diff --git a/OSBindings/Mac/Clock Signal/Wrappers/CSMachine.h b/OSBindings/Mac/Clock Signal/Wrappers/CSMachine.h
index c91487b7f..332328f1c 100644
--- a/OSBindings/Mac/Clock Signal/Wrappers/CSMachine.h	
+++ b/OSBindings/Mac/Clock Signal/Wrappers/CSMachine.h	
@@ -14,8 +14,8 @@
 
 - (void)runForNumberOfCycles:(int)numberOfCycles;
 
-- (int)idealSamplingRateFromRange:(NSRange)range;
-- (void)setAudioSamplingRate:(int)samplingRate;
+- (float)idealSamplingRateFromRange:(NSRange)range;
+- (void)setAudioSamplingRate:(float)samplingRate;
 
 - (void)setView:(CSOpenGLView *)view aspectRatio:(float)aspectRatio;
 - (void)drawViewForPixelSize:(CGSize)pixelSize onlyIfDirty:(BOOL)onlyIfDirty;
diff --git a/OSBindings/Mac/Clock Signal/Wrappers/CSMachine.mm b/OSBindings/Mac/Clock Signal/Wrappers/CSMachine.mm
index e3ba9ada8..188abe561 100644
--- a/OSBindings/Mac/Clock Signal/Wrappers/CSMachine.mm	
+++ b/OSBindings/Mac/Clock Signal/Wrappers/CSMachine.mm	
@@ -36,25 +36,25 @@ struct SpeakerDelegate: public Outputs::Speaker::Delegate {
 	}];
 }
 
-- (int)idealSamplingRateFromRange:(NSRange)range {
+- (float)idealSamplingRateFromRange:(NSRange)range {
 	@synchronized(self) {
 		Outputs::Speaker *speaker = self.machine->get_speaker();
 		if(speaker)
 		{
-			return speaker->get_ideal_clock_rate_in_range((int)range.location, (int)(range.location + range.length));
+			return speaker->get_ideal_clock_rate_in_range((float)range.location, (float)(range.location + range.length));
 		}
 		return 0;
 	}
 }
 
-- (void)setAudioSamplingRate:(int)samplingRate {
+- (void)setAudioSamplingRate:(float)samplingRate {
 	@synchronized(self) {
 		_speakerDelegate.machine = self;
 		[self setSpeakerDelegate:&_speakerDelegate sampleRate:samplingRate];
 	}
 }
 
-- (BOOL)setSpeakerDelegate:(Outputs::Speaker::Delegate *)delegate sampleRate:(int)sampleRate {
+- (BOOL)setSpeakerDelegate:(Outputs::Speaker::Delegate *)delegate sampleRate:(float)sampleRate {
 	@synchronized(self) {
 		Outputs::Speaker *speaker = self.machine->get_speaker();
 		if(speaker)
diff --git a/Outputs/Speaker.hpp b/Outputs/Speaker.hpp
index 9742e86df..6cfd434a2 100644
--- a/Outputs/Speaker.hpp
+++ b/Outputs/Speaker.hpp
@@ -24,7 +24,7 @@ class Speaker {
 				virtual void speaker_did_complete_samples(Speaker *speaker, const int16_t *buffer, int buffer_size) = 0;
 		};
 
-		int get_ideal_clock_rate_in_range(int minimum, int maximum)
+		float get_ideal_clock_rate_in_range(float minimum, float maximum)
 		{
 			// return exactly the input rate if possible
 			if(_input_cycles_per_second >= minimum && _input_cycles_per_second <= maximum) return _input_cycles_per_second;
@@ -36,7 +36,7 @@ class Speaker {
 			return maximum;
 		}
 
-		void set_output_rate(int cycles_per_second, int buffer_size)
+		void set_output_rate(float cycles_per_second, int buffer_size)
 		{
 			_output_cycles_per_second = cycles_per_second;
 			if(_buffer_size != buffer_size)
@@ -58,7 +58,7 @@ class Speaker {
 			_delegate = delegate;
 		}
 
-		void set_input_rate(int cycles_per_second)
+		void set_input_rate(float cycles_per_second)
 		{
 			_input_cycles_per_second = cycles_per_second;
 			set_needs_updated_filter_coefficients();
@@ -74,7 +74,7 @@ class Speaker {
 		bool _coefficients_are_dirty;
 		Delegate *_delegate;
 
-		int _input_cycles_per_second, _output_cycles_per_second;
+		float _input_cycles_per_second, _output_cycles_per_second;
 
 		void set_needs_updated_filter_coefficients()
 		{
@@ -181,7 +181,7 @@ template <class T> class Filter: public Speaker {
 			}
 			else
 			{
-				_number_of_taps = (_input_cycles_per_second + _output_cycles_per_second) / _output_cycles_per_second;
+				_number_of_taps = (int)ceilf((_input_cycles_per_second + _output_cycles_per_second) / _output_cycles_per_second);
 				_number_of_taps *= 2;
 				_number_of_taps |= 1;
 			}
@@ -190,7 +190,7 @@ template <class T> class Filter: public Speaker {
 			_buffer_in_progress_pointer = 0;
 
 			_stepper = std::unique_ptr<SignalProcessing::Stepper>(new SignalProcessing::Stepper((uint64_t)_input_cycles_per_second, (uint64_t)_output_cycles_per_second));
-			_filter = std::unique_ptr<SignalProcessing::FIRFilter>(new SignalProcessing::FIRFilter((unsigned int)_number_of_taps, (unsigned int)_input_cycles_per_second, 0.0, (float)_output_cycles_per_second / 2.0f, SignalProcessing::FIRFilter::DefaultAttenuation));
+			_filter = std::unique_ptr<SignalProcessing::FIRFilter>(new SignalProcessing::FIRFilter((unsigned int)_number_of_taps, (float)_input_cycles_per_second, 0.0, (float)_output_cycles_per_second / 2.0f, SignalProcessing::FIRFilter::DefaultAttenuation));
 
 			_input_buffer = std::unique_ptr<int16_t>(new int16_t[_number_of_taps]);
 			_input_buffer_depth = 0;