Merge branch 'master' into dev

14 files changed, 555 insertions, 536 deletions

diff --git a/CMakeLists.txt b/CMakeLists.txt
new file mode 100644
index 0000000..0b569aa
--- /dev/null
+++ b/CMakeLists.txt
@@ -0,0 +1,25 @@
+# HUB75 RGB LED matrix library utilizing ESP32 DMA Engine

+# https://github.com/mrfaptastic/ESP32-HUB75-MatrixPanel-I2S-DMA

+# MIT License

+

+cmake_minimum_required(VERSION 3.5)

+

+idf_component_register(SRCS "esp32_i2s_parallel_dma.c" "ESP32-HUB75-MatrixPanel-I2S-DMA.cpp" "ESP32-HUB75-MatrixPanel-leddrivers.cpp"

+                       INCLUDE_DIRS "."

+                       REQUIRES arduino)

+

+# In case you are running into issues with "missing" header files from 3rd party libraries like "Adafruit_GFX.h"

+# you can add them to the REQUIRES section above (depending on how you added the library as component!) like

+# REQUIRES arduino Adafruit-GFX-Library)

+

+# Example to build with USE_GFX_ROOT or NO_GFX / just uncomment the appropriate line

+# target_compile_options(${COMPONENT_TARGET} PUBLIC -DUSE_GFX_ROOT)

+# target_compile_options(${COMPONENT_TARGET} PUBLIC -DNO_GFX)

+

+# You can also use multiple options like this

+# target_compile_options(${COMPONENT_TARGET} PUBLIC -DNO_GFX -DNO_FAST_FUNCTIONS)

+

+# All options can be found here:

+# https://github.com/mrfaptastic/ESP32-HUB75-MatrixPanel-I2S-DMA/blob/master/doc/BuildOptions.md

+

+project(ESP32-HUB75-MatrixPanel-I2S-DMA)

diff --git a/ESP32-HUB75-MatrixPanel-I2S-DMA.cpp b/ESP32-HUB75-MatrixPanel-I2S-DMA.cpp
index a9fae2f..60df498 100644
--- a/ESP32-HUB75-MatrixPanel-I2S-DMA.cpp
+++ b/ESP32-HUB75-MatrixPanel-I2S-DMA.cpp
@@ -3,18 +3,18 @@
 
 
 #if defined(ESP32_SXXX)
-	#pragma message "Compiling for ESP32-Sx MCUs"	
+    #pragma message "Compiling for ESP32-Sx MCUs"   
 #elif defined(ESP32_CXXX)
-	#pragma message "Compiling for ESP32-Cx MCUs"		
+    #pragma message "Compiling for ESP32-Cx MCUs"       
 #elif CONFIG_IDF_TARGET_ESP32 || defined(ESP32)
-	#pragma message "Compiling for original 520kB SRAM ESP32."	
+    #pragma message "Compiling for original (released 2016) 520kB SRAM ESP32."  
 #else
     #error "Compiling for something unknown!"
 #endif
 
 
 // Credits: Louis Beaudoin <https://github.com/pixelmatix/SmartMatrix/tree/teensylc>
-// and Sprite_TM: 			https://www.esp32.com/viewtopic.php?f=17&t=3188 and https://www.esp32.com/viewtopic.php?f=13&t=3256
+// and Sprite_TM:           https://www.esp32.com/viewtopic.php?f=17&t=3188 and https://www.esp32.com/viewtopic.php?f=13&t=3256
 
 /*
 
@@ -89,7 +89,7 @@ bool MatrixPanel_I2S_DMA::allocateDMAmemory()
     size_t _dma_linked_list_memory_required     = 0; 
     size_t _total_dma_capable_memory_reserved   = 0;   
     
-	// 1. Calculate the amount of DMA capable memory that's actually available
+    // 1. Calculate the amount of DMA capable memory that's actually available
     #if SERIAL_DEBUG    
         Serial.printf_P(PSTR("Panel Width: %d pixels.\r\n"),  PIXELS_PER_ROW);
         Serial.printf_P(PSTR("Panel Height: %d pixels.\r\n"), m_cfg.mx_height);
@@ -100,12 +100,12 @@ bool MatrixPanel_I2S_DMA::allocateDMAmemory()
         
         Serial.println(F("DMA memory blocks available before any malloc's: "));
         heap_caps_print_heap_info(MALLOC_CAP_DMA);
-		Serial.println(F("******************************************************************"));
+        Serial.println(F("******************************************************************"));
         Serial.printf_P(PSTR("We're going to need %d bytes of SRAM just for the frame buffer(s).\r\n"), _frame_buffer_memory_required);  
         Serial.printf_P(PSTR("The total amount of DMA capable SRAM memory is %d bytes.\r\n"), heap_caps_get_free_size(MALLOC_CAP_DMA));
         Serial.printf_P(PSTR("Largest DMA capable SRAM memory block is %d bytes.\r\n"), heap_caps_get_largest_free_block(MALLOC_CAP_DMA));          
-		Serial.println(F("******************************************************************"));	    
-		
+        Serial.println(F("******************************************************************"));        
+        
     #endif
 
     // Can we potentially fit the framebuffer into the DMA capable memory that's available?
@@ -117,21 +117,21 @@ bool MatrixPanel_I2S_DMA::allocateDMAmemory()
 
       return false;
     }
-	
-	// Alright, theoretically we should be OK, so let us do this, so
-	// lets allocate a chunk of memory for each row (a row could span multiple panels if chaining is in place)
+    
+    // Alright, theoretically we should be OK, so let us do this, so
+    // lets allocate a chunk of memory for each row (a row could span multiple panels if chaining is in place)
   dma_buff.rowBits.reserve(ROWS_PER_FRAME);
 
   // iterate through number of rows
-	for (int malloc_num =0; malloc_num < ROWS_PER_FRAME; ++malloc_num)
-	{
+    for (int malloc_num =0; malloc_num < ROWS_PER_FRAME; ++malloc_num)
+    {
     auto ptr = std::make_shared<rowBitStruct>(PIXELS_PER_ROW, PIXEL_COLOR_DEPTH_BITS, m_cfg.double_buff);
 
     if (ptr->data == nullptr){
       #if SERIAL_DEBUG
-		      Serial.printf_P(PSTR("ERROR: Couldn't malloc rowBitStruct %d! Critical fail.\r\n"), malloc_num);
+              Serial.printf_P(PSTR("ERROR: Couldn't malloc rowBitStruct %d! Critical fail.\r\n"), malloc_num);
       #endif
-    			return false;
+                return false;
           // TODO: should we release all previous rowBitStructs here???
     }
 
@@ -141,7 +141,7 @@ bool MatrixPanel_I2S_DMA::allocateDMAmemory()
         Serial.printf_P(PSTR("Malloc'ing %d bytes of memory @ address %ud for frame row %d.\r\n"), ptr->size()*_num_frame_buffers, (unsigned int)ptr->getDataPtr(), malloc_num);
      #endif
 
-	}
+    }
 
     _total_dma_capable_memory_reserved += _frame_buffer_memory_required;    
 
@@ -156,7 +156,7 @@ bool MatrixPanel_I2S_DMA::allocateDMAmemory()
     // aka. Calculate the lowest LSBMSB_TRANSITION_BIT value that will fit in memory
     int numDMAdescriptorsPerRow = 0;
     lsbMsbTransitionBit = 0;
-	
+    
 
     while(1) {
         numDMAdescriptorsPerRow = 1;
@@ -184,7 +184,7 @@ bool MatrixPanel_I2S_DMA::allocateDMAmemory()
     #endif
 
 
-   #ifndef IGNORE_REFRESH_RATE	
+   #ifndef IGNORE_REFRESH_RATE  
     // calculate the lowest LSBMSB_TRANSITION_BIT value that will fit in memory that will meet or exceed the configured refresh rate
     while(1) {           
         int psPerClock = 1000000000000UL/m_cfg.i2sspeed;
@@ -203,7 +203,7 @@ bool MatrixPanel_I2S_DMA::allocateDMAmemory()
 
