diff --git a/ClockReceiver/Dispatcher.hpp b/ClockReceiver/Dispatcher.hpp
index 460f2aa8a..98fec3cd4 100644
--- a/ClockReceiver/Dispatcher.hpp
+++ b/ClockReceiver/Dispatcher.hpp
@@ -11,7 +11,7 @@
 
 namespace Dispatcher {
 
-template <int max, typename SequencerT>
+template <typename SequencerT>
 struct Dispatcher {
 
 	/// Perform @c target.perform<n>() for the input range `start <= n < end`.
@@ -21,7 +21,7 @@ struct Dispatcher {
 		// Minor optimisation: do a comparison with end once outside the loop and if it implies so
 		// then do no further comparisons within the loop. This is somewhat targetted at expected
 		// use cases.
-		if(end < max) {
+		if(end < SequencerT::max) {
 			dispatch<true>(target, start, end, args...);
 		} else {
 			dispatch<false>(target, start, end, args...);
@@ -30,7 +30,7 @@ struct Dispatcher {
 
 private:
 	template <bool use_end, typename... Args> void dispatch(SequencerT &target, int start, int end, Args&&... args) {
-		static_assert(max < 2048);
+		static_assert(SequencerT::max < 2048);
 
 		// Yes, macros, yuck. But I want an actual switch statement for the dispatch to start
 		// and to allow a simple [[fallthrough]] through all subsequent steps up until end.
@@ -38,15 +38,15 @@ private:
 		//
 		// Sensible choices by the optimiser are assumed.
 
-#define index(n)									\
-	case n:											\
-		if constexpr (n <= max) {					\
-			if constexpr (n == max) return;			\
-			if constexpr (n < max) {				\
-				if(use_end && end == n) return;		\
-			}										\
-			target.template perform<n>(args...);	\
-		}											\
+#define index(n)										\
+	case n:												\
+		if constexpr (n <= SequencerT::max) {			\
+			if constexpr (n == SequencerT::max) return;	\
+			if constexpr (n < SequencerT::max) {		\
+				if(use_end && end == n) return;			\
+			}											\
+			target.template perform<n>(args...);		\
+		}												\
 	[[fallthrough]];
 
 #define index2(n)		index(n);		index(n+1);