fix: util vertical fov math

[Dawntic]

fixes rubbish math for calculating vertical FOV.

Summary by CodeRabbit

• Refactor
  • Improved camera field-of-view calculation: now derives vertical FOV from a camera FOV value using explicit degree-to-radian conversion and trigonometric mapping, yielding more accurate and consistent vertical FOV across screen aspects.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

---

📝 Walkthrough

Walkthrough

GetVerticalFOVRad in src/Utils/Game.cpp was rewritten to compute vertical FOV by converting a camera FOV (degrees) to radians, using tan to obtain half-width, deriving half-height via the screen aspect, and returning 2 * atan(unitHalfHeight).

Changes

Cohort / File(s)	Summary
FOV Calculation Refactor src/Utils/Game.cpp	Replaced scalar/base-based vertical FOV computation with a deg→rad conversion: compute unitHalfWidth = tan(cameraFOVRad/2), derive unitHalfHeight = unitHalfWidth / screenAspect, and return 2 * atan(unitHalfHeight).

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~20 minutes

Suggested reviewers

• doodlum
• alandtse

Poem

🐇
I hopped through radians, degrees in tow,
Tangent whiskers twitching so,
Half-width, half-height, then back to view —
Two atans and the world feels new. ✨

🚥 Pre-merge checks | ✅ 2 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (2 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly and concisely summarizes the main change: fixing vertical FOV math in the utility code.

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

✨ Finishing Touches

• 📝 Generate docstrings

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment

---

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: 27eb9e9
Using base ref: 27eb9e9
Base commit date: 2026-02-18T09:49:12-08:00 (Wednesday, February 18, 2026 09:49 AM)
No actionable suggestions for changed features.

[coderabbitai]

Actionable comments posted: 1

🧹 Nitpick comments (1)

src/Utils/Game.cpp (1)

144-149: Math is correct — standard tan/atan perspective FOV conversion.

The formula vFOV = 2·atan(tan(hFOV/2) / aspect) is algebraically sound and correctly derives vertical FOV from the horizontal camera FOV and screen aspect ratio.

Minor nit: consider using XM_PI (already available via the DirectX Math headers in scope) instead of the magic literal 3.14159265359f, and std::tanf/std::atanf for explicit single-precision math — both are style-only.

♻️ Optional style cleanup

-        float hFOVRad = cameraFOVDeg * (3.14159265359f / 180.0f);
-        float unitHalfWidth = tan(hFOVRad / 2);
+        float hFOVRad = cameraFOVDeg * (XM_PI / 180.0f);
+        float unitHalfWidth = std::tanf(hFOVRad / 2.0f);
         float unitHalfHeight = unitHalfWidth / (globals::state->screenSize.x / globals::state->screenSize.y);
-        float vFOVRad = 2.0f * atan(unitHalfHeight);
+        float vFOVRad = 2.0f * std::atanf(unitHalfHeight);

Also, per the Conventional Commits coding guideline, consider adding a scope to the PR title: fix(util): vertical fov calculation.

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@src/Utils/Game.cpp` around lines 144 - 149, Replace the magic pi literal and
C math calls with the single-precision, symbolic versions: use XM_PI (from
DirectX Math) instead of 3.14159265359f and call std::tanf/std::atanf (or
tanf/atanf) for the trig operations; update the calculations around
cameraFOVDeg, hFOVRad, unitHalfWidth, unitHalfHeight, and vFOVRad in Game.cpp
accordingly to use XM_PI and single-precision trig functions to improve clarity
and type-safety.

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Inline comments:
In `@src/Utils/Game.cpp`:
- Around line 144-149: The FOV calculation in this function returns a too-narrow
vertical FOV in VR because it uses the combined stereo width; update the logic
in the method that reads cameraFOVDeg/ computes
hFOVRad/unitHalfWidth/unitHalfHeight/vFOVRad to branch on REL::Module::IsVR():
when VR is active use a per-eye width (globals::state->screenSize.x * 0.5f) for
the aspect ratio calculation, otherwise keep the existing
globals::state->screenSize.x; ensure you adjust the unitHalfHeight computation
accordingly and keep the final vFOVRad calculation the same, and also verify
that REL::RelocationID(513786, 388785) correctly returns the intended FOV value
in VR.

---

Nitpick comments:
In `@src/Utils/Game.cpp`:
- Around line 144-149: Replace the magic pi literal and C math calls with the
single-precision, symbolic versions: use XM_PI (from DirectX Math) instead of
3.14159265359f and call std::tanf/std::atanf (or tanf/atanf) for the trig
operations; update the calculations around cameraFOVDeg, hFOVRad, unitHalfWidth,
unitHalfHeight, and vFOVRad in Game.cpp accordingly to use XM_PI and
single-precision trig functions to improve clarity and type-safety.

[github-actions]

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