diff --git a/GPU2D.cpp b/GPU2D.cpp
index 8479d97e..de3636e7 100644
--- a/GPU2D.cpp
+++ b/GPU2D.cpp
@@ -334,7 +334,7 @@ u16* GPU2D::GetOBJExtPal(u32 pal)
     {
         if (Num)
         {
-            if (GPU::VRAMMap_BOBJExtPal & (1<<7))
+            if (GPU::VRAMMap_BOBJExtPal & (1<<8))
                 memcpy(dst, &GPU::VRAM_I[(pal << 9)], 256*2);
             else
                 memset(dst, 0, 256*2);
@@ -864,7 +864,7 @@ void GPU2D::DrawSprite_Rotscale(u16* attrib, u16* rotparams, u32 boundwidth, u32
     if (DispCnt & 0x10)
     {
         tilenum <<= ((DispCnt >> 20) & 0x3);
-        ytilefactor = (width >> 3);
+        ytilefactor = (width >> 3) << ((attrib[0] & 0x2000) ? 1:0);
     }
     else
     {
@@ -901,6 +901,34 @@ void GPU2D::DrawSprite_Rotscale(u16* attrib, u16* rotparams, u32 boundwidth, u32
     if (attrib[0] & 0x2000)
     {
         // 256-color
+        tilenum <<= 5;
+        ytilefactor <<= 5;
+        u32 pixelsaddr = (Num ? 0x06600000 : 0x06400000) + tilenum;
+
+        u32 extpal = (DispCnt & 0x80000000);
+
+        u16* pal;
+        if (extpal) pal = GetOBJExtPal(attrib[2] >> 12);
+        else        pal = (u16*)&GPU::Palette[Num ? 0x600 : 0x200];
+
+        for (; xoff < boundwidth;)
+        {
+            if ((u32)rotX < width && (u32)rotY < height)
+            {
+                u8 color;
+
+                // blaaaarg
+                color = GPU::ReadVRAM_OBJ<u8>(pixelsaddr + ((rotY>>11)*ytilefactor) + ((rotY&0x700)>>5) + ((rotX>>11)*64) + ((rotX&0x700)>>8));
+
+                if (color)
+                    dst[xpos] = pal[color] | prio;
+            }
+
+            rotX += rotA;
+            rotY += rotC;
+            xoff++;
+            xpos++;
+        }
     }
     else
     {
@@ -949,7 +977,7 @@ void GPU2D::DrawSprite_Normal(u16* attrib, u32 width, s32 xpos, u32 ypos, u32* d
     }
     else
     {
-        tilenum += ((ypos >> 3) * 0x20) << ((attrib[0] & 0x2000) ? 1:0);
+        tilenum += ((ypos >> 3) * 0x20);
     }
 
     u32 wmask = width - 8; // really ((width - 1) & ~0x7)
diff --git a/GPU3D_Soft.cpp b/GPU3D_Soft.cpp
index 4549d4db..6ea1ba49 100644
--- a/GPU3D_Soft.cpp
+++ b/GPU3D_Soft.cpp
@@ -250,15 +250,15 @@ void RenderPolygon(Polygon* polygon)
         if (vlnext->FinalPosition[1] == vlcur->FinalPosition[1])
             lfactor = 0;
         else
-            lfactor = (((y - vlcur->FinalPosition[1]) << 12) + 0x00) / (vlnext->FinalPosition[1] - vlcur->FinalPosition[1]);
+            lfactor = (((y+1 - vlcur->FinalPosition[1]) << 12) + 0x00) / (vlnext->FinalPosition[1] - vlcur->FinalPosition[1]);
 
         if (vrnext->FinalPosition[1] == vrcur->FinalPosition[1])
             rfactor = 0;
         else
-            rfactor = (((y - vrcur->FinalPosition[1]) << 12) + 0x00) / (vrnext->FinalPosition[1] - vrcur->FinalPosition[1]);
+            rfactor = (((y+1 - vrcur->FinalPosition[1]) << 12) + 0x00) / (vrnext->FinalPosition[1] - vrcur->FinalPosition[1]);
 
-        s32 xl = vlcur->FinalPosition[0] + (((vlnext->FinalPosition[0] - vlcur->FinalPosition[0]) * lfactor) >> 12);
-        s32 xr = vrcur->FinalPosition[0] + (((vrnext->FinalPosition[0] - vrcur->FinalPosition[0]) * rfactor) >> 12);
+        s32 xl = vlcur->FinalPosition[0] + ((((vlnext->FinalPosition[0] - vlcur->FinalPosition[0]) * lfactor) + 0x800) >> 12);
+        s32 xr = vrcur->FinalPosition[0] + ((((vrnext->FinalPosition[0] - vrcur->FinalPosition[0]) * rfactor) + 0x800) >> 12);
 //printf("y:%d xl:%d xr:%d    %08X\n", y, xl, xr, rfactor); // y: 48 143
         if (xl > xr) // TODO: handle it in a more elegant way
         {
diff --git a/melonDS.depend b/melonDS.depend
index f479ee80..adc19479 100644
--- a/melonDS.depend
+++ b/melonDS.depend
@@ -10,7 +10,7 @@
 
 1481161027 c:\documents\sources\melonds\types.h
 
-1488237361 source:c:\documents\sources\melonds\nds.cpp
+1488281931 source:c:\documents\sources\melonds\nds.cpp
 	<stdio.h>
 	<string.h>
 	"NDS.h"
@@ -87,7 +87,7 @@
 	"NDS.h"
 	"SPI.h"
 
-1488226954 source:c:\documents\sources\melonds\gpu2d.cpp
+1488281814 source:c:\documents\sources\melonds\gpu2d.cpp
 	<stdio.h>
 	<string.h>
 	"NDS.h"
@@ -146,16 +146,16 @@
 	"NDS.h"
 	"NDSCart.h"
 
-1488238607 c:\documents\sources\melonds\gpu3d.h
+1488239485 c:\documents\sources\melonds\gpu3d.h
 
-1488239360 source:c:\documents\sources\melonds\gpu3d.cpp
+1488239513 source:c:\documents\sources\melonds\gpu3d.cpp
 	<stdio.h>
 	<string.h>
 	"NDS.h"
 	"GPU.h"
 	"FIFO.h"
 
-1488237540 source:c:\documents\sources\melonds\gpu3d_soft.cpp
+1488244847 source:c:\documents\sources\melonds\gpu3d_soft.cpp
 	<stdio.h>
 	<string.h>
 	"NDS.h"