diff --git a/gltf/gltfpack.cpp b/gltf/gltfpack.cpp
index ab68c0fe..e6afb21c 100644
--- a/gltf/gltfpack.cpp
+++ b/gltf/gltfpack.cpp
@@ -250,6 +250,68 @@ static bool canTransformMesh(const Mesh& mesh)
 	return true;
 }
 
+static void detachMesh(Mesh& mesh, cgltf_data* data, const std::vector<NodeInfo>& nodes, const Settings& settings)
+{
+	// mesh is already instanced, skip
+	if (!mesh.instances.empty())
+		return;
+
+	// mesh is already world space, skip
+	if (mesh.nodes.empty())
+		return;
+
+	// note: when -kn is specified, we keep mesh-node attachment so that named nodes can be transformed
+	if (settings.keep_nodes)
+		return;
+
+	// we keep skinned meshes or meshes with morph targets as is
+	// in theory we could transform both, but in practice transforming morph target meshes is more involved,
+	// and reparenting skinned meshes leads to incorrect bounding box generated in three.js
+	if (mesh.skin || mesh.targets)
+		return;
+
+	bool any_animated = false;
+	for (size_t j = 0; j < mesh.nodes.size(); ++j)
+		any_animated |= nodes[mesh.nodes[j] - data->nodes].animated;
+
+	// animated meshes will be anchored to the same node that they used to be in to retain the animation
+	if (any_animated)
+		return;
+
+	int scene = nodes[mesh.nodes[0] - data->nodes].scene;
+	bool any_other_scene = false;
+	for (size_t j = 0; j < mesh.nodes.size(); ++j)
+		any_other_scene |= scene != nodes[mesh.nodes[j] - data->nodes].scene;
+
+	// we only merge instances when all nodes have a single consistent scene
+	if (scene < 0 || any_other_scene)
+		return;
+
+	// we only merge multiple instances together if requested
+	// this often makes the scenes faster to render by reducing the draw call count, but can result in larger files
+	if (mesh.nodes.size() > 1 && !settings.mesh_merge && !settings.mesh_instancing)
+		return;
+
+	// prefer instancing if possible, use merging otherwise
+	if (mesh.nodes.size() > 1 && settings.mesh_instancing)
+	{
+		mesh.instances.resize(mesh.nodes.size());
+
+		for (size_t j = 0; j < mesh.nodes.size(); ++j)
+			cgltf_node_transform_world(mesh.nodes[j], mesh.instances[j].data);
+
+		mesh.nodes.clear();
+		mesh.scene = scene;
+	}
+	else if (canTransformMesh(mesh))
+	{
+		mergeMeshInstances(mesh);
+
+		assert(mesh.nodes.empty());
+		mesh.scene = scene;
+	}
+}
+
 static bool isExtensionSupported(const ExtensionInfo* extensions, size_t count, const char* name)
 {
 	for (size_t i = 0; i < count; ++i)
@@ -269,9 +331,7 @@ static void process(cgltf_data* data, const char* input_path, const char* output
 	}
 
 	for (size_t i = 0; i < animations.size(); ++i)
-	{
 		processAnimation(animations[i], settings);
-	}
 
 	std::vector<NodeInfo> nodes(data->nodes_count);
 
@@ -279,68 +339,7 @@ static void process(cgltf_data* data, const char* input_path, const char* output
 	markAnimated(data, nodes, animations);
 
 	for (size_t i = 0; i < meshes.size(); ++i)
-	{
-		Mesh& mesh = meshes[i];
-
-		// mesh is already instanced, skip
-		if (!mesh.instances.empty())
-			continue;
-
-		// mesh is already world space, skip
-		if (mesh.nodes.empty())
-			continue;
-
-		// note: when -kn is specified, we keep mesh-node attachment so that named nodes can be transformed
-		if (settings.keep_nodes)
-			continue;
-
-		// we keep skinned meshes or meshes with morph targets as is
-		// in theory we could transform both, but in practice transforming morph target meshes is more involved,
-		// and reparenting skinned meshes leads to incorrect bounding box generated in three.js
-		if (mesh.skin || mesh.targets)
-			continue;
-
-		bool any_animated = false;
-		for (size_t j = 0; j < mesh.nodes.size(); ++j)
-			any_animated |= nodes[mesh.nodes[j] - data->nodes].animated;
-
-		// animated meshes will be anchored to the same node that they used to be in to retain the animation
-		if (any_animated)
-			continue;
-
-		int scene = nodes[mesh.nodes[0] - data->nodes].scene;
-		bool any_other_scene = false;
-		for (size_t j = 0; j < mesh.nodes.size(); ++j)
-			any_other_scene |= scene != nodes[mesh.nodes[j] - data->nodes].scene;
-
-		// we only merge instances when all nodes have a single consistent scene
-		if (scene < 0 || any_other_scene)
-			continue;
-
-		// we only merge multiple instances together if requested
-		// this often makes the scenes faster to render by reducing the draw call count, but can result in larger files
-		if (mesh.nodes.size() > 1 && !settings.mesh_merge && !settings.mesh_instancing)
-			continue;
-
-		// prefer instancing if possible, use merging otherwise
-		if (mesh.nodes.size() > 1 && settings.mesh_instancing)
-		{
-			mesh.instances.resize(mesh.nodes.size());
-
-			for (size_t j = 0; j < mesh.nodes.size(); ++j)
-				cgltf_node_transform_world(mesh.nodes[j], mesh.instances[j].data);
-
-			mesh.nodes.clear();
-			mesh.scene = scene;
-		}
-		else if (canTransformMesh(mesh))
-		{
-			mergeMeshInstances(mesh);
-
-			assert(mesh.nodes.empty());
-			mesh.scene = scene;
-		}
-	}
+		detachMesh(meshes[i], data, nodes, settings);
 
 	// material information is required for mesh and image processing
 	std::vector<MaterialInfo> materials(data->materials_count);
@@ -406,9 +405,7 @@ static void process(cgltf_data* data, const char* input_path, const char* output
 #endif
 
 	for (size_t i = 0; i < meshes.size(); ++i)
-	{
 		processMesh(meshes[i], settings);
-	}
 
 #ifndef NDEBUG
 	meshes.insert(meshes.end(), debug_meshes.begin(), debug_meshes.end());