diff --git a/Processors/pio_16bit_parallel.pio b/Processors/pio_16bit_parallel.pio
index b0b3599..bca2dc6 100644
--- a/Processors/pio_16bit_parallel.pio
+++ b/Processors/pio_16bit_parallel.pio
@@ -33,6 +33,8 @@ next:
    jmp y--, next side 1
 
 .wrap_target
+// Transmit an 8 bit value (LS 8 bits of 32 bits).
+public start_8:
 // Transmit a 16 bit value (LS 16 bits of 32 bits).
 public start_tx:
    // Fetch the next 32 bit value from the TX FIFO and set TFT_WR high.
@@ -43,16 +45,6 @@ public start_tx:
    // Set WR high and jump back to start.
    jmp start_tx side 1
 
-// Transmit an 8 bit value (LS 8 bits of 32 bits).
-public start_8:
-   // Fetch the next 32 bit value from the TX FIFO and set TFT_WR high.
-   pull side 1
-   // Write the first byte (LSB) and sets WR low. This also 
-   // shifts the unused top 16 bits through.
-   out pins, 32 side 0 [1]     
-   // Jump to start
-   jmp start_tx side 1
-
 // Transmit a set window command sequence.
 public set_addr_window:
    // Loop count in x (to send caset, paset and ramwr commands).
@@ -75,7 +67,7 @@ send_xy:
    out pins, 8 side 0 [1]
    // Loop until 4 bytes sent, TFT_WR high.
    jmp !osre, send_xy side 1 [1]
-end_set_addr:
+end_set_addr: // Jump here since delay needed before DC change
    // Loop back for next command and write last command.
    jmp x--, pull_cmd side 1
    // Set DC high.
diff --git a/Processors/pio_16bit_parallel.pio.h b/Processors/pio_16bit_parallel.pio.h
index b9cd4c7..6d792f0 100644
--- a/Processors/pio_16bit_parallel.pio.h
+++ b/Processors/pio_16bit_parallel.pio.h
@@ -13,12 +13,12 @@
 // ------ //
 
 #define tft_io_wrap_target 7
-#define tft_io_wrap 23
+#define tft_io_wrap 20
 
 #define tft_io_offset_block_fill 0u
+#define tft_io_offset_start_8 7u
 #define tft_io_offset_start_tx 7u
-#define tft_io_offset_start_8 10u
-#define tft_io_offset_set_addr_window 13u
+#define tft_io_offset_set_addr_window 10u
 
 static const uint16_t tft_io_program_instructions[] = {
     0x98a0, //  0: pull   block           side 1     
@@ -32,27 +32,24 @@ static const uint16_t tft_io_program_instructions[] = {
     0x98a0, //  7: pull   block           side 1     
     0x7100, //  8: out    pins, 32        side 0 [1] 
     0x1807, //  9: jmp    7               side 1     
-    0x98a0, // 10: pull   block           side 1     
-    0x7100, // 11: out    pins, 32        side 0 [1] 
-    0x1807, // 12: jmp    7               side 1     
-    0xf822, // 13: set    x, 2            side 1     
-    0xe000, // 14: set    pins, 0                    
-    0x80a0, // 15: pull   block                      
-    0x7000, // 16: out    pins, 32        side 0     
-    0x0036, // 17: jmp    !x, 22                     
-    0x98a0, // 18: pull   block           side 1     
-    0xe001, // 19: set    pins, 1                    
-    0x7108, // 20: out    pins, 8         side 0 [1] 
-    0x19f4, // 21: jmp    !osre, 20       side 1 [1] 
-    0x184e, // 22: jmp    x--, 14         side 1     
-    0xe001, // 23: set    pins, 1                    
+    0xf822, // 10: set    x, 2            side 1     
+    0xe000, // 11: set    pins, 0                    
+    0x80a0, // 12: pull   block                      
+    0x7000, // 13: out    pins, 32        side 0     
+    0x0033, // 14: jmp    !x, 19                     
+    0x98a0, // 15: pull   block           side 1     
+    0xe001, // 16: set    pins, 1                    
+    0x7108, // 17: out    pins, 8         side 0 [1] 
+    0x19f1, // 18: jmp    !osre, 17       side 1 [1] 
+    0x184b, // 19: jmp    x--, 11         side 1     
+    0xe001, // 20: set    pins, 1                    
             //     .wrap
 };
 
 #if !PICO_NO_HARDWARE
 static const struct pio_program tft_io_program = {
     .instructions = tft_io_program_instructions,
-    .length = 24,
+    .length = 21,
     .origin = -1,
 };