         #if SERIAL_DEBUG  
           Serial.printf_P(PSTR("lsbMsbTransitionBit of %d gives %d Hz refresh: \r\n"), lsbMsbTransitionBit, actualRefreshRate);        
-		    #endif
+            #endif
 
         if (actualRefreshRate > m_cfg.min_refresh_rate) 
           break;
@@ -218,7 +218,7 @@ bool MatrixPanel_I2S_DMA::allocateDMAmemory()
     Serial.printf_P(PSTR("Raised lsbMsbTransitionBit to %d/%d to meet minimum refresh rate\r\n"), lsbMsbTransitionBit, PIXEL_COLOR_DEPTH_BITS - 1);
     #endif
 
-	#endif
+    #endif
 
   /***
    * Step 2a: lsbMsbTransition bit is now finalised - recalculate the DMA descriptor count required, which is used for
@@ -238,7 +238,7 @@ bool MatrixPanel_I2S_DMA::allocateDMAmemory()
 
         #if SERIAL_DEBUG  
           Serial.printf_P(PSTR("rowColorDepthStruct struct is too large, split DMA payload required. Adding %d DMA descriptors\n"), PIXEL_COLOR_DEPTH_BITS-1);
-		    #endif
+            #endif
 
         numDMAdescriptorsPerRow += PIXEL_COLOR_DEPTH_BITS-1; 
         // Note: If numDMAdescriptorsPerRow is even just one descriptor too large, DMA linked list will not correctly loop.
@@ -251,14 +251,14 @@ bool MatrixPanel_I2S_DMA::allocateDMAmemory()
 
     _dma_linked_list_memory_required = numDMAdescriptorsPerRow * ROWS_PER_FRAME * _num_frame_buffers * sizeof(lldesc_t);
   #if SERIAL_DEBUG
-		Serial.printf_P(PSTR("Descriptors for lsbMsbTransitionBit of %d/%d with %d frame rows require %d bytes of DMA RAM with %d numDMAdescriptorsPerRow.\r\n"), lsbMsbTransitionBit, PIXEL_COLOR_DEPTH_BITS - 1, ROWS_PER_FRAME, _dma_linked_list_memory_required, numDMAdescriptorsPerRow);    
-	#endif   
+        Serial.printf_P(PSTR("Descriptors for lsbMsbTransitionBit of %d/%d with %d frame rows require %d bytes of DMA RAM with %d numDMAdescriptorsPerRow.\r\n"), lsbMsbTransitionBit, PIXEL_COLOR_DEPTH_BITS - 1, ROWS_PER_FRAME, _dma_linked_list_memory_required, numDMAdescriptorsPerRow);    
+    #endif   
 
     _total_dma_capable_memory_reserved += _dma_linked_list_memory_required;
 
     // Do a final check to see if we have enough space for the additional DMA linked list descriptors that will be required to link it all up!
     if(_dma_linked_list_memory_required > heap_caps_get_largest_free_block(MALLOC_CAP_DMA)) {
-#if SERIAL_DEBUG     		
+#if SERIAL_DEBUG            
        Serial.println(F("ERROR: Not enough SRAM left over for DMA linked-list descriptor memory reservation! Oh so close!\r\n"));
 #endif  
         return false;
@@ -271,21 +271,21 @@ bool MatrixPanel_I2S_DMA::allocateDMAmemory()
     dmadesc_a = (lldesc_t *)heap_caps_malloc(desccount * sizeof(lldesc_t), MALLOC_CAP_DMA);
     assert("Can't allocate descriptor framebuffer a");
     if(!dmadesc_a) {
-#if SERIAL_DEBUG     		
+#if SERIAL_DEBUG            
         Serial.println(F("ERROR: Could not malloc descriptor framebuffer a."));
-#endif		
+#endif      
         return false;
     }
-	
+    
     if (m_cfg.double_buff) // reserve space for second framebuffer linked list
     {
         //lldesc_t * dmadesc_b = (lldesc_t *)heap_caps_malloc(desccount * sizeof(lldesc_t), MALLOC_CAP_DMA);
         dmadesc_b = (lldesc_t *)heap_caps_malloc(desccount * sizeof(lldesc_t), MALLOC_CAP_DMA);
         assert("Could not malloc descriptor framebuffer b.");
         if(!dmadesc_b) {
-#if SERIAL_DEBUG     			
+#if SERIAL_DEBUG                
             Serial.println(F("ERROR: Could not malloc descriptor framebuffer b."));
-#endif			
+#endif          
             return false;
         }
     }
@@ -294,13 +294,13 @@ bool MatrixPanel_I2S_DMA::allocateDMAmemory()
     Serial.println(F("*** ESP32-HUB75-MatrixPanel-I2S-DMA: Memory Allocations Complete ***"));
     Serial.printf_P(PSTR("Total memory that was reserved: %d kB.\r\n"), _total_dma_capable_memory_reserved/1024);
     Serial.printf_P(PSTR("... of which was used for the DMA Linked List(s): %d kB.\r\n"), _dma_linked_list_memory_required/1024);
-	
+    
     Serial.printf_P(PSTR("Heap Memory Available: %d bytes total. Largest free block: %d bytes.\r\n"), heap_caps_get_free_size(0), heap_caps_get_largest_free_block(0));
     Serial.printf_P(PSTR("General RAM Available: %d bytes total. Largest free block: %d bytes.\r\n"), heap_caps_get_free_size(MALLOC_CAP_DEFAULT), heap_caps_get_largest_free_block(MALLOC_CAP_DEFAULT));
 #endif
 
     // Just os we know
-  	initialized = true;
+    initialized = true;
 
     return true;
 
@@ -397,11 +397,11 @@ void MatrixPanel_I2S_DMA::configureDMA(const HUB75_I2S_CFG& _cfg)
 
    #if SERIAL_DEBUG  
       Serial.printf_P(PSTR("configureDMA(): Configured LL structure. %d DMA Linked List descriptors populated.\r\n"), current_dmadescriptor_offset);
-	  
-	  if ( desccount != current_dmadescriptor_offset)
-	  {
-		Serial.printf_P(PSTR("configureDMA(): ERROR! Expected descriptor count of %d != actual DMA descriptors of %d!\r\n"), desccount, current_dmadescriptor_offset);		  
-	  }
+      
+      if ( desccount != current_dmadescriptor_offset)
+      {
+        Serial.printf_P(PSTR("configureDMA(): ERROR! Expected descriptor count of %d != actual DMA descriptors of %d!\r\n"), desccount, current_dmadescriptor_offset);        
+      }
     #endif  
 
     //End markers for DMA LL
@@ -429,9 +429,9 @@ void MatrixPanel_I2S_DMA::configureDMA(const HUB75_I2S_CFG& _cfg)
         .desccount_b=desccount,
         .lldesc_b=dmadesc_b,
         .clkphase=_cfg.clkphase,
-		.int_ena_out_eof=_cfg.double_buff
+        .int_ena_out_eof=_cfg.double_buff
     };
-	
+    
     // Setup I2S
     i2s_parallel_driver_install(ESP32_I2S_DEVICE, &dma_cfg);
     i2s_parallel_send_dma(ESP32_I2S_DEVICE, &dmadesc_a[0]);
@@ -439,12 +439,12 @@ void MatrixPanel_I2S_DMA::configureDMA(const HUB75_I2S_CFG& _cfg)
     #if SERIAL_DEBUG  
       Serial.println(F("configureDMA(): DMA setup completed on ESP32_I2S_DEVICE.")); 
     #endif       
-		
+        
 } // end initMatrixDMABuff
 
 
 /* There are 'bits' set in the frameStruct that we simply don't need to set every single time we change a pixel / DMA buffer co-ordinate.
- * 	For example, the bits that determine the address lines, we don't need to set these every time. Once they're in place, and assuming we
+ *  For example, the bits that determine the address lines, we don't need to set these every time. Once they're in place, and assuming we
  *  don't accidentally clear them, then we don't need to set them again.
  *  So to save processing, we strip this logic out to the absolute bare minimum, which is toggling only the R,G,B pixels (bits) per co-ord.
  *
@@ -477,36 +477,36 @@ void IRAM_ATTR MatrixPanel_I2S_DMA::updateMatrixDMABuffer(int16_t x_coord, int16
   }
 
