diff --git a/macros/Load.Macs.s b/macros/Load.Macs.s
new file mode 100644
index 0000000..9d96ddd
--- /dev/null
+++ b/macros/Load.Macs.s
@@ -0,0 +1,113 @@
+* Load Manager macros
+*   by Dave Klimas
+;
+; Copyright Apple Computer, Inc. 1986, 1987
+; and Roger Wagner Publishing, Inc. 1988
+; All Rights Reserved
+;
+_LoaderInitialization  MAC
+                       Tool  $111
+                       <<<
+_LoaderStartUp         MAC
+                       Tool  $211
+                       <<<
+_LoaderShutDown        MAC
+                       Tool  $311
+                       <<<
+~LoaderVersion         MAC
+                       PHA
+_LoaderVersion         MAC
+                       Tool  $411
+                       <<<
+_LoaderReset           MAC
+                       Tool  $511
+                       <<<
+~LoaderStatus          MAC
+                       PHA
+_LoaderStatus          MAC
+                       Tool  $611
+                       <<<
+~InitialLoad           MAC
+                       PHS   5
+                       PHWL  ]1;]2
+                       PHW   ]3
+_InitialLoad           MAC
+                       Tool  $911
+                       <<<
+~InitialLoad2          MAC
+                       PHS   5
+                       PHWL  ]1;]2
+                       PxW   ]3;]4
+_InitialLoad2          MAC
+                       Tool  $2011
+                       <<<
+~Restart               MAC
+                       PHS   5
+                       PHW   ]1
+_Restart               MAC
+                       Tool  $A11
+                       <<<
+~LoadSegNum            MAC
+                       PHS   2
+                       PxW   ]1;]2;]3
+_LoadSegNum            MAC
+                       Tool  $B11
+                       <<<
+~UnloadSegNum          MAC
+                       PxW   ]1;]2;]3
+_UnloadSegNum          MAC
+                       Tool  $C11
+                       <<<
+~LoadSegName           MAC
+                       PHS   4
+                       PHW   ]1
+                       PxL   ]2;]3
+_LoadSegName           MAC
+                       Tool  $D11
+                       <<<
+~UnloadSeg             MAC
+                       PHS   3
+                       PHL   ]1
+_UnloadSeg             MAC
+                       Tool  $E11
+                       <<<
+~GetLoadSegInfo        MAC
+                       PxW   ]1;]2;]3
+                       PHL   ]4
+_GetLoadSegInfo        MAC
+                       Tool  $F11
+                       <<<
+~GetUserID             MAC
+                       P1SL  ]1
+_GetUserID             MAC
+                       Tool  $1011
+                       <<<
+~GetUserID2            MAC
+                       P1SL  ]1
+_GetUserID2            MAC
+                       Tool  $2111
+                       <<<
+~LGetPathname          MAC
+                       PHS   2
+                       PxW   ]1;]2
+_LGetPathname          MAC
+                       Tool  $1111
+                       <<<
+~LGetPathname2         MAC
+                       PHS   2
+                       PxW   ]1;]2
+_LGetPathname2         MAC
+                       Tool  $2211
+                       <<<
+~UserShutDown          MAC
+                       PHA
+                       PxW   ]1;]2
+_UserShutDown          MAC
+                       Tool  $1211
+                       <<<
+~RenamePathname        MAC
+                       PxL   ]1;]2
+_RenamePathname        MAC
+                       Tool  $1311
+                       <<<
+
diff --git a/macros/Util.Macs.s b/macros/Util.Macs.s
index 0a685aa..ef26b6e 100644
--- a/macros/Util.Macs.s
+++ b/macros/Util.Macs.s
@@ -339,6 +339,11 @@ Tool        MAC
             JSL       $E10000        ; go to dispatcher
             <<<
 
+UserTool    MAC
+            LDX       #]1            ; load tool call #
+            JSL       $E10008        ; go to dispatcher
+            <<<
+
 **************************************************
 * Auto-menu item macros                          *
 * This is one alternative for defining a menu    *
diff --git a/src/Core.s b/src/Core.s
index 4077e2c..84dea71 100644
--- a/src/Core.s
+++ b/src/Core.s
@@ -1,4 +1,5 @@
                   use       Util.Macs.s
+                  use       Load.Macs.s
                   use       Locator.Macs.s
                   use       Mem.Macs.s
                   use       Misc.Macs.s
@@ -39,6 +40,12 @@ EngineStartUp     ENT
                   plb
 
                   jsr       ToolStartUp         ; Start up the toolbox tools we rely on
+                  jsr       _CoreStartUp
+
+                  plb
+                  rtl
+
+_CoreStartUp
                   jsr       SoundStartUp        ; Start up any sound/music tools
                   jsr       IntStartUp          ; Enable certain iterrupts
 
