shithub: cstory

Info  •  Files  •  Log  •  Branches

Change double to float

A quick test with godbolt shows this produces smaller (and presumably
faster) code.

--- a/src/Backends/Audio/SoftwareMixer.cpp

+++ b/src/Backends/Audio/SoftwareMixer.cpp

@@ -151,7 +151,7 @@

 			for (unsigned int frames_done = 0; frames_done < frames_total; ++frames_done)

 			{

 				const size_t position_integral = (size_t)sound->position;

-				const double position_fractional = sound->position - position_integral;

+				const float position_fractional = sound->position - position_integral;

 				// Get two samples, and normalise them to 0-1

 				const float sample1 = (sound->samples[position_integral] - 128.0f) / 128.0f;