   /* LED Brightness Compensation. Because if we do a basic "red & mask" for example, 
-	 * we'll NEVER send the dimmest possible colour, due to binary skew.
-	 * i.e. It's almost impossible for color_depth_idx of 0 to be sent out to the MATRIX unless the 'value' of a color is exactly '1'
+     * we'll NEVER send the dimmest possible colour, due to binary skew.
+     * i.e. It's almost impossible for color_depth_idx of 0 to be sent out to the MATRIX unless the 'value' of a color is exactly '1'
    * https://ledshield.wordpress.com/2012/11/13/led-brightness-to-your-eye-gamma-correction-no/
-	 */
+     */
 #ifndef NO_CIE1931
-	red   = lumConvTab[red];
-	green = lumConvTab[green];
-	blue  = lumConvTab[blue]; 	
+    red   = lumConvTab[red];
+    green = lumConvTab[green];
+    blue  = lumConvTab[blue];   
 #endif
 
-	/* When using the drawPixel, we are obviously only changing the value of one x,y position, 
-	 * however, the two-scan panels paint TWO lines at the same time
-	 * and this reflects the parallel in-DMA-memory data structure of uint16_t's that are getting
-	 * pumped out at high speed.
-	 * 
-	 * So we need to ensure we persist the bits (8 of them) of the uint16_t for the row we aren't changing.
-	 * 
-	 * The DMA buffer order has also been reversed (refer to the last code in this function)
-	 * so we have to check for this and check the correct position of the MATRIX_DATA_STORAGE_TYPE
-	 * data.
-	 */
+    /* When using the drawPixel, we are obviously only changing the value of one x,y position, 
+     * however, the two-scan panels paint TWO lines at the same time
+     * and this reflects the parallel in-DMA-memory data structure of uint16_t's that are getting
+     * pumped out at high speed.
+     * 
+     * So we need to ensure we persist the bits (8 of them) of the uint16_t for the row we aren't changing.
+     * 
+     * The DMA buffer order has also been reversed (refer to the last code in this function)
+     * so we have to check for this and check the correct position of the MATRIX_DATA_STORAGE_TYPE
+     * data.
+     */
 
 #ifndef ESP32_SXXX
     // We need to update the correct uint16_t in the rowBitStruct array, that gets sent out in parallel
     // 16 bit parallel mode - Save the calculated value to the bitplane memory in reverse order to account for I2S Tx FIFO mode1 ordering
-	// Irrelevant for ESP32-S2 the way the FIFO ordering works is different - refer to page 679 of S2 technical reference manual
+    // Irrelevant for ESP32-S2 the way the FIFO ordering works is different - refer to page 679 of S2 technical reference manual
     x_coord & 1U ? --x_coord : ++x_coord;
 #endif 
 
-	
+    
     uint16_t _colorbitclear = BITMASK_RGB1_CLEAR, _colorbitoffset = 0;
 
     if (y_coord >= ROWS_PER_FRAME){    // if we are drawing to the bottom part of the panel
@@ -520,15 +520,15 @@ void IRAM_ATTR MatrixPanel_I2S_DMA::updateMatrixDMABuffer(int16_t x_coord, int16
     do {
         --color_depth_idx;
 //        uint8_t mask = (1 << (color_depth_idx COLOR_DEPTH_LESS_THAN_8BIT_ADJUST)); // expect 24 bit color (8 bits per RGB subpixel)
-	#if PIXEL_COLOR_DEPTH_BITS < 8
-		uint8_t mask = (1 << (color_depth_idx+MASK_OFFSET)); // expect 24 bit color (8 bits per RGB subpixel)
-	#else
-		uint8_t mask = (1 << (color_depth_idx)); // expect 24 bit color (8 bits per RGB subpixel)
-	#endif		
+    #if PIXEL_COLOR_DEPTH_BITS < 8
+        uint8_t mask = (1 << (color_depth_idx+MASK_OFFSET)); // expect 24 bit color (8 bits per RGB subpixel)
+    #else
+        uint8_t mask = (1 << (color_depth_idx)); // expect 24 bit color (8 bits per RGB subpixel)
+    #endif      
         uint16_t RGB_output_bits = 0;
 
         /* Per the .h file, the order of the output RGB bits is:
-          * BIT_B2, BIT_G2, BIT_R2,    BIT_B1, BIT_G1, BIT_R1	  */
+          * BIT_B2, BIT_G2, BIT_R2,    BIT_B1, BIT_G1, BIT_R1     */
         RGB_output_bits |= (bool)(blue & mask);   // --B
         RGB_output_bits <<= 1;
         RGB_output_bits |= (bool)(green & mask);  // -BG
@@ -555,11 +555,11 @@ void MatrixPanel_I2S_DMA::updateMatrixDMABuffer(uint8_t red, uint8_t green, uint
 {
   if ( !initialized ) return;
   
-	/* https://ledshield.wordpress.com/2012/11/13/led-brightness-to-your-eye-gamma-correction-no/ */	 
+    /* https://ledshield.wordpress.com/2012/11/13/led-brightness-to-your-eye-gamma-correction-no/ */     
 #ifndef NO_CIE1931
-	red 	= lumConvTab[red];
-	green	= lumConvTab[green];
-	blue 	= lumConvTab[blue]; 	
+    red     = lumConvTab[red];
+    green   = lumConvTab[green];
+    blue    = lumConvTab[blue];     
 #endif
 
   for(uint8_t color_depth_idx=0; color_depth_idx<PIXEL_COLOR_DEPTH_BITS; color_depth_idx++)  // color depth - 8 iterations
@@ -567,23 +567,23 @@ void MatrixPanel_I2S_DMA::updateMatrixDMABuffer(uint8_t red, uint8_t green, uint
     // let's precalculate RGB1 and RGB2 bits than flood it over the entire DMA buffer
     uint16_t RGB_output_bits = 0;
 //    uint8_t mask = (1 << color_depth_idx COLOR_DEPTH_LESS_THAN_8BIT_ADJUST); // 24 bit color
-	#if PIXEL_COLOR_DEPTH_BITS < 8
-		uint8_t mask = (1 << (color_depth_idx+MASK_OFFSET)); // expect 24 bit color (8 bits per RGB subpixel)
-	#else
-		uint8_t mask = (1 << (color_depth_idx)); // expect 24 bit color (8 bits per RGB subpixel)
-	#endif		
+    #if PIXEL_COLOR_DEPTH_BITS < 8
+        uint8_t mask = (1 << (color_depth_idx+MASK_OFFSET)); // expect 24 bit color (8 bits per RGB subpixel)
+    #else
+        uint8_t mask = (1 << (color_depth_idx)); // expect 24 bit color (8 bits per RGB subpixel)
+    #endif      
 
-	/* Per the .h file, the order of the output RGB bits is:
-	 * BIT_B2, BIT_G2, BIT_R2,    BIT_B1, BIT_G1, BIT_R1	  */
+    /* Per the .h file, the order of the output RGB bits is:
+     * BIT_B2, BIT_G2, BIT_R2,    BIT_B1, BIT_G1, BIT_R1      */
     RGB_output_bits |= (bool)(blue & mask);   // --B
     RGB_output_bits <<= 1;
     RGB_output_bits |= (bool)(green & mask);  // -BG
     RGB_output_bits <<= 1;
     RGB_output_bits |= (bool)(red & mask);    // BGR
-	
-	// Duplicate and shift across so we have have 6 populated bits of RGB1 and RGB2 pin values suitable for DMA buffer
+    
+    // Duplicate and shift across so we have have 6 populated bits of RGB1 and RGB2 pin values suitable for DMA buffer
     RGB_output_bits |= RGB_output_bits << BITS_RGB2_OFFSET;  //BGRBGR
-	
+    
     //Serial.printf("Fill with: 0x%#06x\n", RGB_output_bits);
 