@@ -51,22 +58,24 @@ EngineStartUp     ENT
                   jsr       InitTiles           ; Initialize the tile subsystem
 
                   jsr       InitTimers          ; Initialize the timer subsystem
-
-                  plb
-                  rtl
+                  rts
 
 EngineShutDown    ENT
                   phb
                   phk
                   plb
 
-                  jsr       IntShutDown
-                  jsr       SoundShutDown
+                  jsr       _CoreShutDown
                   jsr       ToolShutDown
 
                   plb
                   rtl
 
+_CoreShutDown
+                  jsr       IntShutDown
+                  jsr       SoundShutDown
+                  rts
+
 ToolStartUp
                   _TLStartUp                    ; normal tool initialization
                   pha
@@ -274,9 +283,10 @@ EngineReset
 ;  6. Game Boy Advanced     : 30 x 20   240 x 160 (19,200 bytes ( 60.0%))
 ;  7. Ancient Land of Y's   : 36 x 16   288 x 128 (18,432 bytes ( 57.6%))
 ;  8. Game Boy Color        : 20 x 18   160 x 144 (11,520 bytes ( 36.0%))
-; 
+;  9. Agony (Amiga)         : 36 x 24   288 x 192 (27,648 bytes ( 86.4%))
+;
 ;  X = mode number OR width in pixels (must be multiple of 2)
-;  Y = hwight in pixels (if X > 8)
+;  Y = height in pixels (if X > 8)
 
 ScreenModeWidth   dw        320,272,256,256,280,256,240,288,160,320
 ScreenModeHeight  dw        200,192,200,176,160,160,160,128,144,1
@@ -285,7 +295,11 @@ SetScreenMode     ENT
                   phb
                   phk
                   plb
+                  jsr       _SetScreenMode
+                  plb
+                  rtl
 
+_SetScreenMode
                   cpx       #9
                   bcs       :direct             ; if x > 8, then assume X and Y are the dimensions
 
@@ -326,11 +340,10 @@ SetScreenMode     ENT
                   plx
 
                   jsr       SetScreenRect
-                  jsr       FillScreen
-
+                  jmp       FillScreen          ; tail return
 :exit
-                  plb
-                  rtl
+                  rts
+
 
 WaitForKey        sep       #$20
                   stal      KBD_STROBE_REG      ; clear the strobe
@@ -348,6 +361,10 @@ ClearKbdStrobe    sep       #$20
 ; Read the keyboard and paddle controls and return in a game-controller-like format
 LastKey           db        0
 ReadControl       ENT
+                  jsr       _ReadControl
+                  rtl
+
+_ReadControl      
                   pea       $0000               ; low byte = key code, high byte = %------AB 
 
                   sep       #$20
@@ -391,7 +408,7 @@ ReadControl       ENT
 :KbdDown
                   rep       #$20
                   pla
-                  rtl
+                  rts
 
                   put       Memory.s
                   put       Graphics.s
diff --git a/src/Defs.s b/src/Defs.s
index 8609c53..111e272 100644
--- a/src/Defs.s
+++ b/src/Defs.s
@@ -78,8 +78,9 @@ BG1TileMapWidth        equ   82
 BG1TileMapHeight       equ   84
 BG1TileMapPtr          equ   86
 
-SCBArrayPtr            equ   90          ; USed for palette binding
-Next                   equ   94
+SCBArrayPtr            equ   90          ; Used for palette binding
+SpriteBanks            equ   94          ; Bank bytes for the sprite data and sprite mask
+Next                   equ   96
 
 BankLoad               equ   128
 
diff --git a/src/GTE.s b/src/GTE.s
index 74db746..030b172 100644
--- a/src/GTE.s
+++ b/src/GTE.s
@@ -74,3 +74,21 @@ AllocBank          EXT
 ScreenAddr         EXT
 OneSecondCounter   EXT
 BlitBuff           EXT
+
+; Helper function to load the GTE User Toolset
+GTEInstall
+                   php
+;                   ~InitialLoad userId;localToolPath;#0
+
+                   pea       $8000                ; User tool
+                   pea       $00A5                ; Tool 165
+                   PushLong  toolPtr
+                   _SetTSPtr
+                   
+                   plp
+                   rtl
+
+; Look for the tool set in the System Tools folder and then next to the application
+sysToolPath        strl       '*:System:Tools:ToolGTE'
+localToolPath      strl       '9:ToolGTE'
+toolPtr            adrl       0
\ No newline at end of file
diff --git a/src/Sprite.s b/src/Sprite.s
index 207a1e9..78160a6 100644
--- a/src/Sprite.s
+++ b/src/Sprite.s
@@ -1,4 +1,4 @@
-; Functions for sprie handling.  Mostly maintains the sprite list and provides
+; Functions for sprite handling.  Mostly maintains the sprite list and provides
 ; utility functions to calculate sprite/tile intersections
 ;
 ; The sprite plane actually covers two banks so that more than 32K can be used as a virtual 
@@ -31,6 +31,8 @@ InitSprites
            dex
            bpl    :loop3
 
