diff --git a/code/graphics_system/dsl.s b/code/graphics_system/dsl.s
index 4dc6643..5127dd0 100644
--- a/code/graphics_system/dsl.s
+++ b/code/graphics_system/dsl.s
@@ -8,6 +8,10 @@ load_uimm  equ lui ; dst_reg, immediate value (unsigned)
 load_word  equ lw  ; dst_reg, offset(src_reg0) (offset is immediate value)
 ; Store
 store_word equ sw ; src_reg, dst_address 
+; Shift
+shift_ll equ sll ;
+shift_rl equ srl ;
+shift_ra equ sra ;
 ; Addition
 add_s  equ add   ; dst_reg, reg_a, reg_b (signed)
 add_u  equ add   ; dst_reg, reg_a, reg_b (unsigned)
diff --git a/code/graphics_system/gp.s b/code/graphics_system/gp.s
index 54e4c04..12c0ebd 100644
--- a/code/graphics_system/gp.s
+++ b/code/graphics_system/gp.s
@@ -66,7 +66,8 @@ gp_ModeSetting_DipArea     equ 0xE1 << gcmd_offset | 0x1 << gp_ModeSetting_DrawA
 
 ; GP0(E3h) - Set Drawing Area top left (X1,Y1)
 ; GP0(E4h) - Set Drawing Area bottom right (X2,Y2)
-; Sets the drawing area corners. The Render commands GP0(20h..7Fh) are automatically clipping any pixels that are outside of this region.
+; Sets the drawing area corners. The Render commands GP0(20h..7Fh) 
+; are automatically clipping any pixels that are outside of this region.
 ; 0  - 9   X-coordinate (0..1023)
 ; 10 - 18  Y-coordinate (0..511)   ; \ on v0 GPU (max 1 MB VRAM)
 ; 19 - 23  Not used (zero)         ; /
diff --git a/code/graphics_system/hello_image.s b/code/graphics_system/hello_image.s
index 0c82ae9..c3d5a79 100644
--- a/code/graphics_system/hello_image.s
+++ b/code/graphics_system/hello_image.s
@@ -92,15 +92,15 @@ main:
 	stack_alloc gp_draw_tri_flat__sp_size ; (used for following call)
 		move     rarg_0, reg_io_offset      ; (used for following call)
 		load_imm rarg_1, Color_PS_GoldenPoppy
-		load_imm rarg_2, 100 * -1 + Display_HalfHeight << gp_b16_Y | -100 + Display_HalfWidth << gp_b16_X :: sw rarg_2, 0 * gp_vec2($sp)
-		load_imm rarg_2,  20 * -1 + Display_HalfHeight << gp_b16_Y |   20 + Display_HalfWidth << gp_b16_X :: sw rarg_2, 1 * gp_vec2($sp)
-		load_imm rarg_2,  50 * -1 + Display_HalfHeight << gp_b16_Y |   30 + Display_HalfWidth << gp_b16_X :: sw rarg_2, 2 * gp_vec2($sp)
+		load_imm rarg_2, 100 * -1 + Display_HalfHeight << gp_b16_Y | -100 + Display_HalfWidth << gp_b16_X :: store_word rarg_2, 0 * gp_vec2($sp)
+		load_imm rarg_2,  20 * -1 + Display_HalfHeight << gp_b16_Y |   20 + Display_HalfWidth << gp_b16_X :: store_word rarg_2, 1 * gp_vec2($sp)
+		load_imm rarg_2,  50 * -1 + Display_HalfHeight << gp_b16_Y |   30 + Display_HalfWidth << gp_b16_X :: store_word rarg_2, 2 * gp_vec2($sp)
 	jump_nlink gp_draw_tri_flat :: nop
 ; Bonus traingle
 		load_imm rarg_1,  Color_PS_CadmiumRed
-		load_imm rarg_2,   50 * -1 + Display_HalfHeight << gp_b16_Y | -100 + Display_HalfWidth << gp_b16_X :: sw rarg_2, 0 * gp_vec2($sp)
-		load_imm rarg_2,    0 * -1 + Display_HalfHeight << gp_b16_Y |   20 + Display_HalfWidth << gp_b16_X :: sw rarg_2, 1 * gp_vec2($sp)
-		load_imm rarg_2, -100 * -1 + Display_HalfHeight << gp_b16_Y |   30 + Display_HalfWidth << gp_b16_X :: sw rarg_2, 2 * gp_vec2($sp)
+		load_imm rarg_2,   50 * -1 + Display_HalfHeight << gp_b16_Y | -100 + Display_HalfWidth << gp_b16_X :: store_word rarg_2, 0 * gp_vec2($sp)
+		load_imm rarg_2,    0 * -1 + Display_HalfHeight << gp_b16_Y |   20 + Display_HalfWidth << gp_b16_X :: store_word rarg_2, 1 * gp_vec2($sp)
+		load_imm rarg_2, -100 * -1 + Display_HalfHeight << gp_b16_Y |   30 + Display_HalfWidth << gp_b16_X :: store_word rarg_2, 2 * gp_vec2($sp)
 	jump_nlink gp_draw_tri_flat :: nop
 	stack_release gp_draw_tri_flat__sp_size
 ; Gourand shaded triangle
@@ -114,16 +114,13 @@ main:
 
 ; Copy image contents to vram
 	gcmd_push gp0, rtmp_1, gp_Blit_CPU_VM
-	gcmd_push gp0, rtmp_1, 		(Image_SizeY / 2) * -1 + Display_HalfHeight << gp_b16_Y | -(Image_SizeX / 2) + Display_HalfWidth << gp_b16_X
+	gcmd_push gp0, rtmp_1, 		0 << gp_b16_Y | Display_Width << gp_b16_X
 	gcmd_push gp0, rtmp_1, 		Image_SizeY << gp_b16_Y | Image_SizeX << gp_b16_X
 	; DMA commands
 	@id         equ rtmp_2
 	@img_cursor equ rtmp_3
-	@divisor    equ rtmp_3
-		load_imm      @id,      Image_ByteSize
-		load_imm      @divisor, word
-		div           @id,      @divisor
-		mov_from_high @id
+		load_imm @id, Image_ByteSize
+		shift_rl @id, @id, (word / 2)
 		load_addr @img_cursor, Image
 	loop_dma: 
 		load_word  rtmp_1, (@img_cursor) :: nop :: store_word rtmp_1, gp0 ; @img_curor -> gp_dma_cpu_vm(word)