diff --git a/include/core2/animctrl.h b/include/core2/animctrl.h
index 4147a512..34062cb2 100644
--- a/include/core2/animctrl.h
+++ b/include/core2/animctrl.h
@@ -3,6 +3,7 @@
 
 #include <ultra64.h>
 #include "generic.h"
+#include "ml/mtx.h"
 
 #ifndef NONMATCHING
 #define animctrl_start(this, file, line) _animctrl_start(this, file, line)
@@ -18,10 +19,10 @@ enum animctrl_playback_e{
 };
 
 typedef struct{
-    Mtx mtx_0;
+    MtxF mtx_0;
     s32 size_40;
     s32 capacity_44;
-    Mtx data[];
+    MtxF data[];
 }AnimMtxList;
 
 typedef struct animation_s{
diff --git a/include/ml/mtx.h b/include/ml/mtx.h
index dd179e4c..9ac1a288 100644
--- a/include/ml/mtx.h
+++ b/include/ml/mtx.h
@@ -1,11 +1,35 @@
 #ifndef __ML_MTX__
 #define __ML_MTX__
 
+typedef struct {
+    f32 m[4][4];
+} MtxF;
+
+void mlMtxGet(MtxF *dst);
+MtxF *mlMtx_get_stack_pointer(void);
+void mlMtxApply(Mtx *mPtr);
+void mlMtxPop(void);
+void mlMtx_push_duplicate(void);
+void mlMtx_push_identity(void);
+void mlMtx_push_translation(f32 x, f32 y, f32 z);
+void mlMtx_push_mtx(f32* mtx);
+void mlMtx_push_multiplied(f32* l_mtx);
+void mlMtx_push_multiplied_2(MtxF * l_mtx, MtxF * r_mtx);
 void mlMtxIdent(void);
-void mlMtxRotPitch(f32 pitch);
-void mlMtxRotYaw(f32 yaw);
-void mlMtxRotRoll(f32 roll);
-void mlMtxScale_xyz(f32 x, f32 y, f32);
+void mlMtxSet(MtxF* arg0);
+void mlMtxRotPitch(f32 arg0);
+void mlMtxRotYaw(f32 arg0);
+void mlMtxRotRoll(f32 arg0);
+void mlMtx_rotate_pitch_deg(f32 arg0);
+void mlMtx_rotate_yaw_deg(f32 arg0);
+void mlMtxRotatePYR(f32 pitch, f32 yaw, f32 roll);
+void mlMtxScale_xyz(f32 x, f32 y, f32 z);
+void mlMtxScale(f32 scale);
+void mlMtx_apply_f3(f32 dst[3], f32 x, f32 y, f32 z);
+void mlMtx_apply_vec3s(s16 dst[3], s16 src[3]);
 void mlMtxTranslate(f32 x, f32 y, f32 z);
 
+MtxF *func_8024DD90(void);
+MtxF *func_8024DD9C(void);
+
 #endif
diff --git a/src/FP/ch/scarfsled.c b/src/FP/ch/scarfsled.c
index 5be4d707..1a4f7d01 100644
--- a/src/FP/ch/scarfsled.c
+++ b/src/FP/ch/scarfsled.c
@@ -1,6 +1,7 @@
 #include <ultra64.h>
 #include "functions.h"
 #include "variables.h"
+#include "ml/mtx.h"
 
 void chScarfSled_update(Actor *this);
 
@@ -75,7 +76,7 @@ void __chScarfSled_func_80386630(Actor *this){
     f32 sp7C[3];
     f32 sp70[3];
     f32 sp64[3];
-    f32 sp24[4][4];
+    MtxF sp24;
 
     func_80343DEC(this);
     mapSpecificFlags_set(9, 1);
@@ -89,7 +90,7 @@ void __chScarfSled_func_80386630(Actor *this){
         mlMtxIdent();
         mlMtxRotYaw(sp64[1]);
         mlMtxRotPitch(sp64[0]);
-        mlMtxGet(sp24);
+        mlMtxGet(&sp24);
         
         sp70[0] = 0.0f;
         sp70[1] = 18.0f;
diff --git a/src/core1/code_13990.c b/src/core1/code_13990.c
index 05687f95..44288189 100644
--- a/src/core1/code_13990.c
+++ b/src/core1/code_13990.c
@@ -6,7 +6,7 @@
 
 #include "ml/mtx.h"
 
-void _guRotateF(Mtx*, f32, f32, f32, f32);
+void _guRotateF(f32 mf[4][4], f32, f32, f32, f32);
 f32 func_80263FF0(f32);
 f32 cosf(f32);
 
@@ -21,9 +21,8 @@ f32 D_80276578 = BAD_DTOR;
 f32 D_8027657C = BAD_DTOR;
 
 /*.bss*/
-Mtx_t D_80282810;
-u8 pad_D_80282850[0x780];
-Mtx * s_mtx_stack;
+MtxF D_80282810[0x1F];
+MtxF * s_mtx_stack;
 
 /*.code*/
 /**
@@ -31,17 +30,19 @@ Mtx * s_mtx_stack;
  * 
  * @param dst 
  */
-void mlMtxGet(f32 *dst) {
+void mlMtxGet(MtxF *dst) {
     s32 row, col;
 
     for(row = 0; row < 4; row++){
         for(col = 0; col < 4; col++){
-            *(dst++) = reinterpret_cast(f32, s_mtx_stack->m[row][col]);
+            dst->m[0][0] = s_mtx_stack->m[row][col];
+            // TODO figure out how to avoid this pointer arithmetic.
+            dst = (MtxF*)&dst->m[0][1];
         }
     }
 }
 
-Mtx *mlMtx_get_stack_pointer(void){
+MtxF *mlMtx_get_stack_pointer(void){
     return s_mtx_stack;
 }
 
@@ -49,23 +50,23 @@ void mlMtxApply(Mtx *mPtr){
     _guMtxF2L(s_mtx_stack, mPtr);
 }
 
-void func_802514BC(Mtx* arg0) {
+void func_802514BC(MtxF *arg0) {
     s32 row;
     s32 col;
     s32 i;
     f32 sum;
     f32 prod[4][4];
 
-    for(row = 0; row < 4; row++, arg0 = &arg0->m[1][0]) {
+    for(row = 0; row < 4; row++, arg0 = (MtxF*)&arg0->m[1][0]) {
         for(col = 0; col < 4; col++) {
             sum = 0.0;
             for(i = 0; i < 4; i++) {
-                sum += reinterpret_cast(f32, arg0->m[0][i]) * reinterpret_cast(f32, s_mtx_stack->m[i][col]);
+                sum += arg0->m[0][i] * s_mtx_stack->m[i][col];
             }
             prod[row][col] = sum;
         }
     }
-    func_80253010(s_mtx_stack, prod, sizeof(Mtx));
+    func_80253010(s_mtx_stack, prod, sizeof(MtxF));
 }
 
 void func_802515D4(f32 arg0[3][3]) {
@@ -79,7 +80,7 @@ void func_802515D4(f32 arg0[3][3]) {
         for(var_v1 = 0; var_v1 < 3; var_v1++){
             var_f0 = 0.0f;
             for(var_v0 = 0; var_v0 < 3; var_v0++){
-                var_f0 += arg0[i][var_v0] * reinterpret_cast(f32, s_mtx_stack->m[var_v0][var_v1]);
+                var_f0 += arg0[i][var_v0] * s_mtx_stack->m[var_v0][var_v1];
             }
             sp1C[i][var_v1] = var_f0;
         }
@@ -88,7 +89,7 @@ void func_802515D4(f32 arg0[3][3]) {
 
     for( i = 0; i < 3; i++){
         for(var_v1 = 0; var_v1 < 3; var_v1++){
-            reinterpret_cast(f32, s_mtx_stack->m[i][var_v1]) = sp1C[i][var_v1];
+            s_mtx_stack->m[i][var_v1] = sp1C[i][var_v1];
 
         }
     }
@@ -109,8 +110,8 @@ void mlMtx_push_duplicate(void) {
     f32 *var_a3;
     f32 *var_a2;
 