+; Precalculate some bank values
+           jsr    _CacheSpriteBanks
            rts
 
 ; Run through the list of tile store offsets that this sprite was last drawn into and mark
@@ -116,23 +118,27 @@ _ClearSpriteFromTileStore
 ; any sprite that needs to be re-drawn has been marked as dirty.
 ;
 ; In the first phase, we run through the list of dirty sprites and erase them from their
-; OLD_VBUFF_ADDR.  This clears the sprite plane buffers.  We also interate through the
+; OLD_VBUFF_ADDR.  This clears the sprite plane buffers.  We also iterate through the
 ; TILE_STORE_ADDR_X array and mark all of the tile store location that this sprite had occupied
 ; as dirty, as well as removing this sprite from the TS_SPRITE_FLAG bitfield.
 ;
-; A final aspect is that any of the sprites idicated in the TS_SPRITE_FLAG are marked to be
+; A final aspect is that any of the sprites indicated in the TS_SPRITE_FLAG are marked to be
 ; drawn in the next phase (since a portion of their content may have been erased if they overlap)
 ; 
 ; In the second phase, the sprite is re-drawn into the sprite plane buffers and the appropriate
-; Tile Store locations are marked as dirty 
-;
-; 
+; Tile Store locations are marked as dirty. It is important to recognize that the sprites themselves
+; can be marked dirty, and the underlying tiles in the tile store are independently marked dirty.
+
 forceSpriteFlag ds 2
 _RenderSprites
 
 ; First step is to look at the StartX and StartY values.  If the offsets have changed from the
-; last time that the frame was redered, then we need to mark all of the sprites as dirty so that
+; last time that the frame was rendered, then we need to mark all of the sprites as dirty so that
 ; the tiles on which they were located at the previous frame will be refreshed
+;
+; OPTIMIZATION NOTE: Shoud check that the sprite actually chanegs position.  If the screen scrolles
+;                    by +X, but the sprite moves by -X (so it's relative position is unchanged), then
+;                    it does NOT need to be marked as dirty.
 
             stz   forceSpriteFlag
             lda   StartX
@@ -169,7 +175,7 @@ _RenderSprites
 
 ; Second step is to scan the list of sprites.  A sprite is either clean or dirty.  If it's dirty,
 ; then its position had changed, so we need to add tiles to the dirty queue to make sure the
-; playfield gets update.  If it's clean, we can skip everything.
+; playfield gets updated.  If it's clean, we can skip everything.
 
             ldy   #0
 :loop       lda   _Sprites+SPRITE_STATUS,y       ; If the status is zero, that's the sentinel value
@@ -371,7 +377,7 @@ draw_8x16h
              adc   #{8*SPRITE_PLANE_SPAN}
              tax
              tya
-             adc   #{128*32}                      ; 32 tiles to the next verical one, each tile is 128 bytes
+             adc   #{128*32}                      ; 32 tiles to the next vertical one, each tile is 128 bytes
              tay
              jmp   _DrawTile8x8
 
@@ -586,8 +592,8 @@ _DrawTile8x8
 ]line       equ   0
             lup   8
             lda:  tiledata+32+{]line*4},y
-            andl  spritemask+{]line*256},x
-            stal  spritemask+{]line*256},x
+            andl  spritemask+{]line*SPRITE_PLANE_SPAN},x
+            stal  spritemask+{]line*SPRITE_PLANE_SPAN},x
             
             ldal  spritedata+{]line*SPRITE_PLANE_SPAN},x
             and:  tiledata+32+{]line*4},y
@@ -619,8 +625,8 @@ _DrawTile16x16
 ]line       equ   0
             lup   8
             lda:  tiledata+32+{]line*4},y
-            andl  spritemask+{]line*256},x
-            stal  spritemask+{]line*256},x
+            andl  spritemask+{]line*SPRITE_PLANE_SPAN},x
+            stal  spritemask+{]line*SPRITE_PLANE_SPAN},x
             
