diff --git a/config/SOUE01/symbols.txt b/config/SOUE01/symbols.txt
index d0f907b6..6e601e2e 100644
--- a/config/SOUE01/symbols.txt
+++ b/config/SOUE01/symbols.txt
@@ -22199,7 +22199,7 @@ PSMTXTrans = .text:0x803B8770; // type:function size:0x34
 fn_803B87B0 = .text:0x803B87B0; // type:function size:0x4C
 fn_803B8800 = .text:0x803B8800; // type:function size:0x28
 fn_803B8830 = .text:0x803B8830; // type:function size:0x58
-fn_803B8890 = .text:0x803B8890; // type:function size:0xA4
+PSMTXQuat = .text:0x803B8890; // type:function size:0xA4
 fn_803B8940 = .text:0x803B8940; // type:function size:0x174
 fn_803B8AC0 = .text:0x803B8AC0; // type:function size:0xA4
 C_MTXLightPerspective = .text:0x803B8B70; // type:function size:0xF8
diff --git a/configure.py b/configure.py
index c7f04a70..b072e5f1 100644
--- a/configure.py
+++ b/configure.py
@@ -348,7 +348,7 @@ def nw4rLib(lib_name, objects, extra_cflags=[]):
             Object(NonMatching, "m/m3d/m_bmdl.cpp"),
             Object(Matching, "m/m3d/m_calc_ratio.cpp"),
             Object(NonMatching, "m/m3d/m_fanm.cpp"),
-            Object(NonMatching, "m/m3d/m_mdl.cpp"),
+            Object(Matching, "m/m3d/m_mdl.cpp"),
             Object(NonMatching, "m/m3d/m_scnLeaf.cpp"),
             Object(Matching, "m/m3d/m_smdl.cpp"),
             Object(Matching, "m/m_allocator.cpp"),
diff --git a/include/m/m3d/m_mdl.h b/include/m/m3d/m_mdl.h
index 66193ff7..d118e841 100644
--- a/include/m/m3d/m_mdl.h
+++ b/include/m/m3d/m_mdl.h
@@ -22,10 +22,7 @@ class callback_c {
 
 class mdl_c : public smdl_c {
 public:
-    // TODO this is not the correct callback class, since NSMBW has
-    // the A, B, C names and also the _CALC_X from IScnObjCallback,
-    // so there must be a different callback interface
-    class mdlCallback_c : public nw4r::g3d::IScnObjCallback {
+    class mdlCallback_c : public nw4r::g3d::ICalcWorldCallback {
     public:
         mdlCallback_c();
         virtual ~mdlCallback_c();
diff --git a/include/nw4r/g3d/g3d_calcworld.h b/include/nw4r/g3d/g3d_calcworld.h
index 7ea59a78..7c9ef69d 100644
--- a/include/nw4r/g3d/g3d_calcworld.h
+++ b/include/nw4r/g3d/g3d_calcworld.h
@@ -23,6 +23,17 @@ struct FuncObjCalcWorld {
 
 class WorldMtxManip {};
 
+// Name from ketteiban
+class ICalcWorldCallback {
+public:
+    virtual ~ICalcWorldCallback() {}
+
+    virtual void ExecCallbackA(nw4r::g3d::ChrAnmResult *, nw4r::g3d::ResMdl, nw4r::g3d::FuncObjCalcWorld *) {}
+    virtual void ExecCallbackB(nw4r::g3d::WorldMtxManip *, nw4r::g3d::ResMdl, nw4r::g3d::FuncObjCalcWorld *) {}
+    virtual void ExecCallbackC(nw4r::math::MTX34 *, nw4r::g3d::ResMdl, nw4r::g3d::FuncObjCalcWorld *) {}
+};
+
+
 void CalcWorld(math::MTX34 *, u32 *, const u8 *, const math::MTX34 *, ResMdl, AnmObjChr *, FuncObjCalcWorld *, u32);
 