-    var_a2 = s_mtx_stack + 1;
-    var_a3 = s_mtx_stack;
+    var_a2 = &(s_mtx_stack + 1)->m[0][0];
+    var_a3 = &s_mtx_stack->m[0][0];
     for(i = 0; i < 16; i++) {
         var_a2[i] = var_a3[i];
     }
@@ -123,7 +124,7 @@ void mlMtx_push_duplicate(void) {
  */
 void mlMtx_push_identity(void){
     s32 i;
-    f32 *v0 = ++s_mtx_stack;
+    f32 *v0 = &(++s_mtx_stack)->m[0][0];
     for(i = 0; i<3; i++){
         v0[0] = 1.0f;
         v0[1] = 0.0f;
@@ -141,14 +142,13 @@ void mlMtx_push_identity(void){
  * @param x 
  * @param y 
  * @param z 
- * @return f32* 
  */
-f32 *mlMtx_push_translation(f32 x, f32 y, f32 z){
-    f32 * var_v0 = ++s_mtx_stack;
+void mlMtx_push_translation(f32 x, f32 y, f32 z){
+    f32 * var_v0 = &(++s_mtx_stack)->m[0][0];
     *(var_v0++) = 1.0f; *(var_v0++) = 0.0f; *(var_v0++) = 0.0f; *(var_v0++) = 0.0f;
     *(var_v0++) = 0.0f; *(var_v0++) = 1.0f; *(var_v0++) = 0.0f; *(var_v0++) = 0.0f;
     *(var_v0++) = 0.0f; *(var_v0++) = 0.0f; *(var_v0++) = 1.0f; *(var_v0++) = 0.0f;
-    *(var_v0++) = x; *(var_v0++) = y; *(var_v0++) = z; *(var_v0++) = 1.0f;
+    *(var_v0++) = x;    *(var_v0++) = y;    *(var_v0++) = z;    *(var_v0++) = 1.0f;
 }
 
 /**
@@ -162,7 +162,7 @@ void mlMtx_push_mtx(f32* mtx) {
     s_mtx_stack++;
     for(var_v0 = 0; var_v0 < 4; var_v0++){
         for(j = 0; j < 4; j++){
-            reinterpret_cast(f32, s_mtx_stack->m[var_v0][j]) = *mtx++;
+            s_mtx_stack->m[var_v0][j] = *mtx++;
         }
     }
 }
@@ -176,15 +176,15 @@ void mlMtx_push_mtx(f32* mtx) {
 void mlMtx_push_multiplied(f32* l_mtx) {
     s32 i;
     s32 j;
-    Mtx* var_a2;
+    MtxF* var_a2;
 
     var_a2 = s_mtx_stack + 1;
     for(i  = 0; i < 4; i++, l_mtx += 4){
         for(j = 0; j < 4; j++){
-            reinterpret_cast(f32, var_a2->m[i][j]) = l_mtx[0] * reinterpret_cast(f32, s_mtx_stack->m[0][j])
-                                                   + l_mtx[1] * reinterpret_cast(f32, s_mtx_stack->m[1][j])
-                                                   + l_mtx[2] * reinterpret_cast(f32, s_mtx_stack->m[2][j])
-                                                   + l_mtx[3] * reinterpret_cast(f32, s_mtx_stack->m[3][j]);
+            var_a2->m[i][j] = l_mtx[0] * s_mtx_stack->m[0][j]
+                            + l_mtx[1] * s_mtx_stack->m[1][j]
+                            + l_mtx[2] * s_mtx_stack->m[2][j]
+                            + l_mtx[3] * s_mtx_stack->m[3][j];
         }
     }
     s_mtx_stack = var_a2;
@@ -197,21 +197,21 @@ void mlMtx_push_multiplied(f32* l_mtx) {
  * @param l_mtx 
  * @param r_mtx 
  */
-void mlMtx_push_multiplied_2(Mtx * l_mtx, Mtx * r_mtx) {
+void mlMtx_push_multiplied_2(MtxF * l_mtx, MtxF * r_mtx) {
     s32 row;
     s32 col;
-    Mtx * dst = (s_mtx_stack + 1);
+    MtxF * dst = (s_mtx_stack + 1);
     
-    for (row = 0; row < 4; row++, r_mtx = &r_mtx->m[1][0])
+    for (row = 0; row < 4; row++, r_mtx = (MtxF*)&r_mtx->m[1][0])
     {
         for (col = 0; col < 4; col++)
         {
-            reinterpret_cast(f32, dst->m[row][col]) =
+            dst->m[row][col] =
             (
-                reinterpret_cast(f32, r_mtx->m[0][0]) * reinterpret_cast(f32, l_mtx->m[0][col]) +
-                reinterpret_cast(f32, r_mtx->m[0][1]) * reinterpret_cast(f32, l_mtx->m[1][col]) +
-                reinterpret_cast(f32, r_mtx->m[0][2]) * reinterpret_cast(f32, l_mtx->m[2][col]) +
-                reinterpret_cast(f32, r_mtx->m[0][3]) * reinterpret_cast(f32, l_mtx->m[3][col])
+                r_mtx->m[0][0] * l_mtx->m[0][col] +
+                r_mtx->m[0][1] * l_mtx->m[1][col] +
+                r_mtx->m[0][2] * l_mtx->m[2][col] +
+                r_mtx->m[0][3] * l_mtx->m[3][col]
             );
         }
     }
@@ -221,7 +221,7 @@ void mlMtx_push_multiplied_2(Mtx * l_mtx, Mtx * r_mtx) {
 //mlMtx
 void mlMtxIdent(void){
     s32 i;
-    f32 *v0 = s_mtx_stack = &D_80282810;
+    f32 *v0 = &(s_mtx_stack = &D_80282810[0])->m[0][0];
     for(i = 0; i<3; i++){
         v0[0] = 1.0f;
         v0[1] = 0.0f;
@@ -234,28 +234,31 @@ void mlMtxIdent(void){
 }
 
 void func_80251B5C(f32 x, f32 y, f32 z){
-    f32 * var_v0 = s_mtx_stack = &D_80282810;
+    f32 * var_v0 = &(s_mtx_stack = &D_80282810[0])->m[0][0];
     *(var_v0++) = 1.0f; *(var_v0++) = 0.0f; *(var_v0++) = 0.0f; *(var_v0++) = 0.0f;
     *(var_v0++) = 0.0f; *(var_v0++) = 1.0f; *(var_v0++) = 0.0f; *(var_v0++) = 0.0f;
     *(var_v0++) = 0.0f; *(var_v0++) = 0.0f; *(var_v0++) = 1.0f; *(var_v0++) = 0.0f;
     *(var_v0++) = x; *(var_v0++) = y; *(var_v0++) = z; *(var_v0++) = 1.0f;
 }
 
-void mlMtxSet(Mtx* arg0) {
+void mlMtxSet(MtxF* arg0) {
     s32 i, j;
-    f32 *v0;
+    MtxF *v0;
     
-    v0 = s_mtx_stack = &D_80282810;
+    v0 = s_mtx_stack = D_80282810;
     for(i = 0; i < 4*4; i+=4){
         for(j = 0; j < 4; j++){
-            *(v0++) = *(((f32*)arg0)++);
+            v0->m[0][0] = arg0->m[0][0];
+            // TODO figure out how to avoid this pointer arithmetic.
+            v0 = (MtxF*)(&v0->m[0][1]);
+            arg0 = (MtxF*)(&arg0->m[0][1]); 
         }
     }
 }
 
 void mlMtxRotate(f32 a, f32 x, f32 y, f32 z) {
-    _guRotateF(s_mtx_stack + 1, a, x, y, z);
-    guMtxCatF(s_mtx_stack + 1, s_mtx_stack, s_mtx_stack);
+    _guRotateF((s_mtx_stack + 1)->m, a, x, y, z);
+    guMtxCatF((s_mtx_stack + 1)->m, s_mtx_stack->m, s_mtx_stack->m);
 }
 
 void mlMtxRotPitch(f32 arg0) {
@@ -268,20 +271,20 @@ void mlMtxRotPitch(f32 arg0) {
         arg0 *= D_80276578;
         sin = sinf(arg0);
         cos = cosf(arg0);
-        var_f18 = reinterpret_cast(f32, s_mtx_stack->m[1][0]);
-        var_f10 = reinterpret_cast(f32, s_mtx_stack->m[2][0]);
-        reinterpret_cast(f32, s_mtx_stack->m[1][0]) = var_f18*cos + var_f10*sin;
-        reinterpret_cast(f32, s_mtx_stack->m[2][0]) = var_f18*-sin + var_f10*cos;
+        var_f18 = s_mtx_stack->m[1][0];
+        var_f10 = s_mtx_stack->m[2][0];
+        s_mtx_stack->m[1][0] = var_f18*cos + var_f10*sin;
+        s_mtx_stack->m[2][0] = var_f18*-sin + var_f10*cos;
 
-        var_f18 = reinterpret_cast(f32, s_mtx_stack->m[1][1]);
-        var_f10 = reinterpret_cast(f32, s_mtx_stack->m[2][1]);
-        reinterpret_cast(f32, s_mtx_stack->m[1][1]) = var_f18*cos + var_f10*sin;
-        reinterpret_cast(f32, s_mtx_stack->m[2][1]) = var_f18*-sin + var_f10*cos;
+        var_f18 = s_mtx_stack->m[1][1];
+        var_f10 = s_mtx_stack->m[2][1];
+        s_mtx_stack->m[1][1] = var_f18*cos + var_f10*sin;
+        s_mtx_stack->m[2][1] = var_f18*-sin + var_f10*cos;
 