     // iterate rows
@@ -622,7 +622,7 @@ void MatrixPanel_I2S_DMA::clearFrameBuffer(bool _buff_id){
   int row_idx = dma_buff.rowBits.size();
   do {
     --row_idx;
-	
+    
     ESP32_I2S_DMA_STORAGE_TYPE* row = dma_buff.rowBits[row_idx]->getDataPtr(0, _buff_id);   // set pointer to the HEAD of a buffer holding data for the entire matrix row
 
     ESP32_I2S_DMA_STORAGE_TYPE abcde = (ESP32_I2S_DMA_STORAGE_TYPE)row_idx;
@@ -635,15 +635,15 @@ void MatrixPanel_I2S_DMA::clearFrameBuffer(bool _buff_id){
     // fill all x_pixels except color_index[0] (LSB) ones, this also clears all color data to 0's black
     do {
       --x_pixel;
-	  
-	  if ( m_cfg.driver == HUB75_I2S_CFG::SM5266P) {
-		// modifications here for row shift register type SM5266P 
-		// https://github.com/mrfaptastic/ESP32-HUB75-MatrixPanel-I2S-DMA/issues/164
-		row[x_pixel] = abcde & (0x18 << BITS_ADDR_OFFSET); // mask out the bottom 3 bits which are the clk di bk inputs  
-	  } else {		  
-		row[x_pixel] = abcde;
-	  }
-	  
+      
+      if ( m_cfg.driver == HUB75_I2S_CFG::SM5266P) {
+        // modifications here for row shift register type SM5266P 
+        // https://github.com/mrfaptastic/ESP32-HUB75-MatrixPanel-I2S-DMA/issues/164
+        row[x_pixel] = abcde & (0x18 << BITS_ADDR_OFFSET); // mask out the bottom 3 bits which are the clk di bk inputs  
+      } else {        
+        row[x_pixel] = abcde;
+      }
+      
     } while(x_pixel!=dma_buff.rowBits[row_idx]->width);
 
     // color_index[0] (LSB) x_pixels must be "marked" with a previous's row address, 'cause  it is used to display
@@ -651,52 +651,52 @@ void MatrixPanel_I2S_DMA::clearFrameBuffer(bool _buff_id){
     abcde = ((ESP32_I2S_DMA_STORAGE_TYPE)row_idx-1) << BITS_ADDR_OFFSET;
     do {
       --x_pixel;
-	  
-	  if ( m_cfg.driver == HUB75_I2S_CFG::SM5266P) {
-		// modifications here for row shift register type SM5266P 
-		// https://github.com/mrfaptastic/ESP32-HUB75-MatrixPanel-I2S-DMA/issues/164
-		row[x_pixel] = abcde & (0x18 << BITS_ADDR_OFFSET); // mask out the bottom 3 bits which are the clk di bk inputs  
-	  } else {		  
-		row[x_pixel] = abcde;
-	  }	  
+      
+      if ( m_cfg.driver == HUB75_I2S_CFG::SM5266P) {
+        // modifications here for row shift register type SM5266P 
+        // https://github.com/mrfaptastic/ESP32-HUB75-MatrixPanel-I2S-DMA/issues/164
+        row[x_pixel] = abcde & (0x18 << BITS_ADDR_OFFSET); // mask out the bottom 3 bits which are the clk di bk inputs  
+      } else {        
+        row[x_pixel] = abcde;
+      }   
       //row[x_pixel] = abcde;
     } while(x_pixel);
-	
-	
-	// modifications here for row shift register type SM5266P 
-	// https://github.com/mrfaptastic/ESP32-HUB75-MatrixPanel-I2S-DMA/issues/164	
-	if ( m_cfg.driver == HUB75_I2S_CFG::SM5266P) {	
-		uint16_t serialCount;
-		uint16_t latch;
-		x_pixel = dma_buff.rowBits[row_idx]->width - 16; // come back 8*2 pixels to allow for 8 writes
-		serialCount = 8;
-		do{
-			serialCount--;
-			latch = row[x_pixel] | (((((ESP32_I2S_DMA_STORAGE_TYPE)row_idx) % 8) == serialCount) << 1) << BITS_ADDR_OFFSET; // data on 'B'
-			row[x_pixel++] = latch| (0x05<< BITS_ADDR_OFFSET); // clock high on 'A'and BK high for update
-			row[x_pixel++] = latch| (0x04<< BITS_ADDR_OFFSET); // clock low on 'A'and BK high for update
-		} while (serialCount);
-	} // end SM5266P	
-	
+    
+    
+    // modifications here for row shift register type SM5266P 
+    // https://github.com/mrfaptastic/ESP32-HUB75-MatrixPanel-I2S-DMA/issues/164    
+    if ( m_cfg.driver == HUB75_I2S_CFG::SM5266P) {  
+        uint16_t serialCount;
+        uint16_t latch;
+        x_pixel = dma_buff.rowBits[row_idx]->width - 16; // come back 8*2 pixels to allow for 8 writes
+        serialCount = 8;
+        do{
+            serialCount--;
+            latch = row[x_pixel] | (((((ESP32_I2S_DMA_STORAGE_TYPE)row_idx) % 8) == serialCount) << 1) << BITS_ADDR_OFFSET; // data on 'B'
+            row[x_pixel++] = latch| (0x05<< BITS_ADDR_OFFSET); // clock high on 'A'and BK high for update
+            row[x_pixel++] = latch| (0x04<< BITS_ADDR_OFFSET); // clock low on 'A'and BK high for update
+        } while (serialCount);
+    } // end SM5266P    
+    
 
     // let's set LAT/OE control bits for specific pixels in each color_index subrows
-	// Need to consider the original ESP32's (WROOM) DMA TX FIFO reordering of bytes...
+    // Need to consider the original ESP32's (WROOM) DMA TX FIFO reordering of bytes...
     uint8_t coloridx = dma_buff.rowBits[row_idx]->color_depth;
     do {
       --coloridx;
 
-	  // switch pointer to a row for a specific color index
-	  row = dma_buff.rowBits[row_idx]->getDataPtr(coloridx, _buff_id);
+      // switch pointer to a row for a specific color index
+      row = dma_buff.rowBits[row_idx]->getDataPtr(coloridx, _buff_id);
 
-	  #ifdef ESP32_SXXX
-		// -1 works better on ESP32-S2 ? Because bytes get sent out in order...
-		row[dma_buff.rowBits[row_idx]->width - 1] |= BIT_LAT;   // -1 pixel to compensate array index starting at 0					
-	  #else
-		// We need to update the correct uint16_t in the rowBitStruct array, that gets sent out in parallel
-		// 16 bit parallel mode - Save the calculated value to the bitplane memory in reverse order to account for I2S Tx FIFO mode1 ordering
-		// Irrelevant for ESP32-S2 the way the FIFO ordering works is different - refer to page 679 of S2 technical reference manual
-		row[dma_buff.rowBits[row_idx]->width - 2] |= BIT_LAT;   // -2 in the DMA array is actually -1 when it's reordered by TX FIFO				  
-	  #endif
+      #ifdef ESP32_SXXX
+        // -1 works better on ESP32-S2 ? Because bytes get sent out in order...
+        row[dma_buff.rowBits[row_idx]->width - 1] |= BIT_LAT;   // -1 pixel to compensate array index starting at 0                 
+      #else
+        // We need to update the correct uint16_t in the rowBitStruct array, that gets sent out in parallel
+        // 16 bit parallel mode - Save the calculated value to the bitplane memory in reverse order to account for I2S Tx FIFO mode1 ordering
+        // Irrelevant for ESP32-S2 the way the FIFO ordering works is different - refer to page 679 of S2 technical reference manual
+        row[dma_buff.rowBits[row_idx]->width - 2] |= BIT_LAT;   // -2 in the DMA array is actually -1 when it's reordered by TX FIFO                  
+      #endif
 