             ldal  spritedata+{]line*SPRITE_PLANE_SPAN},x
             and:  tiledata+32+{]line*4},y
@@ -663,8 +669,8 @@ SPRITE_ROW_STRIDE equ 8*SPRITE_PLANE_SPAN
 ]line       equ   0
             lup   8
             lda:  tiledata+TILE_ROW_STRIDE+32+{]line*4},y
-            andl  spritemask+SPRITE_ROW_STRIDE+{]line*256},x
-            stal  spritemask+SPRITE_ROW_STRIDE+{]line*256},x
+            andl  spritemask+SPRITE_ROW_STRIDE+{]line*SPRITE_PLANE_SPAN},x
+            stal  spritemask+SPRITE_ROW_STRIDE+{]line*SPRITE_PLANE_SPAN},x
             
             ldal  spritedata+SPRITE_ROW_STRIDE+{]line*SPRITE_PLANE_SPAN},x
             and:  tiledata+TILE_ROW_STRIDE+32+{]line*4},y
@@ -717,8 +723,8 @@ _DrawTile8x8V
 ]line       equ   0
             lup   8
             lda:  tiledata+32+{{7-]line}*4},y
-            andl  spritemask+{]line*256},x
-            stal  spritemask+{]line*256},x
+            andl  spritemask+{]line*SPRITE_PLANE_SPAN},x
+            stal  spritemask+{]line*SPRITE_PLANE_SPAN},x
             
             ldal  spritedata+{]line*SPRITE_PLANE_SPAN},x
             and:  tiledata+32+{{7-]line}*4},y
@@ -744,13 +750,20 @@ _DrawTile8x8V
 ; resgion to all $FFFF
 ; 
 ; X = address is sprite plane -- erases an 8x8 region
+_CacheSpriteBanks
+            lda    #>spritemask
+            and    #$FF00
+            ora    #^spritedata
+            sta    SpriteBanks
+            rts
+
 SPRITE_PLANE_SPAN equ 256
 
 _EraseTileSprite
             phb                                   ; Save the bank to switch to the sprite plane
 
-            pea    #^spritedata
-            plb
+            pei    SpriteBanks
+            plb                                   ; pop the data bank (low byte)
 
             lda    #0
             sta:   {0*SPRITE_PLANE_SPAN}+0,x
@@ -770,8 +783,7 @@ _EraseTileSprite
             sta:   {7*SPRITE_PLANE_SPAN}+0,x
             sta:   {7*SPRITE_PLANE_SPAN}+2,x
 
-            pea    #^spritemask
-            plb
+            plb                                  ; pop the mask bank (high byte)
 
             lda    #$FFFF
             sta:   {0*SPRITE_PLANE_SPAN}+0,x
@@ -791,7 +803,6 @@ _EraseTileSprite
             sta:   {7*SPRITE_PLANE_SPAN}+0,x
             sta:   {7*SPRITE_PLANE_SPAN}+2,x
 