-        var_f18 = reinterpret_cast(f32, s_mtx_stack->m[1][2]);
-        var_f10 = reinterpret_cast(f32, s_mtx_stack->m[2][2]);
-        reinterpret_cast(f32, s_mtx_stack->m[1][2]) = var_f18*cos + var_f10*sin;
-        reinterpret_cast(f32, s_mtx_stack->m[2][2]) = var_f18*-sin + var_f10*cos;
+        var_f18 = s_mtx_stack->m[1][2];
+        var_f10 = s_mtx_stack->m[2][2];
+        s_mtx_stack->m[1][2] = var_f18*cos + var_f10*sin;
+        s_mtx_stack->m[2][2] = var_f18*-sin + var_f10*cos;
     }
 }
 
@@ -297,10 +300,10 @@ void mlMtxRotYaw(f32 arg0) {
         sin = sinf(arg0);
         cos = cosf(arg0);
         for(i = 0; i < 3; i++){
-            var_f18 = reinterpret_cast(f32, s_mtx_stack->m[0][i]);
-            var_f10 = reinterpret_cast(f32, s_mtx_stack->m[2][i]);
-            reinterpret_cast(f32, s_mtx_stack->m[0][i]) = var_f18*cos - var_f10*sin;
-            reinterpret_cast(f32, s_mtx_stack->m[2][i]) = var_f18*sin + var_f10*cos;
+            var_f18 = s_mtx_stack->m[0][i];
+            var_f10 = s_mtx_stack->m[2][i];
+            s_mtx_stack->m[0][i] = var_f18*cos - var_f10*sin;
+            s_mtx_stack->m[2][i] = var_f18*sin + var_f10*cos;
         }
     }
 }
@@ -315,20 +318,20 @@ void mlMtxRotRoll(f32 arg0) {
         arg0 *= D_8027657C;
         sin = sinf(arg0);
         cos = cosf(arg0);
-        var_f18 = reinterpret_cast(f32, s_mtx_stack->m[0][0]);
-        var_f10 = reinterpret_cast(f32, s_mtx_stack->m[1][0]);
-        reinterpret_cast(f32, s_mtx_stack->m[0][0]) = var_f18*cos + var_f10*sin;
-        reinterpret_cast(f32, s_mtx_stack->m[1][0]) = var_f18*-sin + var_f10*cos;
+        var_f18 = s_mtx_stack->m[0][0];
+        var_f10 = s_mtx_stack->m[1][0];
+        s_mtx_stack->m[0][0] = var_f18*cos + var_f10*sin;
+        s_mtx_stack->m[1][0] = var_f18*-sin + var_f10*cos;
 
-        var_f18 = reinterpret_cast(f32, s_mtx_stack->m[0][1]);
-        var_f10 = reinterpret_cast(f32, s_mtx_stack->m[1][1]);
-        reinterpret_cast(f32, s_mtx_stack->m[0][1]) = var_f18*cos + var_f10*sin;
-        reinterpret_cast(f32, s_mtx_stack->m[1][1]) = var_f18*-sin + var_f10*cos;
+        var_f18 = s_mtx_stack->m[0][1];
+        var_f10 = s_mtx_stack->m[1][1];
+        s_mtx_stack->m[0][1] = var_f18*cos + var_f10*sin;
+        s_mtx_stack->m[1][1] = var_f18*-sin + var_f10*cos;
 
-        var_f18 = reinterpret_cast(f32, s_mtx_stack->m[0][2]);
-        var_f10 = reinterpret_cast(f32, s_mtx_stack->m[1][2]);
-        reinterpret_cast(f32, s_mtx_stack->m[0][2]) = var_f18*cos + var_f10*sin;
-        reinterpret_cast(f32, s_mtx_stack->m[1][2]) = var_f18*-sin + var_f10*cos;
+        var_f18 = s_mtx_stack->m[0][2];
+        var_f10 = s_mtx_stack->m[1][2];
+        s_mtx_stack->m[0][2] = var_f18*cos + var_f10*sin;
+        s_mtx_stack->m[1][2] = var_f18*-sin + var_f10*cos;
     }
 }
 
@@ -345,10 +348,10 @@ void mlMtx_rotate_pitch_deg(f32 arg0) {
         sin = sinf(arg0);
         cos = cosf(arg0);
         for(i = 0; i < 3; i++){
-            var_f18 = reinterpret_cast(f32, s_mtx_stack->m[1][i]);
-            var_f10 = reinterpret_cast(f32, s_mtx_stack->m[2][i]);
-            reinterpret_cast(f32, s_mtx_stack->m[1][i]) = var_f18*cos + var_f10*sin;
-            reinterpret_cast(f32, s_mtx_stack->m[2][i]) = var_f18*-sin +var_f10*cos;
+            var_f18 = s_mtx_stack->m[1][i];
+            var_f10 = s_mtx_stack->m[2][i];
+            s_mtx_stack->m[1][i] = var_f18*cos + var_f10*sin;
+            s_mtx_stack->m[2][i] = var_f18*-sin +var_f10*cos;
         }
     }
 }
@@ -365,10 +368,10 @@ void mlMtx_rotate_yaw_deg(f32 arg0) {
         sin = sinf(arg0);
         cos = cosf(arg0);
         for(i = 0; i < 3; i++){
-            var_f18 = reinterpret_cast(f32, s_mtx_stack->m[0][i]);
-            var_f10 = reinterpret_cast(f32, s_mtx_stack->m[2][i]);
-            reinterpret_cast(f32, s_mtx_stack->m[0][i]) = var_f18*cos - var_f10*sin;
-            reinterpret_cast(f32, s_mtx_stack->m[2][i]) = var_f18*sin + var_f10*cos;
+            var_f18 = s_mtx_stack->m[0][i];
+            var_f10 = s_mtx_stack->m[2][i];
+            s_mtx_stack->m[0][i] = var_f18*cos - var_f10*sin;
+            s_mtx_stack->m[2][i] = var_f18*sin + var_f10*cos;
         }
     }
 }
@@ -383,25 +386,25 @@ void mlMtxRotatePYR(f32 pitch, f32 yaw, f32 roll){
 void mlMtxScale_xyz(f32 x, f32 y, f32 z){
     int i;
     for(i = 0; i < 3; i++){
-        reinterpret_cast(f32, s_mtx_stack->m[0][i]) *= x;
-        reinterpret_cast(f32, s_mtx_stack->m[1][i]) *= y;
-        reinterpret_cast(f32, s_mtx_stack->m[2][i]) *= z;
+        s_mtx_stack->m[0][i] *= x;
+        s_mtx_stack->m[1][i] *= y;
+        s_mtx_stack->m[2][i] *= z;
     }
 }
 
 void mlMtxScale(f32 scale){
     int i;
     for(i = 0; i < 3; i++){
-        reinterpret_cast(f32, s_mtx_stack->m[0][i]) *= scale;
-        reinterpret_cast(f32, s_mtx_stack->m[1][i]) *= scale;
-        reinterpret_cast(f32, s_mtx_stack->m[2][i]) *= scale;
+        s_mtx_stack->m[0][i] *= scale;
+        s_mtx_stack->m[1][i] *= scale;
+        s_mtx_stack->m[2][i] *= scale;
     }
 }
 
 void func_80252330(f32 x, f32 y, f32 z){
-    reinterpret_cast(f32, s_mtx_stack->m[3][0]) = x;
-    reinterpret_cast(f32, s_mtx_stack->m[3][1]) = y;
-    reinterpret_cast(f32, s_mtx_stack->m[3][2]) = z;
+    s_mtx_stack->m[3][0] = x;
+    s_mtx_stack->m[3][1] = y;
+    s_mtx_stack->m[3][2] = z;
 }
 
 void func_8025235C(f32 arg0[3], f32 arg1[3]) {
@@ -413,10 +416,10 @@ void func_8025235C(f32 arg0[3], f32 arg1[3]) {
     sp0[2] = arg1[2];
     
     for(i = 0; i < 3; i++){
-        arg0[i] = sp0[0]*reinterpret_cast(f32, s_mtx_stack->m[0][i]) 
-                  + sp0[1]*reinterpret_cast(f32, s_mtx_stack->m[1][i]) 
-                  + sp0[2]*reinterpret_cast(f32, s_mtx_stack->m[2][i]) 
-                  + reinterpret_cast(f32, s_mtx_stack->m[3][i]);
+        arg0[i] = sp0[0]*s_mtx_stack->m[0][i] 
+                + sp0[1]*s_mtx_stack->m[1][i] 
+                + sp0[2]*s_mtx_stack->m[2][i] 
+                + s_mtx_stack->m[3][i];
     }
 }
 