       // need to disable OE before/after latch to hide row transition
       // Should be one clock or more before latch, otherwise can get ghosting
@@ -704,21 +704,21 @@ void MatrixPanel_I2S_DMA::clearFrameBuffer(bool _buff_id){
       do {
         --_blank;
         
-	  #ifdef ESP32_SXXX		
-			row[0 + _blank] |= BIT_OE;
-			row[dma_buff.rowBits[row_idx]->width - _blank - 1 ] |= BIT_OE;    // (LAT pulse is (width-2) -1 pixel to compensate array index starting at 0
- 	  #else
-		  
-			// Original ESP32 WROOM FIFO Ordering Sucks
-			uint8_t _blank_row_tx_fifo_tmp = 0 + _blank;
-			(_blank_row_tx_fifo_tmp & 1U) ? --_blank_row_tx_fifo_tmp : ++_blank_row_tx_fifo_tmp; 
-			row[_blank_row_tx_fifo_tmp] |= BIT_OE;
-			
-			_blank_row_tx_fifo_tmp = dma_buff.rowBits[row_idx]->width - _blank - 1; // (LAT pulse is (width-2) -1 pixel to compensate array index starting at 0
-			(_blank_row_tx_fifo_tmp & 1U) ? --_blank_row_tx_fifo_tmp : ++_blank_row_tx_fifo_tmp; 
-			row[_blank_row_tx_fifo_tmp] |= BIT_OE;
-			
-	  #endif
+      #ifdef ESP32_SXXX     
+            row[0 + _blank] |= BIT_OE;
+            row[dma_buff.rowBits[row_idx]->width - _blank - 1 ] |= BIT_OE;    // (LAT pulse is (width-2) -1 pixel to compensate array index starting at 0
+      #else
+          
+            // Original ESP32 WROOM FIFO Ordering Sucks
+            uint8_t _blank_row_tx_fifo_tmp = 0 + _blank;
+            (_blank_row_tx_fifo_tmp & 1U) ? --_blank_row_tx_fifo_tmp : ++_blank_row_tx_fifo_tmp; 
+            row[_blank_row_tx_fifo_tmp] |= BIT_OE;
+            
+            _blank_row_tx_fifo_tmp = dma_buff.rowBits[row_idx]->width - _blank - 1; // (LAT pulse is (width-2) -1 pixel to compensate array index starting at 0
+            (_blank_row_tx_fifo_tmp & 1U) ? --_blank_row_tx_fifo_tmp : ++_blank_row_tx_fifo_tmp; 
+            row[_blank_row_tx_fifo_tmp] |= BIT_OE;
+            
+      #endif
 
       } while (_blank);
 
@@ -758,23 +758,26 @@ void MatrixPanel_I2S_DMA::brtCtrlOE(int brt, const bool _buff_id){
       int x_coord = dma_buff.rowBits[row_idx]->width;
       do {
         --x_coord;
+        
+       // clear OE bit for all other pixels
+        row[x_coord] &= BITMASK_OE_CLEAR;       
 
         // Brightness control via OE toggle - disable matrix output at specified x_coord
         if((coloridx > lsbMsbTransitionBit || !coloridx) && ((x_coord) >= brt)){
           row[x_coord] |= BIT_OE; // Disable output after this point.
-		  continue;  
+          continue;  
         }
         // special case for the bits *after* LSB through (lsbMsbTransitionBit) - OE is output after data is shifted, so need to set OE to fractional brightness
         if(coloridx && coloridx <= lsbMsbTransitionBit) {
-          // divide brightness in half for each bit below lsbMsbTransitionBit
-          int lsbBrightness = brt >> (lsbMsbTransitionBit - coloridx + 1);
-          if((x_coord) >= lsbBrightness)
-            row[x_coord] |= BIT_OE;  // Disable output after this point.
-            continue;
+            // divide brightness in half for each bit below lsbMsbTransitionBit
+            int lsbBrightness = brt >> (lsbMsbTransitionBit - coloridx + 1);
+            if((x_coord) >= lsbBrightness) {
+                row[x_coord] |= BIT_OE;  // Disable output after this point.
+                continue;
+            }
         }
 
-        // clear OE bit for all other pixels
-        row[x_coord] &= BITMASK_OE_CLEAR;
+ 
       } while(x_coord);
 
       // need to disable OE before/after latch to hide row transition
@@ -783,14 +786,14 @@ void MatrixPanel_I2S_DMA::brtCtrlOE(int brt, const bool _buff_id){
       do {
         --_blank;
 
-	  #ifdef ESP32_SXXX		
-			row[0 + _blank] |= BIT_OE;
- 	  #else 
-			// Original ESP32 WROOM FIFO Ordering Sucks
-			uint8_t _blank_row_tx_fifo_tmp = 0 + _blank;
-			(_blank_row_tx_fifo_tmp & 1U) ? --_blank_row_tx_fifo_tmp : ++_blank_row_tx_fifo_tmp; 
-			row[_blank_row_tx_fifo_tmp] |= BIT_OE;					
-	  #endif
+      #ifdef ESP32_SXXX     
+            row[0 + _blank] |= BIT_OE;
+      #else 
+            // Original ESP32 WROOM FIFO Ordering Sucks
+            uint8_t _blank_row_tx_fifo_tmp = 0 + _blank;
+            (_blank_row_tx_fifo_tmp & 1U) ? --_blank_row_tx_fifo_tmp : ++_blank_row_tx_fifo_tmp; 
+            row[_blank_row_tx_fifo_tmp] |= BIT_OE;                  
+      #endif
 
         //row[0 + _blank] |= BIT_OE;
         // no need, has been done already
@@ -866,9 +869,9 @@ void MatrixPanel_I2S_DMA::hlineDMA(int16_t x_coord, int16_t y_coord, int16_t l,
 
   /* LED Brightness Compensation */
 #ifndef NO_CIE1931
-	red   = lumConvTab[red];
-	green = lumConvTab[green];
-	blue  = lumConvTab[blue]; 	
+    red   = lumConvTab[red];
+    green = lumConvTab[green];
+    blue  = lumConvTab[blue];   
 #endif
 
   uint16_t _colorbitclear = BITMASK_RGB1_CLEAR, _colorbitoffset = 0;
@@ -887,14 +890,14 @@ void MatrixPanel_I2S_DMA::hlineDMA(int16_t x_coord, int16_t y_coord, int16_t l,
     // let's precalculate RGB1 and RGB2 bits than flood it over the entire DMA buffer
     uint16_t RGB_output_bits = 0;
 //    uint8_t mask = (1 << color_depth_idx COLOR_DEPTH_LESS_THAN_8BIT_ADJUST);
-	#if PIXEL_COLOR_DEPTH_BITS < 8
-		uint8_t mask = (1 << (color_depth_idx+MASK_OFFSET)); // expect 24 bit color (8 bits per RGB subpixel)
-	#else
-		uint8_t mask = (1 << (color_depth_idx)); // expect 24 bit color (8 bits per RGB subpixel)
-	#endif		
+    #if PIXEL_COLOR_DEPTH_BITS < 8
+        uint8_t mask = (1 << (color_depth_idx+MASK_OFFSET)); // expect 24 bit color (8 bits per RGB subpixel)
+    #else
+        uint8_t mask = (1 << (color_depth_idx)); // expect 24 bit color (8 bits per RGB subpixel)
+    #endif      
 