 void CalcWorld(math::MTX34 *, u32 *, const u8 *, const math::MTX34 *, ResMdl, AnmObjChr *, FuncObjCalcWorld *);
diff --git a/include/nw4r/g3d/g3d_scnmdlsmpl.h b/include/nw4r/g3d/g3d_scnmdlsmpl.h
index f788a0e9..8082a4a2 100644
--- a/include/nw4r/g3d/g3d_scnmdlsmpl.h
+++ b/include/nw4r/g3d/g3d_scnmdlsmpl.h
@@ -4,6 +4,7 @@
 #include "nw4r/math/math_types.h"
 #include "nw4r/g3d/g3d_resmdl.h"
 #include "nw4r/g3d/g3d_scnobj.h"
+#include "nw4r/g3d/g3d_calcworld.h"
 
 namespace nw4r {
 namespace g3d {
@@ -75,6 +76,10 @@ class ScnMdlSimple : public ScnLeaf {
         return mNumViewMtx;
     }
 
+    void SetCalcWorldCallback(ICalcWorldCallback *cb) {
+        mpCalcWorldCallback = cb;
+    }
+
 private:
     ResMdl mResMdl;               // at 0xE8
     math::MTX34 *mWldMatrixArray; // at 0xEC
@@ -90,6 +95,8 @@ class ScnMdlSimple : public ScnLeaf {
     void *mByteCodeMix;     // at 0x10C
     void *mByteCodeDrawOpa; // at 0x110
     void *mByteCodeDrawXlu; // at 0x114
+    UNKWORD WORD_0x118;
+    ICalcWorldCallback *mpCalcWorldCallback; // at 0x11C
 
     NW4R_G3D_TYPE_OBJ_DECL(ScnMdlSimple);
 };
diff --git a/src/m/m3d/m_mdl.cpp b/src/m/m3d/m_mdl.cpp
index acd027e0..9d908857 100644
--- a/src/m/m3d/m_mdl.cpp
+++ b/src/m/m3d/m_mdl.cpp
@@ -14,7 +14,8 @@ mdl_c::mdlCallback_c::~mdlCallback_c() {}
 
 void mdl_c::mdlCallback_c::ExecCallbackA(nw4r::g3d::ChrAnmResult *result, nw4r::g3d::ResMdl mdl,
         nw4r::g3d::FuncObjCalcWorld *o) {
-    nw4r::g3d::ChrAnmResult *resPtr = &mpNodes[o->GetNodeId()];
+    u16 nodeId = o->GetNodeId();
+    nw4r::g3d::ChrAnmResult *resPtr = &mpNodes[nodeId];
     if (mCalcRatio.is0x18() && !mCalcRatio.isEnd()) {
         if (!mCalcRatio.is0x19()) {
             *result = *resPtr;
@@ -23,7 +24,8 @@ void mdl_c::mdlCallback_c::ExecCallbackA(nw4r::g3d::ChrAnmResult *result, nw4r::
             f32 f2 = mCalcRatio.get0x10();
             f32 f1 = mCalcRatio.get0x14();
 
-            // TODO the math operators cause ps inline assembly
+            // TODO clean up this code, what does it even do, why do operators
+            // break
             if ((flags & 8) == 0) {
                 result->VEC3_0x4.x = (result->VEC3_0x4.x * f1 + resPtr->VEC3_0x4.x * f2);
                 result->VEC3_0x4.y = (result->VEC3_0x4.y * f1 + resPtr->VEC3_0x4.y * f2);
@@ -36,8 +38,8 @@ void mdl_c::mdlCallback_c::ExecCallbackA(nw4r::g3d::ChrAnmResult *result, nw4r::
 
