From 580f9e05cbcdc723c59469d6828f2ceae66b85f3 Mon Sep 17 00:00:00 2001
From: Joshua Ashton <joshua@froggi.es>
Date: Tue, 9 May 2023 02:52:35 +0000
Subject: [PATCH] color_helpers: Optimize pq_to_nits + nits_to_pq

Make these take in vector types, saves ~0.4ms
---
 src/color_helpers.cpp |  4 ++--
 src/color_helpers.h   | 20 +++++++++++---------
 2 files changed, 13 insertions(+), 11 deletions(-)

diff --git a/src/color_helpers.cpp b/src/color_helpers.cpp
index 15dbf04..b860b36 100644
--- a/src/color_helpers.cpp
+++ b/src/color_helpers.cpp
@@ -483,7 +483,7 @@ glm::vec3 calcEOTFToLinear( const glm::vec3 & input, EOTF eotf, const tonemappin
     }
     else if ( eotf == EOTF_PQ )
     {
-        return glm::vec3( pq_to_nits( input.r ),  pq_to_nits( input.g ), pq_to_nits( input.b ) );
+        return pq_to_nits( input );
     }
 
     return glm::vec3(0);
@@ -502,7 +502,7 @@ glm::vec3 calcLinearToEOTF( const glm::vec3 & input, EOTF eotf, const tonemappin
     }
     else if ( eotf == EOTF_PQ )
     {
-        return glm::vec3( nits_to_pq( input.r ), nits_to_pq( input.g ), nits_to_pq( input.b ) );
+        return glm::vec3( nits_to_pq(input) );
     }
 
     return glm::vec3(0);
diff --git a/src/color_helpers.h b/src/color_helpers.h
index a33f26d..8c2ec25 100644
--- a/src/color_helpers.h
+++ b/src/color_helpers.h
@@ -45,7 +45,8 @@ inline float linear_to_srgb( float fVal )
     return ( fVal < 0.0031308f ) ? fVal * 12.92f : std::pow( fVal, 1.0f / 2.4f ) * 1.055f - 0.055f;
 }
 
-inline float pq_to_nits( float pq )
+template <typename T>
+inline T pq_to_nits( const T& pq )
 {
     const float c1 = 0.8359375f;
     const float c2 = 18.8515625f;
@@ -54,23 +55,24 @@ inline float pq_to_nits( float pq )
     const float oo_m1 = 1.0f / 0.1593017578125f;
     const float oo_m2 = 1.0f / 78.84375f;
 
-    float num = std::max(std::pow(pq, oo_m2) - c1, 0.0f);
-    float den = c2 - c3 * std::pow(pq, oo_m2);
+    T num = glm::max(glm::pow(pq, T(oo_m2)) - c1, T(0.0f));
+    T den = c2 - c3 * glm::pow(pq, T(oo_m2));
 
-    return 10000.0f * std::pow(num / den, oo_m1);
+    return glm::pow(num / den, T(oo_m1)) * 10000.0f;
 }
 
-inline float nits_to_pq( float nits )
+template <typename T>
+inline T nits_to_pq( const T& nits )
 {
-    float y = clamp(nits / 10000.0f, 0.0f, 1.0f);
+    T y = glm::clamp(nits / 10000.0f, T(0.0f), T(1.0f));
     const float c1 = 0.8359375f;
     const float c2 = 18.8515625f;
     const float c3 = 18.6875f;
     const float m1 = 0.1593017578125f;
     const float m2 = 78.84375f;
-    float num = c1 + c2 * std::pow(y, m1);
-    float den = 1.0 + c3 * std::pow(y, m1);
-    float n = std::pow(num / den, m2);
+    T num = c1 + c2 * glm::pow(y, T(m1));
+    T den = T(1.0) + c3 * glm::pow(y, T(m1));
+    T n = glm::pow(num / den, T(m2));
     return n;
 }