@@ -424,17 +427,17 @@ void func_80252434(f32 arg0[3], f32 arg1[3]) {
     s32 i;
     
     for(i = 0; i < 3; i++){
-        arg0[i] = arg1[0]*reinterpret_cast(f32, s_mtx_stack->m[0][i]) 
-                  + arg1[1]*reinterpret_cast(f32, s_mtx_stack->m[1][i]) 
-                  + arg1[2]*reinterpret_cast(f32, s_mtx_stack->m[2][i]) 
-                  + reinterpret_cast(f32, s_mtx_stack->m[3][i]);
+        arg0[i] = arg1[0]*s_mtx_stack->m[0][i] 
+                + arg1[1]*s_mtx_stack->m[1][i] 
+                + arg1[2]*s_mtx_stack->m[2][i] 
+                + s_mtx_stack->m[3][i];
     }
 }
 
 void mlMtx_apply_f3(f32 dst[3], f32 x, f32 y, f32 z) {
-    dst[0] = x*((f32 (*)[4])s_mtx_stack)[0][0] + y*((f32 (*)[4])s_mtx_stack)[1][0] + z*((f32 (*)[4])s_mtx_stack)[2][0] + ((f32 (*)[4])s_mtx_stack)[3][0];
-    dst[1] = x*((f32 (*)[4])s_mtx_stack)[0][1] + y*((f32 (*)[4])s_mtx_stack)[1][1] + z*((f32 (*)[4])s_mtx_stack)[2][1] + ((f32 (*)[4])s_mtx_stack)[3][1];
-    dst[2] = x*((f32 (*)[4])s_mtx_stack)[0][2] + y*((f32 (*)[4])s_mtx_stack)[1][2] + z*((f32 (*)[4])s_mtx_stack)[2][2] + ((f32 (*)[4])s_mtx_stack)[3][2];
+    dst[0] = x* s_mtx_stack->m[0][0] + y*s_mtx_stack->m[1][0] + z*s_mtx_stack->m[2][0] + s_mtx_stack->m[3][0];
+    dst[1] = x* s_mtx_stack->m[0][1] + y*s_mtx_stack->m[1][1] + z*s_mtx_stack->m[2][1] + s_mtx_stack->m[3][1];
+    dst[2] = x* s_mtx_stack->m[0][2] + y*s_mtx_stack->m[1][2] + z*s_mtx_stack->m[2][2] + s_mtx_stack->m[3][2];
 }
 
 void mlMtx_apply_vec3s(s16 dst[3], s16 src[3]) {
@@ -442,7 +445,7 @@ void mlMtx_apply_vec3s(s16 dst[3], s16 src[3]) {
     f32 sp0[3];
     f32 (*temp_v0)[4];
 
-    temp_v0 = s_mtx_stack;
+    temp_v0 = &s_mtx_stack->m[0];
     sp0[0] = (f32) src[0];
     sp0[1] = (f32) src[1];
     sp0[2] = (f32) src[2];
@@ -489,19 +492,14 @@ void func_8025276C(s32 arg0[3], s32 arg1[3], s32 arg2[3], s32 arg3[3]) {
 }
 
 void mlMtxTranslate(f32 x, f32 y, f32 z) {
-    f32 *temp_t6;
-    f32 *phi_v0;
     f32 phi_f18;
     f32 phi_f16;
-    f32 phi_f10;
     s32 phi_v1;
 
-    phi_v0 = s_mtx_stack;
-    for(phi_v1 = 0; phi_v1 < 0xC; phi_v1 +=4){
-        phi_v0 = (u32)s_mtx_stack + phi_v1;
-        phi_f18 = phi_v0[0] * x;
-        phi_f16 = phi_v0[4] * y;
-        phi_v0[0xC] += phi_f18 + phi_f16 + (phi_v0[8] * z);
+    for(phi_v1 = 0; phi_v1 < 3; phi_v1++){
+        phi_f18 = s_mtx_stack->m[0][phi_v1] * x;
+        phi_f16 = s_mtx_stack->m[1][phi_v1] * y;
+        s_mtx_stack->m[3][phi_v1] += phi_f18 + phi_f16 + (s_mtx_stack->m[2][phi_v1] * z);
     }
 }
 
@@ -509,7 +507,7 @@ void func_80252A38(f32 x, f32 y, f32 z) {
     s32 var_v1;
 
     for(var_v1 = 0; var_v1 != 3; var_v1++){
-        reinterpret_cast(f32, s_mtx_stack->m[3][var_v1]) += reinterpret_cast(f32, s_mtx_stack->m[0][var_v1])*x + reinterpret_cast(f32, s_mtx_stack->m[1][var_v1])*y + reinterpret_cast(f32, s_mtx_stack->m[2][var_v1])*z;
+        s_mtx_stack->m[3][var_v1] += s_mtx_stack->m[0][var_v1]*x + s_mtx_stack->m[1][var_v1]*y + s_mtx_stack->m[2][var_v1]*z;
     }
 }
 
diff --git a/src/core1/code_EAF0.c b/src/core1/code_EAF0.c
index 5ea939c3..ce6ae4af 100644
--- a/src/core1/code_EAF0.c
+++ b/src/core1/code_EAF0.c
@@ -2,6 +2,8 @@
 #include "functions.h"
 #include "variables.h"
 
+#include "ml/mtx.h"
+
 extern void guPerspective(Mtx *, u16*, f32, f32, f32, f32, f32);
 extern f32 ml_dotProduct_vec3f(f32[3], f32[3]);
 
@@ -21,8 +23,8 @@ f32 D_80280ECC;
 f32 D_80280ED0[4][4];
 Vp s_viewport_stack[8];
 int D_80280F90;
-f32 D_80280F98[4][4];
-Mtx D_80280FD8;
+MtxF D_80280F98;
+MtxF D_80280FD8;
 s32 s_viewport_stack_index; //viewport indx
 
 
@@ -386,11 +388,11 @@ bool func_8024DD34(f32 arg0, f32 arg1, f32 arg2) {
     return ((D_80280ED0[2][0]*arg0 + D_80280ED0[2][1]*arg1 + D_80280ED0[2][2]*arg2 + D_80280ED0[2][3]) <= 0.0f);
 }
 
-Mtx *func_8024DD90(void){
+MtxF *func_8024DD90(void){
     return &D_80280F98;
 }
 
-Mtx *func_8024DD9C(void){
+MtxF *func_8024DD9C(void){
     return &D_80280FD8;
 }
 
@@ -450,7 +452,7 @@ f32 D_80281058[3];
 f32 D_80281068[3];
 f32 D_80281078[3];
 f32 D_80281088[3];
-f32 D_80281098[4][4];
+MtxF D_80281098;
 
 bool func_8024E030(f32 arg0[3], f32 *arg1)
 {
@@ -488,7 +490,7 @@ void func_8024E258(void){
     func_8024C5A8(D_80281088);
     for(i = 0; i < 4; i++){
         for(j = 0; j < 4; j++){
-            D_80281098[i][j] = D_80280F98[i][j];
+            D_80281098.m[i][j] = D_80280F98.m[i][j];
         }
     }
 }
@@ -501,7 +503,7 @@ void func_8024E2FC(void){
     ml_vec3f_copy(D_80280EA0, D_80281088);
     for(i = 0; i < 4; i++){
         for(j = 0; j < 4; j++){
-            D_80280F98[i][j] = D_80281098[i][j];
+            D_80280F98.m[i][j] = D_80281098.m[i][j];
         }
     }
 }
diff --git a/src/core2/ch/bubble.c b/src/core2/ch/bubble.c
index 1f2c8b8d..49b339a6 100644
--- a/src/core2/ch/bubble.c
+++ b/src/core2/ch/bubble.c
@@ -7,7 +7,7 @@ typedef struct {
 }ActorLocal_Core2_3ECE0;
 
 void chBubble_update(Actor *this);
-extern void func_802F32C4(s32, f32[3], f32, ActorMarker *, void(*)(f32[3], f32, ActorMarker *));
+extern void func_802F32C4(s32, f32[3], f32, ActorMarker *, s32(*)(f32[3], f32, ActorMarker *));
 
 /* .data */
 ActorInfo chBubble = { 
@@ -17,7 +17,7 @@ ActorInfo chBubble = {
 };
 
 /* .code */
-int __chBubble_func_802C5C70(f32 arg0[3], f32 arg1, ActorMarker *marker){
+s32 __chBubble_func_802C5C70(f32 arg0[3], f32 arg1, ActorMarker *marker){
     s32 pad2C;
     Actor *this = marker_getActor(marker);
     ActorLocal_Core2_3ECE0 *local = (ActorLocal_Core2_3ECE0 *) &this->local;
diff --git a/src/core2/code_630D0.c b/src/core2/code_630D0.c
index f8b1279d..5d5aef8c 100644
--- a/src/core2/code_630D0.c
+++ b/src/core2/code_630D0.c
@@ -2,11 +2,14 @@
 #include "functions.h"
 #include "variables.h"
 #include "animation.h"
+#include "ml/mtx.h"
+
+void *defrag(void *);
 
 void animMtxList_setBoneless(AnimMtxList **this_ptr, BKAnimationList *anim_list){
     AnimMtxList * this;
-    Mtx *end_ptr;
-    Mtx *i_ptr;
+    MtxF *end_ptr;
+    MtxF *i_ptr;
     
     this = *this_ptr;
     if(this->capacity_44 < anim_list->cnt_4){
@@ -16,14 +19,14 @@ void animMtxList_setBoneless(AnimMtxList **this_ptr, BKAnimationList *anim_list)
     }
 
     this->size_40 = anim_list->cnt_4;
-    end_ptr = (Mtx *)(this->size_40*sizeof(Mtx) + (s32)this +sizeof(AnimMtxList));
+    end_ptr = (MtxF *)(this->size_40*sizeof(MtxF) + (s32)this +sizeof(AnimMtxList));
     mlMtxIdent();
     for(i_ptr = this->data; i_ptr < end_ptr; i_ptr++){
         mlMtxGet(i_ptr);
     }
 }
 