             nw4r::math::QUAT q1;
             nw4r::math::QUAT q2;
+
             C_QUATMtx(q1, resPtr->mMtx);
-            // TODO ...
             if ((flags & 0x20) == 0) {
                 C_QUATMtx(q2, result->mMtx);
             } else {
@@ -46,12 +48,28 @@ void mdl_c::mdlCallback_c::ExecCallbackA(nw4r::g3d::ChrAnmResult *result, nw4r::
                 q2.z = 0.0f;
                 q2.w = 1.0f;
             }
-            nw4r::math::QUAT q3;
-            C_QUATSlerp(q1, q2, q3, f1);
-            PSMTXQuat(result->mMtx, q3);
-            result->VEC3_0x10 = (flags & 0x40) == 0 ?
-                    nw4r::math::VEC3(result->VEC3_0x10.x * f1, result->VEC3_0x10.y * f1, result->VEC3_0x10.z * f1) :
-                    result->VEC3_0x10;
+
+            C_QUATSlerp(q1, q2, q1, f1);
+            // TODO weird hacks to force register order, probably more inlines
+            f32 tmp3, tmp2, tmp1;
+            tmp1 = result->mMtx._03;
+            tmp2 = result->mMtx._13;
+            tmp3 = result->mMtx._23;
+            nw4r::math::VEC3 tmp(tmp1, tmp2, tmp3);
+            PSMTXQuat(result->mMtx, q1);
+            result->mMtx._03 = tmp.x;
+            result->mMtx._13 = tmp.y;
+            result->mMtx._23 = tmp.z;
+            if ((flags & 0x40) == 0) {
+                result->mMtx._03 = tmp.x * f1;
+                result->mMtx._13 = tmp.y * f1;
+                result->mMtx._23 = tmp.z * f1;
+            }
+            u32 flags2 = result->mFlags & ~(0x80000000 | 0x80000040 | 0x80000020 | 0x80000008);
+            result->mMtx._03 += resPtr->mMtx._03 * f2;
+            result->mMtx._13 += resPtr->mMtx._13 * f2;
+            result->mMtx._23 += resPtr->mMtx._23 * f2;
+            result->mFlags = flags2;
             *resPtr = *result;
         }
     } else {
@@ -59,7 +77,7 @@ void mdl_c::mdlCallback_c::ExecCallbackA(nw4r::g3d::ChrAnmResult *result, nw4r::
     }
 
     if (mpBaseCallback != nullptr) {
-        mpBaseCallback->timingA(mNumNode, result, mdl);
+        mpBaseCallback->timingA(nodeId, result, mdl);
     }
 }
 
@@ -69,9 +87,11 @@ void mdl_c::mdlCallback_c::ExecCallbackB(nw4r::g3d::WorldMtxManip *m, nw4r::g3d:
     if (mpBaseCallback != nullptr) {
         mpBaseCallback->timingB(nodeId, m, mdl);
     }
+    // Not sure what this does
     u32 num = mdl.GetResNodeNumEntries();
-    u32 numInc = num + 1;
-    o->SetNodeId(numInc - (numInc / num) * num);
+    u32 nodeInc = nodeId + 1;
+    u32 tmp = (nodeInc / num);
+    o->SetNodeId(nodeInc - tmp * num);
 }
 
 void mdl_c::mdlCallback_c::ExecCallbackC(nw4r::math::MTX34 *mat, nw4r::g3d::ResMdl mdl, nw4r::g3d::FuncObjCalcWorld *) {
@@ -185,7 +205,7 @@ bool mdl_c::create(nw4r::g3d::ResMdl mdl, mdl_c::mdlCallback_c *cb, mAllocator_c
     }
     nw4r::g3d::ScnMdlSimple *sMdl;
     sMdl = nw4r::g3d::G3dObj::DynamicCast<nw4r::g3d::ScnMdlSimple>(mpScnLeaf);
-    sMdl->SetCallback(mpCallback);
+    sMdl->SetCalcWorldCallback(mpCallback);
     sMdl->EnableScnObjCallbackTiming(nw4r::g3d::ScnObj::TIMING_ALL);
     setCallback(nullptr);
     return true;