-            pla
             plb
             rts
 
diff --git a/src/Tool.s b/src/Tool.s
new file mode 100644
index 0000000..c1ba0dd
--- /dev/null
+++ b/src/Tool.s
@@ -0,0 +1,131 @@
+; Toolbox wrapper for the GTE library.  Implemented as a user tool
+;
+; Ref: Toolbox Reference, Volume 2, Appendix A
+; Ref: IIgs Tech Note #73
+
+ToStrip         equ   $E10184
+
+_CallTable
+                dw    {_CTEnd-_CallTable}/4,0
+                adrl  _TSBootInit-1
+                adrl  _TSStartUp-1
+                adrl  _TSShutDown-1
+                adrl  _TSVersion-1
+                adrl  _TSReset-1
+                adrl  _TSStatus-1
+                adrl  _TSReserved-1
+                adrl  _TSReserved-1
+
+                adrl  _TSSetScreenMode
+_CTEnd
+
+; Do nothing when the tool set is installed
+_TSBootInit
+                lda   #0
+                clc
+                rtl
+
+; Call the regular GTE startup function after setting the Work Area Point (WAP).  The caller much provide
+; one page of Bank 0 memory for the tool set's private use
+;
+; X = 
+_TSStartUp
+:zpToUse        equ     7
+
+                pea     #$8000
+                txa
+                and     #$00FF
+                pha
+
+                pea     $0000
+                lda     :zpToUse+6,s
+                pha
+                _SetWAP
+
+                jsr     _CoreStartUp
+
+_TSShutDown
+                cmp     #0                 ; Acc is low word of the WAP (direct page)
+                beq     :inactive
+
+                phd
+                pha
+                pld                        ; Set the direct page for the toolset
+
+                phx                        ; Preserve the X register
+                jsr     _CoreShutDown      ; Shut down GTE
+                pla
+
+                pea     $8000
+                and     #$00FF
+                pha
+                pea     $0000              ; Set WAP to null
+                pea     $0000
+                _SetWAP
+
+                pld                        ; Restore the direct page
+
+:inactive
+                lda     #0
+                clc
+                rtl
+
+_TSVersion
+                lda     #$0100     ; Version 1
+                sta     7,s
+
+                lda     #0
+                clc
+                rtl
+
+_TSReset
+                lda     #0
+                clc
+                rtl
+
+; Check the WAP values in the A, Y registers
+_TSStatus
+                sta    1,s
+                tya
+                ora    1,s
+                sta    1,s        ; 0 if WAP is null, non-zero if WAP is set
+
+                lda     #0
+                clc
+                rtl
+
+_TSReserved
+                txa
+                xba
+                ora     #$00FF            ; Put error code $FF in the accumulator
+                sec
+                rtl
+
+_TSSetScreenMode
+                phd                       ; Preserve the direct page
+                pha
+                pld
+
+                lda     9,s
+                tay
+                lda     9,s
+                tax
+                jsr     _SetScreenMode
+                pld
+
+                ldx     #0                 ; No error
+                ldy     #4                 ; Remove the 4 input bytes
+                jml     ToStrip
+
+_TSReadControl
+                phd                       ; Preserve the direct page
+                pha
+                pld
+
+                jsr     _ReadControl
+                sta     9,s
+
+                pld
+                ldx     #0                 ; No error
+                ldy     #0                 ; Remove zero input bytes
+                jml     ToStrip
diff --git a/src/blitter/Template.s b/src/blitter/Template.s
index cdb36b4..8dea8d2 100644
--- a/src/blitter/Template.s
+++ b/src/blitter/Template.s
@@ -526,11 +526,11 @@ BuildBank
 ; to cover the full screen vertical scrolling.
 ;
 ; The 'base' location is always assumed to be on a 4kb ($1000) boundary.  We make sure that
-; the code is assembled on a page boundary to help will alignment
+; the code is assembled on a page boundary to help with alignment
                    ds    \,$00                      ; pad to the next page boundary
 base
-entry_1            ldx   #0000                      ; Used for LDA 00,x addressing
-entry_2            ldy   #0000                      ; Used for LDA (00),y addressing
+entry_1            ldx   #0000                      ; Used for LDA 00,x addressing (Dynamic Tiles)
+entry_2            ldy   #0000                      ; Used for LDA (00),y addressing (Second Layer; BG1)
 entry_3            lda   #0000                      ; Sets screen address (right edge)
                    tcs
 
@@ -691,22 +691,17 @@ epilogue_1         tsc
 ; Standard Two-level Mix (23 bytes)
 ;
 ;   Optimal     = 18 cycles (LDA/AND/ORA/PHA/JMP)
-;  16-bit write = 21 cycles 
-;   8-bit low   = 30 cycles
-;   8-bit high  = 29 cycles
+;  16-bit write = 20 cycles 
+;   8-bit low   = 28 cycles
+;   8-bit high  = 27 cycles
 ;
 ;  start     lda  (00),y        ; 6
 ;            and  #MASK         ; 3
 ;            ora  #DATA         ; 3 = 12 cycles to load the data
 ;            bcs  alt_exit      ; 2/3
 ;            pha                ; 4
-;  out       brl  next          ; 4 Fast-path completes in 5 additional cycles
-;
-;  alt_exit  bvs  r_edge        ; 2/3 
-;            clc                ; 2
-;            brl  l_jmp_rtn     ; 3
-;  r_edge    rep   #$41
-;            brl  r_jmp_rtn     ; 3
+;  out       jmp  next          ; 3 Fast-path completes in 5 additional cycles
+;  alt_exit  jmp  jmp_rtn       ; 3 
 ;
 ;
 ; For dynamic masked tiles, we re-write bytes 2 - 8 as this, which mostly
@@ -716,7 +711,6 @@ epilogue_1         tsc
 ;  start     lda  (00),y        ; 6
 ;            and  $80,x         ; 5
 ;            ora  $00,x         ; 5 = 16 cycles to load the data
-;            bcc  *+4           ; 
 ;            bcs  alt_exit      ; 2/3
 ;            pha
 ;            ...