-Mtx *animMtxList_get(AnimMtxList *this, s32 arg1){
+MtxF *animMtxList_get(AnimMtxList *this, s32 arg1){
     if (arg1 == -1){
         return &this->mtx_0;
     }
@@ -40,7 +43,7 @@ AnimMtxList *animMtxList_new(void){
     this->size_40 = 0;
     this->capacity_44 = 0;
     mlMtxIdent();
-    mlMtxGet(this);
+    mlMtxGet(&this->mtx_0);
     return this;
 }
 
@@ -50,14 +53,13 @@ s32 animMtxList_len(AnimMtxList* this){
     return 1;
 }
 
-void mlMtxSet(Mtx*);
-void func_8033A5B8(s32, s32, f32[4],f32[3], f32[3]);
+void func_8033A5B8(BoneTransformList *this, s32 bone_id, f32 arg2[4], f32 scale[3], f32 arg4[3]);
 
 void animMtxList_setBoned(AnimMtxList **this_ptr, BKAnimationList *anim_list, BoneTransformList *arg2){
     AnimMtxList * this;
-    Mtx *start_ptr;
-    Mtx *end_ptr;
-    Mtx *i_ptr;
+    MtxF *start_ptr;
+    MtxF *end_ptr;
+    MtxF *i_ptr;
     BKAnimation *s0;
     f32 tmp_f0;
     s32 pad94[1];
diff --git a/src/core2/code_637D0.c b/src/core2/code_637D0.c
index b3118452..c81f66d1 100644
--- a/src/core2/code_637D0.c
+++ b/src/core2/code_637D0.c
@@ -1,13 +1,15 @@
 #include <ultra64.h>
 #include "functions.h"
 #include "variables.h"
+#include "ml/mtx.h"
 
 extern void func_80252C08(f32[3], f32[3], f32, f32[3]);
-extern void func_80252CC4(f32[3], f32[3], f32, s32);
+extern void func_80252CC4(f32[3], f32[3], f32, f32[3]);
 extern void mlMtx_apply_f3(f32[3], f32, f32, f32);
-extern  s32 func_802EBAE0(BKModelUnk14List *arg0, f32 position[3], f32 rotation[3], f32 scale, s32 arg4, s32 arg5, f32 arg6[3], f32 arg7, f32 arg8[3]);
+extern  s32 func_802EBAE0(BKModelUnk14List *arg0, f32 position[3], f32 rotation[3], f32 scale, f32 arg4[3], AnimMtxList* arg5, f32 arg6[3], f32 arg7, f32 arg8[3]);
+extern MtxF *animMtxList_get(AnimMtxList *this, s32 arg1);
 /* .bss */
-Mtx D_80380880;
+MtxF D_80380880;
 
 /* .code */
 bool func_802EA760(BKModelUnk14List *arg0, s32 arg1, f32 arg2[3], f32 rotation[3], f32 scale, f32 arg5[3], f32 arg6[3], f32 *arg7) {
@@ -18,7 +20,7 @@ bool func_802EA760(BKModelUnk14List *arg0, s32 arg1, f32 arg2[3], f32 rotation[3
         return FALSE;
     }
     start = sizeof(BKModelUnk14List) + arg0->cnt0*sizeof(BKModelUnk14_0) + (s32)arg0;
-    temp_v0 = arg0->cnt2*sizeof(BKModelUnk14_1) + start + arg1 *sizeof(BKModelUnk14_2);
+    temp_v0 = (BKModelUnk14_2*)(arg0->cnt2*sizeof(BKModelUnk14_1) + start + arg1 *sizeof(BKModelUnk14_2));
 
     arg6[0] = (f32) temp_v0->unk2[0];
     arg6[1] = (f32) temp_v0->unk2[1];
@@ -67,7 +69,7 @@ s32 func_802EA864(BKModelUnk14List *arg0, f32 position[3], f32 rotation[3], f32
         mlMtxIdent();
         func_80252EC8(spB0, sp8C);
         func_80252CC4(position, rotation, scale, arg4);
-        func_8025235C(&sp78, arg5);
+        func_8025235C(sp78, arg5);
         for(j = 0; j < 3; j++){
             if (((sp78[j] + arg6 / scale) <= spA4[j]) || (sp98[j] <= (sp78[j] - arg6 / scale))) 
                 break;
@@ -93,7 +95,7 @@ s32 func_802EAB34(BKModelUnk14List *arg0, f32 position[3], f32 rotation[3], f32
 
     // tmp = ;
     tmp = (BKModelUnk14_0 *)(arg0 + 1);
-    iPtr = tmp + arg0->cnt0;
+    iPtr = (BKModelUnk14_1*)(tmp + arg0->cnt0);
     end_ptr = iPtr + arg0->cnt2;
     for(iPtr = iPtr; iPtr < end_ptr; iPtr++){
         spA0[0] = (f32) iPtr->unk4[0];
@@ -108,7 +110,7 @@ s32 func_802EAB34(BKModelUnk14List *arg0, f32 position[3], f32 rotation[3], f32
         mlMtxIdent();
         func_80252DDC(spA0, sp94);
         func_80252CC4(position, rotation, scale, arg4);
-        func_8025235C(&sp78, arg5);
+        func_8025235C(sp78, arg5);
         if (!(temp_f20 / 2 <= (sp78[2] - arg6 / scale)) && !((sp78[2] + arg6 / scale) <= -(temp_f20 / 2))) {
             if (!(((arg6 / scale + temp_f24) *  (arg6 / scale + temp_f24)) <= ((sp78[0] * sp78[0]) + (sp78[1] * sp78[1])))) {
                 return iPtr->unkD;
@@ -119,7 +121,7 @@ s32 func_802EAB34(BKModelUnk14List *arg0, f32 position[3], f32 rotation[3], f32
     return 0;
 }
 
-s32 func_802EAD5C(BKModelUnk14List *arg0, f32 position[3], f32 rotation[3], f32 scale, s32 arg4, f32 arg5[3], f32 arg6) {
+s32 func_802EAD5C(BKModelUnk14List *arg0, f32 position[3], f32 rotation[3], f32 scale, f32 arg4[3], f32 arg5[3], f32 arg6) {
     BKModelUnk14_0 *t0_ptr;
     BKModelUnk14_1 *t1_ptr;
     f32 sp5C[3];
@@ -153,7 +155,7 @@ s32 func_802EAD5C(BKModelUnk14List *arg0, f32 position[3], f32 rotation[3], f32
     return 0;
 }
 
-s32 func_802EAED4(BKModelUnk14List *arg0, f32 position[3], f32 rotation[3], f32 scale, s32 arg4, AnimMtxList *arg5, f32 arg6[3], f32 arg7) {
+s32 func_802EAED4(BKModelUnk14List *arg0, f32 position[3], f32 rotation[3], f32 scale, f32 arg4[3], AnimMtxList *arg5, f32 arg6[3], f32 arg7) {
     f32 spF4[3];
     f32 spE8[3];
     f32 spDC[3];
@@ -242,7 +244,7 @@ s32 func_802EAED4(BKModelUnk14List *arg0, f32 position[3], f32 rotation[3], f32
     return 0;
 }
 
-s32 func_802EB458(BKModelUnk14List *arg0, f32 position[3], f32 rotation[3], f32 scale, s32 arg4, s32 arg5, f32 arg6[3], f32 arg7)
+s32 func_802EB458(BKModelUnk14List *arg0, f32 position[3], f32 rotation[3], f32 scale, f32 arg4[3], AnimMtxList* arg5, f32 arg6[3], f32 arg7)
 {
     BKModelUnk14_0 *ptr_t0;
     f32 spB8[3];
@@ -313,7 +315,7 @@ s32 func_802EB458(BKModelUnk14List *arg0, f32 position[3], f32 rotation[3], f32
     return 0;
 }
 
