compresses moderate DeepColor.w values and dims specular;
replace it with a scaled variant such as (1.0 - exp(-DeepColor.w /
SpecularTonemapK)) where SpecularTonemapK is a tunable constant (e.g., 8–16) so
normal water (0–2 range) preserves previous brightness while still clamping
extremes, and add the missing space before the final *; update the expression
involving reflectionMul, Color::DirectionalLight, DeepColor.w and llDirLightMult
to use the new scaled function/constant and ensure formatting includes a space
before the last multiplication.