     /* Per the .h file, the order of the output RGB bits is:
-      * BIT_B2, BIT_G2, BIT_R2,    BIT_B1, BIT_G1, BIT_R1	  */
+      * BIT_B2, BIT_G2, BIT_R2,    BIT_B1, BIT_G1, BIT_R1     */
     RGB_output_bits |= (bool)(blue & mask);   // --B
     RGB_output_bits <<= 1;
     RGB_output_bits |= (bool)(green & mask);  // -BG
@@ -911,14 +914,14 @@ void MatrixPanel_I2S_DMA::hlineDMA(int16_t x_coord, int16_t y_coord, int16_t l,
     int16_t _l = l;
     do {                 // iterate pixels in a row
         int16_t _x = x_coord + --_l;
-		
-#ifdef ESP32_SXXX		
+        
+#ifdef ESP32_SXXX       
         // ESP 32 doesn't need byte flipping for TX FIFO.
         uint16_t &v = p[_x];
 #else
         // Save the calculated value to the bitplane memory in reverse order to account for I2S Tx FIFO mode1 ordering
         uint16_t &v = p[_x & 1U ? --_x : ++_x];
-#endif		
+#endif      
 
         v &= _colorbitclear;      // reset color bits
         v |= RGB_output_bits;    // set new color bits
@@ -948,9 +951,9 @@ void MatrixPanel_I2S_DMA::vlineDMA(int16_t x_coord, int16_t y_coord, int16_t l,
 
   /* LED Brightness Compensation */
 #ifndef NO_CIE1931
-	red   = lumConvTab[red];
-	green = lumConvTab[green];
-	blue  = lumConvTab[blue]; 	
+    red   = lumConvTab[red];
+    green = lumConvTab[green];
+    blue  = lumConvTab[blue];   
 #endif
 
 #ifndef ESP32_SXXX
@@ -964,15 +967,15 @@ void MatrixPanel_I2S_DMA::vlineDMA(int16_t x_coord, int16_t y_coord, int16_t l,
 
     // let's precalculate RGB1 and RGB2 bits than flood it over the entire DMA buffer
 //    uint8_t mask = (1 << color_depth_idx COLOR_DEPTH_LESS_THAN_8BIT_ADJUST);
-	#if PIXEL_COLOR_DEPTH_BITS < 8
-		uint8_t mask = (1 << (color_depth_idx+MASK_OFFSET)); // expect 24 bit color (8 bits per RGB subpixel)
-	#else
-		uint8_t mask = (1 << (color_depth_idx)); // expect 24 bit color (8 bits per RGB subpixel)
-	#endif		
+    #if PIXEL_COLOR_DEPTH_BITS < 8
+        uint8_t mask = (1 << (color_depth_idx+MASK_OFFSET)); // expect 24 bit color (8 bits per RGB subpixel)
+    #else
+        uint8_t mask = (1 << (color_depth_idx)); // expect 24 bit color (8 bits per RGB subpixel)
+    #endif      
     uint16_t RGB_output_bits = 0;
 
     /* Per the .h file, the order of the output RGB bits is:
-    * BIT_B2, BIT_G2, BIT_R2,    BIT_B1, BIT_G1, BIT_R1	  */
+    * BIT_B2, BIT_G2, BIT_R2,    BIT_B1, BIT_G1, BIT_R1   */
     RGB_output_bits |= (bool)(blue & mask);   // --B
     RGB_output_bits <<= 1;
     RGB_output_bits |= (bool)(green & mask);  // -BG
@@ -1029,4 +1032,4 @@ void MatrixPanel_I2S_DMA::fillRectDMA(int16_t x, int16_t y, int16_t w, int16_t h
   }
 }
 
-#endif  // NO_FAST_FUNCTIONS
\ No newline at end of file
+#endif  // NO_FAST_FUNCTIONS
diff --git a/ESP32-HUB75-MatrixPanel-I2S-DMA.h b/ESP32-HUB75-MatrixPanel-I2S-DMA.h
index 0fb91d1..0a39538 100644
--- a/ESP32-HUB75-MatrixPanel-I2S-DMA.h
+++ b/ESP32-HUB75-MatrixPanel-I2S-DMA.h
@@ -132,7 +132,6 @@
 #define ESP32_I2S_DMA_MODE          I2S_PARALLEL_WIDTH_16   // From esp32_i2s_parallel_v2.h = 16 bits in parallel
 #define ESP32_I2S_DMA_STORAGE_TYPE  uint16_t                // DMA output of one uint16_t at a time.
 #define CLKS_DURING_LATCH           0                       // Not (yet) used. 
-#define ESP32_I2S_DEVICE            I2S_NUM_0
 
 // Panel Upper half RGB (numbering according to order in DMA gpio_bus configuration)
 #define BITS_RGB1_OFFSET 0 // Start point of RGB_X1 bits
diff --git a/ESP32-VirtualMatrixPanel-I2S-DMA.h b/ESP32-VirtualMatrixPanel-I2S-DMA.h
index f4f1834..7228232 100644
--- a/ESP32-VirtualMatrixPanel-I2S-DMA.h
+++ b/ESP32-VirtualMatrixPanel-I2S-DMA.h
@@ -133,7 +133,7 @@ inline VirtualCoords VirtualMatrixPanel::getCoords(int16_t &x, int16_t &y) {
 	//Serial.println("Called Base.");
 	coords.x = coords.y = -1; // By defalt use an invalid co-ordinates that will be rejected by updateMatrixDMABuffer
 
-	if ( x < 0 || x >= width() || y < 0 || y >= height() ) { // Co-ordinates go from 0 to X-1 remember! width() and height() are out of range!
+	if ( x < 0 || x >= virtualResX || y < 0 || y >= virtualResY ) { // Co-ordinates go from 0 to X-1 remember! otherwise they are out of range!
 	//Serial.printf("VirtualMatrixPanel::getCoords(): Invalid virtual display coordinate. x,y: %d, %d\r\n", x, y);
 		return coords;
 	}
@@ -246,9 +246,11 @@ inline void VirtualMatrixPanel::drawPixelRGB888(int16_t x, int16_t y, uint8_t r,
 
 inline void VirtualMatrixPanel::setRotate(bool rotate) {
 	_rotate=rotate;	
-	
+
+#ifndef NO_GFX
 	// We don't support rotation by degrees.
 	if (rotate) { setRotation(1); } else { setRotation(0); }
+#endif
 }
 
 inline void VirtualMatrixPanel::setPhysicalPanelScanRate(PANEL_SCAN_RATE rate) {
diff --git a/component.mk b/component.mk
new file mode 100644
index 0000000..049f190
--- /dev/null
+++ b/component.mk
@@ -0,0 +1 @@
+COMPONENT_ADD_INCLUDEDIRS = .

diff --git a/esp32_i2s_parallel_dma.c b/esp32_i2s_parallel_dma.c
index ec6e17b..6c62514 100644
--- a/esp32_i2s_parallel_dma.c
+++ b/esp32_i2s_parallel_dma.c
@@ -22,7 +22,6 @@
 
 #include <driver/gpio.h>
 #include <driver/periph_ctrl.h>
-#include <rom/gpio.h>
 #include <soc/gpio_sig_map.h>
 
 // For I2S state management.
@@ -47,41 +46,28 @@ volatile bool previousBufferFree      = true;
 
 static void IRAM_ATTR irq_hndlr(void* arg) { // if we use I2S1 (default)
 
-#ifdef ESP32_ORIG
+//i2s_port_t port = *((i2s_port_t*) arg);
 
-    if ( (*(i2s_port_t*)arg) == I2S_NUM_1 ) { // https://www.bogotobogo.com/cplusplus/pointers2_voidpointers_arrays.php
-      //For I2S1
+/* Compiler pre-processor check. Saves a few cycles, no need to cast void ptr to i2s_port_t and then check 120 times second... */
+#if ESP32_I2S_DEVICE == I2S_NUM_0
       SET_PERI_REG_BITS(I2S_INT_CLR_REG(1), I2S_OUT_EOF_INT_CLR_V, 1, I2S_OUT_EOF_INT_CLR_S);
-    } else {
+#else
       // For I2S0
       SET_PERI_REG_BITS(I2S_INT_CLR_REG(0), I2S_OUT_EOF_INT_CLR_V, 1, I2S_OUT_EOF_INT_CLR_S);    
-    }
+#endif
 
-#else
-      // Other ESP32 MCU's only have one I2S 
-      SET_PERI_REG_BITS(I2S_INT_CLR_REG(0), I2S_OUT_EOF_INT_CLR_V, 1, I2S_OUT_EOF_INT_CLR_S);    
+	previousBufferFree 		= true;
 