-s32 func_802EB8A0(BKModelUnk14List *arg0, f32 *position, f32 *rotation, f32 scale, s32 arg4, s32 arg5, f32 *arg6, f32 arg7) {
+s32 func_802EB8A0(BKModelUnk14List *arg0, f32 *position, f32 *rotation, f32 scale, f32 arg4[3], AnimMtxList* arg5, f32 *arg6, f32 arg7) {
     BKModelUnk14_0 *t0_ptr;
     BKModelUnk14_1 *t1_ptr;
     f32 sp74[3];
@@ -326,7 +328,7 @@ s32 func_802EB8A0(BKModelUnk14List *arg0, f32 *position, f32 *rotation, f32 scal
 
     mlMtxIdent();
     func_80252CC4(position, rotation, scale, arg4);
-    func_8025235C(&sp74, arg6);
+    func_8025235C(sp74, arg6);
     t0_ptr = (BKModelUnk14_0 *)(arg0 + 1);
     t1_ptr = (BKModelUnk14_1 *)(t0_ptr + arg0->cnt0);
     i_ptr = (BKModelUnk14_2 *)(t1_ptr  + arg0->cnt2);
@@ -353,11 +355,11 @@ s32 func_802EB8A0(BKModelUnk14List *arg0, f32 *position, f32 *rotation, f32 scal
     return 0;
 }
 
-bool func_802EBA98(BKModelUnk14List *arg0, f32 arg1[3], f32 rotation[3], f32 scale, s32 arg4, f32 arg5[3], f32 arg6, f32 arg7[3]){
+bool func_802EBA98(BKModelUnk14List *arg0, f32 arg1[3], f32 rotation[3], f32 scale, f32 arg4[3], f32 arg5[3], f32 arg6, f32 arg7[3]){
     return func_802EBAE0(arg0, arg1, rotation, scale, arg4, NULL, arg5, arg6, arg7);
 }
 
-s32 func_802EBAE0(BKModelUnk14List *arg0, f32 position[3], f32 rotation[3], f32 scale, s32 arg4, s32 arg5, f32 arg6[3], f32 arg7, f32 arg8[3])
+s32 func_802EBAE0(BKModelUnk14List *arg0, f32 position[3], f32 rotation[3], f32 scale, f32 arg4[3], AnimMtxList* arg5, f32 arg6[3], f32 arg7, f32 arg8[3])
 {
   f32 sp3C[3];
   f32 temp_f0;
@@ -467,7 +469,7 @@ s32 func_802EBD3C(BKModelUnk14List *arg0, f32 arg1[3], f32 rotation[3], f32 scal
     return 0;
 }
 
-s32 func_802EC000(BKModelUnk14List *arg0, f32 arg1[3], f32 rotation[3], f32 scale, s32 arg4, f32 arg5[3], s32 arg6){
+s32 func_802EC000(BKModelUnk14List *arg0, f32 arg1[3], f32 rotation[3], f32 scale, f32 arg4[3], f32 arg5[3], s32 arg6){
     BKModelUnk14_1 *i_ptr;
     f32 sp90[3];
     f32 sp84[3];
@@ -492,9 +494,9 @@ s32 func_802EC000(BKModelUnk14List *arg0, f32 arg1[3], f32 rotation[3], f32 scal
             temp_f20 = (f32) i_ptr->unk0;
             temp_f22 = (f32) i_ptr->unk2;
             mlMtxIdent();
-            func_80252DDC(&sp90, &sp84);
+            func_80252DDC(sp90, sp84);
             func_80252CC4(arg1, rotation, scale, arg4);
-            func_8025235C(&sp68, arg5);
+            func_8025235C(sp68, arg5);
             temp_f0 = (f32) (temp_f22 / 2.0);
             if (!(temp_f0 <= sp68[2]) && !(sp68[2] <= -temp_f0) && !((temp_f20 * temp_f20) <= (sp68[0] * sp68[0] + sp68[1]*sp68[1]))) {
                 return i_ptr->unkD;
@@ -506,7 +508,7 @@ s32 func_802EC000(BKModelUnk14List *arg0, f32 arg1[3], f32 rotation[3], f32 scal
     return 0;
 }
 
-s32 func_802EC238(BKModelUnk14List *arg0, f32 arg1[3], f32 rotation[3], f32 scale, f32 arg4[3], s32 arg5, s32 arg6){
+s32 func_802EC238(BKModelUnk14List *arg0, f32 arg1[3], f32 rotation[3], f32 scale, f32 arg4[3], f32 arg5[3], s32 arg6){
     BKModelUnk14_0 *t0_ptr;
     BKModelUnk14_1 *t1_ptr;
     f32 sp54[3];
diff --git a/src/core2/code_6B030.c b/src/core2/code_6B030.c
index 34673821..8fde81fb 100644
--- a/src/core2/code_6B030.c
+++ b/src/core2/code_6B030.c
@@ -2,6 +2,8 @@
 #include "functions.h"
 #include "variables.h"
 
+#include "ml/mtx.h"
+
 extern s16 D_803A5D00[2][0xF660];
 
 
@@ -49,7 +51,7 @@ Struct_core2_6B030_1 D_80368B28[] = {
 
 /* .bss */
 f32 *D_80380A10;
-Mtx D_80380A18;
+MtxF D_80380A18;
 s32 D_80380A58;
 
 /* .code */
@@ -377,7 +379,7 @@ void func_802F2FCC(Struct64s *arg0, f32 arg1[3], s16 arg2, f32 arg3, ActorMarker
     }
 }
 
-void func_802F32C4(Struct64s *arg0, f32 arg1[3], f32 arg2, ActorMarker *arg3, void(*arg4)(f32[3], f32, ActorMarker *)){
+void func_802F32C4(Struct64s *arg0, f32 arg1[3], f32 arg2, ActorMarker *arg3, s32(*arg4)(f32[3], f32, ActorMarker *)){
     func_802F2FCC(arg0, arg1, 8, arg2, arg3, arg4);
 }
 
diff --git a/src/core2/code_72060.c b/src/core2/code_72060.c
index 3c9c537a..e904b21e 100644
--- a/src/core2/code_72060.c
+++ b/src/core2/code_72060.c
@@ -1,6 +1,7 @@
 #include <ultra64.h>
 #include "functions.h"
 #include "variables.h"
+#include "ml/mtx.h"
 
 typedef struct {
     u8 pad0[0xC];
@@ -10,10 +11,7 @@ typedef struct {
 extern f32  viewport_get_yaw();
 extern int  func_8024DD34(f32, f32, f32);
 extern void func_80251B5C(f32, f32, f32);
-extern void mlMtx_rotate_pitch_deg(f32);
-extern void mlMtx_rotate_yaw_deg(f32);
 extern void func_80252A38(f32, f32, f32);
-extern void mlMtxApply(Mtx *);
 extern f32 vtxList_getGlobalNorm(BKVertexList *);
 extern bool func_8024DB50(f32[3], f32);
 
@@ -202,17 +200,17 @@ void func_802F962C(Gfx **gfx, Mtx **mtx, Vtx **vtx) {
     if ((D_80369280 != NULL) && (D_80369284 != 0)) {
         viewport_get_position_vec3f(D_80381050);
         viewport_get_rotation_vec3f(D_80381060);
-        D_80381090 = (s32)D_80369288 + D_80369288->gfx_list_offset_C + sizeof(BKGfxList);
+        D_80381090 = (Gfx*)((s32)D_80369288 + D_80369288->gfx_list_offset_C + sizeof(BKGfxList));
         temp_s3 = (BKVertexList *)((s32)D_80369288 + D_80369288->vtx_list_offset_10);
         D_8038108C = vtxList_getGlobalNorm(temp_s3);
         func_80349AD0();
         gSPSegment((*gfx)++, 1, osVirtualToPhysical(temp_s3 + 1));
-        gSPSegment((*gfx)++, 0x02, osVirtualToPhysical((s32)D_80369288 + D_80369288->texture_list_offset_8 + sizeof(BKTextureList) + sizeof(BKTextureHeader)));
+        gSPSegment((*gfx)++, 0x02, osVirtualToPhysical((void*)((s32)D_80369288 + D_80369288->texture_list_offset_8 + sizeof(BKTextureList) + sizeof(BKTextureHeader))));
         gSPSetGeometryMode((*gfx)++, G_ZBUFFER);
         gSPDisplayList((*gfx)++, D_80369290);
         gSPSegment((*gfx)++, 0x03, osVirtualToPhysical(&D_803692B0));
 
-        D_80381094 = (BKGeoList *)((s32)D_80369288 + D_80369288->geo_list_offset_4);
+        D_80381094 = (Struct_core2_72060_0 *)((s32)D_80369288 + D_80369288->geo_list_offset_4);
         
         for(phi_s0 = D_80369280->unk1C; phi_s0 < D_80369280->unk1C + D_80369284; phi_s0++) {
             if ((func_802F989C(gfx, mtx, phi_s0) == 0) && (phi_s0->unk0[1] < D_8038104C)) {
diff --git a/src/core2/code_AEDA0.c b/src/core2/code_AEDA0.c
index 2861baaa..e249ef7b 100644
--- a/src/core2/code_AEDA0.c
+++ b/src/core2/code_AEDA0.c
@@ -1,16 +1,15 @@
 #include <ultra64.h>
 #include "functions.h"
 #include "variables.h"
+#include "ml/mtx.h"
 
