diff --git a/CMakeLists.txt b/CMakeLists.txt
index 1de16ae..dfd7357 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -240,9 +240,7 @@ endif()
 add_library(aare_compiler_flags INTERFACE)
 target_compile_features(aare_compiler_flags INTERFACE cxx_std_17)
 
-if(AARE_PYTHON_BINDINGS)
-    add_subdirectory(python)
-endif()
+
 
 #################
 # MSVC specific #
@@ -308,6 +306,9 @@ target_compile_options(
 
 endif() #GCC/Clang specific
 
+if(AARE_PYTHON_BINDINGS)
+    add_subdirectory(python)
+endif()
 
 if(AARE_ASAN)
     message(STATUS "AddressSanitizer enabled")
diff --git a/RELEASE.md b/RELEASE.md
index d9c0888..6ca6f44 100644
--- a/RELEASE.md
+++ b/RELEASE.md
@@ -16,6 +16,7 @@
 ### Bugfixes: 
 
  - multi threaded cluster finder doesnt drop frames if queues are full 
+ - Round before casting in the cluster finder to avoid biasing clusters by truncating
 
 
 ### 2025.11.21
diff --git a/include/aare/ClusterFinder.hpp b/include/aare/ClusterFinder.hpp
index 3f140a9..7d059ff 100644
--- a/include/aare/ClusterFinder.hpp
+++ b/include/aare/ClusterFinder.hpp
@@ -148,12 +148,27 @@ class ClusterFinder {
                         for (int ic = -dx; ic < dx + has_center_pixel_x; ic++) {
                             if (ix + ic >= 0 && ix + ic < frame.shape(1) &&
                                 iy + ir >= 0 && iy + ir < frame.shape(0)) {
-                                CT tmp =
-                                    static_cast<CT>(frame(iy + ir, ix + ic)) -
-                                    static_cast<CT>(
+
+                                // If the cluster type is an integral type, and
+                                // the pedestal is a floating point type then we
+                                // need to round the value before storing it
+                                if constexpr (std::is_integral_v<CT> &&
+                                              std::is_floating_point_v<
+                                                  PEDESTAL_TYPE>) {
+                                    auto tmp = std::lround(
+                                        frame(iy + ir, ix + ic) -
                                         m_pedestal.mean(iy + ir, ix + ic));
-                                cluster.data[i] =
-                                    tmp; // Watch for out of bounds access
+                                    cluster.data[i] = static_cast<CT>(tmp);
+                                }
+                                // On the other hand if both are floating point
+                                // or both are integral then we can just static
+                                // cast directly
+                                else {
+                                    auto tmp =
+                                        frame(iy + ir, ix + ic) -
+                                        m_pedestal.mean(iy + ir, ix + ic);
+                                    cluster.data[i] = static_cast<CT>(tmp);
+                                }
                             }
                             i++;
                         }
diff --git a/include/aare/ProducerConsumerQueue.hpp b/include/aare/ProducerConsumerQueue.hpp
index bf24c25..184ceb4 100644
--- a/include/aare/ProducerConsumerQueue.hpp
+++ b/include/aare/ProducerConsumerQueue.hpp
@@ -28,6 +28,7 @@
 #include <atomic>
 #include <cassert>
 #include <cstdlib>
+#include <cstddef>
 #include <memory>
 #include <stdexcept>
 #include <type_traits>