-#endif
-/*
-    if ( previousBufferOutputLoopCount == 1)
-	{
-		// at this point, the previously active buffer is free, go ahead and write to it
-		previousBufferFree 		= true;
-		////previousBufferOutputLoopCount = 0;
-		//i2s_parallel_set_previous_buffer_not_free();
-	}
-	else { previousBufferOutputLoopCount++; } 
+/*	
+    if(shiftCompleteCallback) { // we've defined a callback function ?
+        shiftCompleteCallback();
+ }
 */
-	previousBufferFree 		= true;
-	
- //   if(shiftCompleteCallback) // we've defined a callback function ?
- //       shiftCompleteCallback();
         
 } // end irq_hndlr
 
 
 // For peripheral setup and configuration
-
 static inline int get_bus_width(i2s_parallel_cfg_bits_t width) {
   switch(width) {
     case I2S_PARALLEL_WIDTH_8:
diff --git a/esp32_i2s_parallel_dma.h b/esp32_i2s_parallel_dma.h
index 1570fe2..d7678bc 100644
--- a/esp32_i2s_parallel_dma.h
+++ b/esp32_i2s_parallel_dma.h
@@ -19,7 +19,11 @@ extern "C" {
 #include <freertos/FreeRTOS.h>
 #include <driver/i2s.h>
 #include <esp_err.h>
+//#include <esp32/rom/lldesc.h>
+//#include <esp32/rom/gpio.h>
 #include <rom/lldesc.h>
+#include <rom/gpio.h>
+
 
 // Get MCU Type and Max CLK Hz for MCU
 #include <esp32_i2s_parallel_mcu_def.h>
@@ -41,20 +45,22 @@ typedef struct {
     int desccount_b;      // only used with double buffering
     lldesc_t * lldesc_b;  // only used with double buffering
     bool clkphase;        // Clock signal phase
-	bool int_ena_out_eof; // Do we raise an interrupt every time the DMA output loops? Don't do this unless we're doing double buffering!
+    bool int_ena_out_eof; // Do we raise an interrupt every time the DMA output loops? Don't do this unless we're doing double buffering!
 } i2s_parallel_config_t;
 
 static inline int i2s_parallel_get_memory_width(i2s_port_t port, i2s_parallel_cfg_bits_t width) {
   switch(width) {
     case I2S_PARALLEL_WIDTH_8:
 
-#ifdef ESP32_ORIG	
-      // IS21 supports space saving single byte 8 bit parallel access
+#ifdef ESP32_ORIG   
+      // Only I2S1 on the legacy ESP32 WROOM MCU supports space saving single byte 8 bit parallel access
       if(port == I2S_NUM_1) {
         return 1;
+      } else {
+        return 2;
       }
 #else 
-		return 1;
+        return 1;
 #endif
 
     case I2S_PARALLEL_WIDTH_16:
@@ -70,10 +76,10 @@ static inline int i2s_parallel_get_memory_width(i2s_port_t port, i2s_parallel_cf
 void link_dma_desc(volatile lldesc_t *dmadesc, volatile lldesc_t *prevdmadesc, void *memory, size_t size);
 
 // I2S DMA Peripheral Setup Functions
-esp_err_t 	i2s_parallel_driver_install(i2s_port_t port, i2s_parallel_config_t* conf);
-esp_err_t 	i2s_parallel_send_dma(i2s_port_t port, lldesc_t* dma_descriptor);
-esp_err_t 	i2s_parallel_stop_dma(i2s_port_t port);
-//i2s_dev_t* 	i2s_parallel_get_dev(i2s_port_t port);
+esp_err_t   i2s_parallel_driver_install(i2s_port_t port, i2s_parallel_config_t* conf);
+esp_err_t   i2s_parallel_send_dma(i2s_port_t port, lldesc_t* dma_descriptor);
+esp_err_t   i2s_parallel_stop_dma(i2s_port_t port);
+//i2s_dev_t*    i2s_parallel_get_dev(i2s_port_t port);
 