-Mtx *func_8024DD9C(void);
-void mlMtxSet(Mtx *);
 void func_80252330(f32, f32, f32);
-void mlMtxApply(Mtx *);
 void func_803382D8(s32 arg0);
 void func_803382E4(s32 arg0);
 void spriteRender_set1Primative(bool boolean);
 void func_803382FC(s32 arg0);
 void func_80338308(s32 arg0, s32 arg1);
+BKSpriteTextureBlock *func_8033EFB0(Struct84s *arg0, s32 arg1);
 
 /* .data */
 Gfx D_80370260[] = {
@@ -263,17 +262,17 @@ void spriteRender_drawWithSegment(Gfx **gfx, Vtx **vtx, BKSprite *sprite, u32 fr
     frame_ptr = spriteGetFramePtr(sprite, frame);
     
     //load palette in indexed pixels
-    var_t2 = frame_ptr + 1;
+    var_t2 = (BKSpriteTextureBlock *)(frame_ptr + 1);
     if (sprite->type & SPRITE_TYPE_CI4) {
         gDPSetTextureLUT((*gfx)++, G_TT_RGBA16);
         palette_mem = ALIGN(frame_ptr + 1, 8);
         gDPLoadTLUT_pal16((*gfx)++, 0, palette_mem);
-        var_t2 = palette_mem + 0x20;
+        var_t2 = (BKSpriteTextureBlock *)(palette_mem + 0x20);
     } else if (sprite->type & SPRITE_TYPE_CI8) {
         gDPSetTextureLUT((*gfx)++, G_TT_RGBA16);
         palette_mem = ALIGN(frame_ptr + 1, 8);
         gDPLoadTLUT_pal256((*gfx)++, palette_mem);
-        var_t2 = palette_mem + 0x200;
+        var_t2 = (BKSpriteTextureBlock *)(palette_mem + 0x200);
     }
     
 //start new vtx seg
@@ -290,7 +289,7 @@ void spriteRender_drawWithSegment(Gfx **gfx, Vtx **vtx, BKSprite *sprite, u32 fr
     // sp1B4 = sp1B4;
     for(sp1BC = 0; sp1BC < frame_ptr->chunkCnt; sp1BC++){
         temp_ra = var_t2->h;
-        tmem = ALIGN(var_t2 + 1, 8); //align
+        tmem = (u8*)ALIGN(var_t2 + 1, 8); //align
 
         //load texture block
         if (sprite->type & SPRITE_TYPE_RGBA16) {
@@ -342,7 +341,7 @@ void spriteRender_drawWithSegment(Gfx **gfx, Vtx **vtx, BKSprite *sprite, u32 fr
                 gSPVertex((*gfx)++, sp1B0, 0, 0);
             }
         }
-        var_t2 = tmem + ((s32) (var_t2->w * var_t2->h) * pixel_size_nibbles / 2);
+        var_t2 = (BKSpriteTextureBlock *)(tmem + ((s32) (var_t2->w * var_t2->h) * pixel_size_nibbles / 2));
     }
     // sp1B4 = reinterpret_cast(Gfx *,sp1B4);
     *vtx = var_a3;
diff --git a/src/core2/code_B8860.c b/src/core2/code_B8860.c
index cf657fce..200879a2 100644
--- a/src/core2/code_B8860.c
+++ b/src/core2/code_B8860.c
@@ -1,10 +1,10 @@
 #include <ultra64.h>
 #include "functions.h"
 #include "variables.h"
+#include "ml/mtx.h"
 
 extern void mlMtxRotatePYR(f32, f32, f32);
 extern void func_80252330(f32, f32, f32);
-extern void mlMtxApply(Mtx *);
 
 typedef struct {
     BKSprite *sprite_0;
diff --git a/src/core2/code_BD100.c b/src/core2/code_BD100.c
index 6d06d4a4..2bd4aa4a 100644
--- a/src/core2/code_BD100.c
+++ b/src/core2/code_BD100.c
@@ -1,11 +1,10 @@
 #include <ultra64.h>
 #include "functions.h"
 #include "variables.h"
+#include "ml/mtx.h"
 
 extern void spriteRender_drawWithSegment(Gfx**, Vtx**, BKSprite *, s32, s32);
-extern void func_80252330(f32, f32, f32);
-extern void mlMtxRotatePYR(f32, f32, f32);
-extern void mlMtxApply(Mtx *);
+extern void func_80252330(f32, f32, f32);\
 extern void func_80349AD0(void);
 extern void func_8024C5A8(f32[3]);
 
diff --git a/src/core2/code_BE2C0.c b/src/core2/code_BE2C0.c
index 86fcffa9..15da711f 100644
--- a/src/core2/code_BE2C0.c
+++ b/src/core2/code_BE2C0.c
@@ -2,6 +2,8 @@
 #include "functions.h"
 #include "variables.h"
 
+#include "ml/mtx.h"
+
 extern f32 func_80255D70(f32);
 
 #define LENGTH_SQ_VEC4F(v) (v[0]*v[0] + v[1]*v[1] + v[2]*v[2] + v[3]*v[3])
@@ -160,7 +162,7 @@ void func_803458E4(f32 arg0[4], f32 arg1[4], f32 arg2[4], f32 arg3) {
 
 }
 
-void func_80345A44(f32 arg0[4], f32 arg1[4][4]) {
+void func_80345A44(f32 arg0[4], MtxF* arg1) {
     f32 temp_f0;
     f32 temp_f2;
     s32 temp_a0;
@@ -168,32 +170,32 @@ void func_80345A44(f32 arg0[4], f32 arg1[4][4]) {
     s32 var_a0;
 
 
-    temp_f0 = arg1[0][0] + arg1[1][1] + arg1[2][2];
+    temp_f0 = arg1->m[0][0] + arg1->m[1][1] + arg1->m[2][2];
     if (temp_f0 > 0.0) {
         temp_f2 = gu_sqrtf(temp_f0 + 1.0);
         arg0[3] = temp_f2 * 0.5;
         temp_f2 = (0.5 / temp_f2);
-        arg0[0] = (arg1[1][2] - arg1[2][1]) * temp_f2;
-        arg0[1] = (arg1[2][0] - arg1[0][2]) * temp_f2;
-        arg0[2] = (arg1[0][1] - arg1[1][0]) * temp_f2;
+        arg0[0] = (arg1->m[1][2] - arg1->m[2][1]) * temp_f2;
+        arg0[1] = (arg1->m[2][0] - arg1->m[0][2]) * temp_f2;
+        arg0[2] = (arg1->m[0][1] - arg1->m[1][0]) * temp_f2;
     }
     else{
         var_a0 = 0;
-        if (arg1[var_a0][var_a0] < arg1[1][1]) {
+        if (arg1->m[var_a0][var_a0] < arg1->m[1][1]) {
             var_a0 = 1;
         }
-        if (arg1[var_a0][var_a0] < arg1[2][2]) {
+        if (arg1->m[var_a0][var_a0] < arg1->m[2][2]) {
             var_a0 = 2;
         }
         temp_a0 = D_80371ED0[var_a0];
         temp_a2 = D_80371ED0[temp_a0];
 
-        temp_f2 = gu_sqrtf((arg1[var_a0][var_a0] - (arg1[temp_a0][temp_a0] + arg1[temp_a2][temp_a2])) + 1.0);
+        temp_f2 = gu_sqrtf((arg1->m[var_a0][var_a0] - (arg1->m[temp_a0][temp_a0] + arg1->m[temp_a2][temp_a2])) + 1.0);
         arg0[var_a0] = temp_f2 * 0.5;
         temp_f2 = (0.5 / temp_f2);
-        arg0[3] = (arg1[temp_a0][temp_a2] - arg1[temp_a2][temp_a0]) * temp_f2;
-        arg0[temp_a0] = (arg1[var_a0][temp_a0] + arg1[temp_a0][var_a0]) * temp_f2;
-        arg0[temp_a2] = (arg1[var_a0][temp_a2] + arg1[temp_a2][var_a0]) * temp_f2;
+        arg0[3] = (arg1->m[temp_a0][temp_a2] - arg1->m[temp_a2][temp_a0]) * temp_f2;
+        arg0[temp_a0] = (arg1->m[var_a0][temp_a0] + arg1->m[temp_a0][var_a0]) * temp_f2;
+        arg0[temp_a2] = (arg1->m[var_a0][temp_a2] + arg1->m[temp_a2][var_a0]) * temp_f2;
     }
 }
 
diff --git a/src/core2/gc/pauseMenu.c b/src/core2/gc/pauseMenu.c
index cfc9ddbc..d8fa2da2 100644
--- a/src/core2/gc/pauseMenu.c
+++ b/src/core2/gc/pauseMenu.c
@@ -1,6 +1,7 @@
 #include <ultra64.h>
 #include "functions.h"
 #include "variables.h"
+#include "ml/mtx.h"
 
