getPCMScale() too large?

Author: mbellew

src/libprojectM/Renderer/BeatDetect.cpp

@@ -109,6 +109,17 @@ void BeatDetect::detectFromSamples()
 }
 
 
+
+float BeatDetect::getPCMScale()
+{
+    // the constant here just depends on the particulars of getBeatVals(), the
+    // range of vol_history, and what "looks right".
+    // larger value means larger, more jagged waveform.
+    return 1.5 / fmax(0.0001f,sqrtf(vol_history));
+}
+
+
+
 void BeatDetect::getBeatVals( float samplerate, unsigned fft_length, float *vdataL, float *vdataR )
 {
     assert( 512==fft_length || 1024==fft_length );    // should be power of 2, expect >= 512

src/libprojectM/Renderer/BeatDetect.hpp

@@ -67,18 +67,7 @@ class DLLEXPORT BeatDetect
         // getPCMScale() was added to address https://github.com/projectM-visualizer/projectm/issues/161
         // Returning 1.0 results in using the raw PCM data, which can make the presets look pretty unresponsive
         // if the application volume is low.
-		float getPCMScale()
-		{
-			// the constant here just depends on the particulars of getBeatVals(), the
-			// range of vol_history, and what "looks right".
-			// larger value means larger, more jagged waveform.
-#ifdef WIN32
-// this is broken?
-#undef max
-			//work0around
-#endif /** WIN32 */
-			return 4.0f / std::max(0.0001f,sqrtf(vol_history));
-		}
+		float getPCMScale();
 
 	private:
 		int beat_buffer_pos;