 #include "zoombox.h"
 
@@ -26,7 +27,6 @@ extern void func_802C5994(void);
 extern void func_802E412C(s32, s32);
 void volatileFlag_set(s32, s32);
 f32 func_8024DE1C(f32, f32, f32 *, f32 *);
-void mlMtxApply(Mtx*);
 void func_80310D2C(void);
 
 s32 level_get(void);
diff --git a/src/core2/gc/zoombox.c b/src/core2/gc/zoombox.c
index 142fcb26..fbc0059b 100644
--- a/src/core2/gc/zoombox.c
+++ b/src/core2/gc/zoombox.c
@@ -2,6 +2,7 @@
 #include "functions.h"
 #include "variables.h"
 #include "zoombox.h"
+#include "ml/mtx.h"
 
 extern f32 func_8024DE1C(f32, f32, f32[3], f32[3]);
 void func_80252330(f32, f32, f32);
@@ -9,6 +10,8 @@ extern f32 func_8033DDB8(void);
 extern void func_8024E60C(s32, void *);
 extern void func_8024E640(s32, void *);
 extern void func_8024E5A8(s32, void *);
+extern void func_80344090(BKSpriteDisplayData *self, s32 frame, Gfx **gfx);
+BKSprite *func_8033B6C4(enum asset_e sprite_id, BKSpriteDisplayData **arg1);
 
 typedef struct struct_18_s{
     s16 sfx_id; /* enum sfx_e */
@@ -838,7 +841,7 @@ void func_803163A8(GcZoombox *this, Gfx **gfx, Mtx **mtx) {
     modelRender_draw(gfx, mtx, sp50, sp5C, this->unk198 * sp34, sp38, this->model);
 }
 
-void func_803164B0(GcZoombox *this, Gfx **gfx, Mtx **mtx, s32 arg3, s32 arg4, s32 arg5, f32 arg6) {
+void func_803164B0(GcZoombox *this, Gfx **gfx, Mtx **mtx, s32 arg3, s32 arg4, BKSpriteDisplayData *arg5, f32 arg6) {
     f32 sp2C[3];
     f32 temp_f12;
 
@@ -1190,7 +1193,7 @@ void gczoombox_update(GcZoombox *this){
                     }//L803176C8
                     if(this->unk1A4_25 == 0){
                          if(this->unk166 - ((this->unk1A4_19)? 0xf : 0xc) >= this->unk16C){
-                              func_803153A8(this->unk30, this->unk0, 0, 0x30);
+                              func_803153A8(this->unk30, this->unk0, 0, sizeof(this->unk0));
                               _gczoombox_memClear(this->unk30, 0x30);
                               this->unk16C = this->unk166;
                               this->unk16E = ((this->unk1A4_19) ? 0xf : 0xc) + this->unk16C;
@@ -1459,9 +1462,9 @@ GcZoombox *gczoombox_new(s32 arg0, GcZoomboxSprite portrait_id, s32 arg2, s32 ar
     func_80318760(this, 18000);
     _gczoombox_memClear( this->unk0, 0x30);
     _gczoombox_memClear( this->unk30, 0x30);
-    _gczoombox_memClear( &this->unk60, 0x30);
-    _gczoombox_memClear( &this->unk90, 0x20);
-    _gczoombox_memClear( &this->unkB0, 0x40);
+    _gczoombox_memClear( this->unk60, 0x30);
+    _gczoombox_memClear( this->unk90, 0x20);
+    _gczoombox_memClear( this->unkB0, 0x40);
     return this;
 }
 
@@ -1768,10 +1771,10 @@ void func_80318C48(GcZoombox *this, s32 arg1) {
                 if (this->unk0[0] == 0xFD) {
                     this->unk0[1] = 0x68;
                 } else {
-                    func_803153A8(&this->unk0[0], &D_803830B0, 0, 0x30);
+                    func_803153A8(&this->unk0[0], D_803830B0, 0, sizeof(D_803830B0));
                     this->unk0[0] = 0xFD;
                     this->unk0[1] = 0x68;
-                    func_803153A8(&D_803830B0, &this->unk0[2], 0, 0x2E);
+                    func_803153A8(D_803830B0, &this->unk0[2], 0, sizeof(this->unk0) - 2);
                 }
             }
             else if (this->unk0[0] == 0xFD) {
@@ -1785,10 +1788,10 @@ void func_80318C48(GcZoombox *this, s32 arg1) {
                     return;
                 }
                 else{
-                    func_803153A8(&this->unk30[0], &D_803830B0, 0, 0x30);
+                    func_803153A8(&this->unk30[0], D_803830B0, 0, sizeof(D_803830B0));
                     this->unk30[0] = 0xFD;
                     this->unk30[1] = 0x68;
-                    func_803153A8(&D_803830B0, &this->unk30[2], 0, 0x2E);
+                    func_803153A8(D_803830B0, &this->unk30[2], 0, sizeof(this->unk30) - 2);
                 }
             }
             else if (this->unk30[0] == 0xFD) {
diff --git a/src/core2/gc/zoombox.h b/src/core2/gc/zoombox.h
index 0a1d21a3..203d1c80 100644
--- a/src/core2/gc/zoombox.h
+++ b/src/core2/gc/zoombox.h
@@ -209,7 +209,7 @@ typedef struct {
 void gczoombox_free(GcZoombox *this);
 void func_803160A8(GcZoombox *this);
 void func_803163A8(GcZoombox *this, Gfx **gfx, Mtx **mtx);
-void func_803164B0(GcZoombox *this, Gfx **gfx, Mtx **mtx, s32 arg3, s32 arg4, s32 arg5, f32 arg6);
+void func_803164B0(GcZoombox *this, Gfx **gfx, Mtx **mtx, s32 arg3, s32 arg4, BKSpriteDisplayData *arg5, f32 arg6);
 void func_80316764(GcZoombox *this, s32 arg1);
 void gczoombox_draw(GcZoombox *this, Gfx **gdl, Mtx ** mptr, void *vptr);
 void func_80316E08(GcZoombox *this);
diff --git a/src/core2/modelRender.c b/src/core2/modelRender.c
index df2dd9b5..3aa25cb1 100644
--- a/src/core2/modelRender.c
+++ b/src/core2/modelRender.c
@@ -3,6 +3,7 @@
 #include "variables.h"
 #include "core2/modelRender.h"
 #include "animation.h"
+#include "ml/mtx.h"
 
 #define ARRAYLEN(x) (sizeof(x) / sizeof((x)[0]))
 
@@ -17,11 +18,9 @@ extern void viewport_set_rotation_vec3f(f32[3]);
 extern void viewport_update(void);
 extern void func_8033BD4C(BKModelBin *);
 extern s32 func_8024DB50(f32[3], f32);
-extern void mlMtx_push_translation(f32, f32, f32);
-extern void mlMtxScale(f32);
-extern void mlMtxApply(Mtx* mtx);
 extern AnimMtxList *animMtxList_new();
 extern AnimMtxList *animMtxList_defrag(AnimMtxList *);
+extern MtxF *animMtxList_get(AnimMtxList *this, s32 arg1);
 
 
 typedef struct{
@@ -662,7 +661,7 @@ struct {
     LookAt *lookat_buffer_end;
     f32 eye_pos[3];
 } D_803837E0;
-Mtx D_80383BF8;
+MtxF D_80383BF8;
 f32 modelRenderCameraPosition[3];
 f32 modelRenderCameraRotation[3];
 BKModelBin *modelRenderModelBin;
@@ -909,9 +908,9 @@ void func_80338CD0(Gfx **gfx, Mtx **mtx, void *arg2){
     if (0 < indx) {
         if (indx <= cmd->unk8) {
             s0 = cmd->unkC;
-            sub_cmd = cmd;
+            sub_cmd = (s32)cmd;
             sub_cmd += *(s32*)(s0 + (indx - 1));
-            func_80339124(gfx, mtx, sub_cmd);
+            func_80339124(gfx, mtx, (BKGeoList*)sub_cmd);
         }
     } else {
         s1 = indx * (-1);
@@ -919,9 +918,9 @@ void func_80338CD0(Gfx **gfx, Mtx **mtx, void *arg2){
         for (s2 = 0; s2 < cmd->unk8; s2++) {
             if (s1 & 1)
             {
-                sub_cmd = cmd;
+                sub_cmd = (s32)cmd;
                 sub_cmd += s0[0];
-                func_80339124(gfx, mtx, sub_cmd);
+                func_80339124(gfx, mtx, (BKGeoList*)sub_cmd);
             }
             s1 >>= 1;
             s0++;