 // For frame buffer flipping / double buffering
 typedef struct {
diff --git a/esp32_i2s_parallel_dma_cxxx.txt b/esp32_i2s_parallel_dma_cxxx.txt
deleted file mode 100644
index 60ebf5f..0000000
--- a/esp32_i2s_parallel_dma_cxxx.txt
+++ /dev/null
@@ -1,313 +0,0 @@
-/*
- * ESP32_I2S_PARALLEL_DMA (Version 3)
- * 
- * Author:      Mrfaptastic @ https://github.com/mrfaptastic/
- * 
- * Description: Multi-ESP32 product DMA setup functions for ESP32 C3/H2 RISC-V chips
- * 
- * ESP32C series doesn't support LCD mode / parallel DMA! 
- *
- */
-
-// Header
-#include "esp32_i2s_parallel_dma.h" 
-
-#if defined(ESP32_CXXX)
-
-
-#include <esp_err.h>
- 
- // Turn on and off a peripheral
-#include <driver/periph_ctrl.h>
-
-// GPIO
-#include <soc/gpio_periph.h>
-#include <hal/gpio_types.h>
-#include <driver/gpio.h>
-#include <driver/periph_ctrl.h>
-
-#include <rom/gpio.h>
-#include <soc/gpio_sig_map.h>
-
-// DMA Linked List Struct
-#include <soc/lldesc.h>
-#include <soc/io_mux_reg.h>
-
-// I2S
-#include <soc/i2s_struct.h>
-#include <soc/i2s_reg.h>
-
-// GDMA
-#include <soc/gdma_channel.h>
-#include <soc/gdma_periph.h>
-#include <soc/gdma_reg.h>
-#include <soc/gdma_struct.h>
-
-// For I2S state management.
-static i2s_parallel_state_t *i2s_state 	= NULL;
-
-// ESP32-S2,S3,C3 only has IS20
-// Original ESP32 has two I2S's, but we'll stick with the lowest common denominator.
-static i2s_dev_t* I2S 					= &I2S0; // Device to use for this library, change if you want.
-
-callback shiftCompleteCallback;
-void setShiftCompleteCallback(callback f) {
-    shiftCompleteCallback = f;
-}
-
-volatile bool previousBufferFree = true;
-
-static void IRAM_ATTR gdma_irq_handler(void* arg) { // if we use I2S1 (default)
-
-    GDMA.intr[0].clr.out_eof = 1;
-
-    // at this point, the previously active buffer is free, go ahead and write to it
-    // previousBufferFree = true;
-    // if(shiftCompleteCallback) // we've defined a callback function ?
-    // shiftCompleteCallback();
-
-    // at this point, the previously active buffer is free, go ahead and write to it
-    previousBufferFree = true;
-
-    if(shiftCompleteCallback) // we've defined a callback function ?
-        shiftCompleteCallback();
-		
-} // end irq_hndlr
-
-
-// For peripheral setup and configuration
-
-static inline int get_bus_width(i2s_parallel_cfg_bits_t width) {
-  switch(width) {
-    case I2S_PARALLEL_WIDTH_8:
-      return 8;
-    case I2S_PARALLEL_WIDTH_16:
-      return 16;
-    case I2S_PARALLEL_WIDTH_24:
-      return 24;
-    default:
-      return -ESP_ERR_INVALID_ARG;
-  }
-}
-
-static void gpio_setup_out(int gpio, int sig) {
-  if(gpio < 0) {
-    return;
-  }
-
-    // Configure GPIO
-    // https://github.com/espressif/esp-idf/blob/d5f58ab13551cd883e8d8478ba367b6e4543ffec/examples/peripherals/gpio/generic_gpio/main/gpio_example_main.c    
-    gpio_config_t io_conf = {};
-    io_conf.intr_type       = GPIO_INTR_DISABLE;
-    io_conf.mode            = GPIO_MODE_DEF_OUTPUT;
-    io_conf.pin_bit_mask    = (1ULL << gpio);
-    io_conf.pull_down_en    = 1;
-    io_conf.pull_up_en      = 0;
-
-    gpio_config(&io_conf);
-
-    // Set IOMUX to GPIO
-    //gpio_iomux_out(gpio, sig, false); // ?? is sig right?
-    gpio_matrix_out(gpio, sig, false, false);
-
-    // Drive Strength to MAX
-    gpio_set_drive_capability((gpio_num_t)gpio, (gpio_drive_cap_t)3);  
-}
-
-
-// DMA Linked List
-// Size must be less than DMA_MAX - need to handle breaking long transfer into two descriptors before call
-// DMA_MAX by the way is the maximum data packet size you can hold in one chunk
-void link_dma_desc(volatile lldesc_t *dmadesc, volatile lldesc_t *prevdmadesc, void *memory, size_t size) 
-{
-    if(size > DMA_MAX) size = DMA_MAX;
-
-    dmadesc->size = size;
-    dmadesc->length = size;
-    dmadesc->buf = memory;
-    dmadesc->eof = 0;
-    dmadesc->sosf = 0;
-    dmadesc->owner = 1;
-    dmadesc->qe.stqe_next = 0;  // will need to set this elsewhere
-    dmadesc->offset = 0;
-
-    // link previous to current
-    if(prevdmadesc)
-        prevdmadesc->qe.stqe_next = (lldesc_t*)dmadesc;
-}
-
-
-
-esp_err_t i2s_parallel_driver_install(i2s_port_t port, i2s_parallel_config_t* cfg) {
-  
-  port = I2S_NUM_0; /// override.
-  
-  if(port < I2S_NUM_0 || port >= I2S_NUM_MAX) {
-    return ESP_ERR_INVALID_ARG;
-  }
-  if(cfg->sample_width < I2S_PARALLEL_WIDTH_8 || cfg->sample_width >= I2S_PARALLEL_WIDTH_MAX) {
-    return ESP_ERR_INVALID_ARG;
-  }
-  if(cfg->sample_rate > I2S_PARALLEL_CLOCK_HZ || cfg->sample_rate < 1) {
-    return ESP_ERR_INVALID_ARG;
-  }
-  uint32_t clk_div_main = I2S_PARALLEL_CLOCK_HZ / cfg->sample_rate / i2s_parallel_get_memory_width(port, cfg->sample_width);
-  if(clk_div_main < 2 || clk_div_main > 0xFF) {
-    return ESP_ERR_INVALID_ARG;
-  }
-  
-    // Setup GPIOs
-    int bus_width = get_bus_width(cfg->sample_width);  
-
-    i2s_dev_t *i2s_dev = I2S; // There's only one I2S device on C3....
-
-    //Figure out which signal numbers to use for routing
-    //printf("Setting up parallel I2S bus at I2S%d\n", i2snum(dev));
-    int sig_data_base =  I2SO_SD_OUT_IDX;
-    int sig_clk       =  I2SO_WS_OUT_IDX;
-
-    //Route the signals
-    for (int x=0; x < bus_width; x++) {
-        gpio_setup_out(cfg->gpio_bus[x], sig_data_base+x);
-    }
-
-    //ToDo: Clk/WS may need inversion?
-    gpio_setup_out(cfg->gpio_clk, sig_clk);
-    
-    // Power on I2S1 (or 0)
-    periph_module_enable(PERIPH_I2S1_MODULE);
-    
-    // Now it's apparently I2S0 !?
-    //i2s_dev_t *i2s_dev = &I2S0;
-
-    // Reset RX (not that we use it)
-    i2s_dev->rx_conf.val = 0;
-    i2s_dev->rx_conf.rx_reset=1;         i2s_dev->rx_conf.rx_reset=0;
-    i2s_dev->rx_conf.rx_fifo_reset=1;    i2s_dev->rx_conf.rx_fifo_reset=0;
-
-    // Reset TX (what we care about)
-    i2s_dev->tx_conf.val = 0;
-    i2s_dev->tx_conf.tx_reset=1;         i2s_dev->tx_conf.tx_reset=0;
-    i2s_dev->tx_conf.tx_fifo_reset=1;    i2s_dev->tx_conf.tx_fifo_reset=0;
-    i2s_dev->tx_conf.tx_chan_equal=1;
-
-    // Device setup
-    i2s_dev->tx_conf1.val = 0;    
-    i2s_dev->tx_conf1.tx_bits_mod=16;//cfg->bits;
-
-    i2s_dev->rx_conf1.val = 0;    
-    i2s_dev->rx_conf1.rx_bits_mod=16; //cfg->bits;
-
-    i2s_dev->tx_conf1.tx_bck_div_num=2;
-    i2s_dev->rx_conf1.rx_bck_div_num=2;
-
-    i2s_dev->tx_clkm_conf.val=0;
-    i2s_dev->tx_clkm_conf.tx_clk_sel=2; // 160mhz
-
-    // clock speed
-    i2s_dev->tx_clkm_conf.tx_clkm_div_num=160/16; // 10Mhz
-    //i2s_dev->tx_clkm_div_conf.val = 0; // 
-    i2s_dev->tx_clkm_div_conf.tx_clkm_div_x = 0; // > ?
-    i2s_dev->tx_clkm_div_conf.tx_clkm_div_y = 0; // 
-    i2s_dev->tx_clkm_div_conf.tx_clkm_div_yn1 = 0;
-    i2s_dev->tx_clkm_div_conf.tx_clkm_div_z = 0;
-
-    i2s_dev->tx_clkm_conf.tx_clk_active=1;  // Start
-
-    
-    //Allocate DMA descriptors
-    i2s_state = malloc(sizeof(i2s_parallel_state_t));
-    assert(i2s_state != NULL);
-    i2s_parallel_state_t *st= i2s_state;
-
-    st->desccount_a = cfg->desccount_a;
-    st->desccount_b = cfg->desccount_b;
-    st->dmadesc_a = cfg->lldesc_a;
-    st->dmadesc_b = cfg->lldesc_b;
-
-    // setup I2S Interrupt
-    // SET_PERI_REG_BITS(I2S_INT_ENA_REG(1), I2S_OUT_EOF_INT_ENA_V, 1, I2S_OUT_EOF_INT_ENA_S);
-    // allocate a level 1 interrupt: lowest priority, as ISR isn't urgent and may take a long time to complete
-
-    esp_intr_alloc(ETS_I2S1_INTR_SOURCE, (int)(ESP_INTR_FLAG_IRAM | ESP_INTR_FLAG_LEVEL1), gdma_irq_handler, NULL, NULL);
-    GDMA.intr[0].ena.out_eof  = 1; //?
-
- 
-    // Reset GDMA device
-    GDMA.channel[0].out.out_conf0.out_rst   = 1; // REG_SET_BIT(DMA_OUT_CONF0_CH0_REG, DMA_OUT_RST_CH0);
-    GDMA.channel[0].out.out_conf0.out_rst   = 0; // REG_CLR_BIT(DMA_OUT_CONF0_CH0_REG, DMA_OUT_RST_CH0);
-    // GDMA.channel[0].out.out_conf0.out_eof_mode = 1; ?
-    GDMA.misc_conf.ahbm_rst_inter = 1;
-    GDMA.misc_conf.ahbm_rst_inter = 0;
-
-    // Setup interrupt
-
-    // Setup outlink
-    GDMA.channel[0].out.out_link.addr       = ((uint32_t)(&st->dmadesc_a[0]));// Set a value here
-    GDMA.channel[0].out.out_peri_sel.sel    = SOC_GDMA_TRIG_PERIPH_I2S0; // 3 = I2S0
-    GDMA.channel[0].out.out_conf0.out_data_burst_en = 1;
-    GDMA.channel[0].out.out_conf0.outdscr_burst_en  = 1;
-    GDMA.channel[0].out.out_link.start = 1;
-
-    while (!GDMA.intr->raw.out_eof) { } // check status    
-
-    i2s_dev->tx_conf.tx_start = 1;
-	
-    return ESP_OK;
-}
-
- esp_err_t i2s_parallel_stop_dma(i2s_port_t port) {
-  if(port < I2S_NUM_0 || port >= I2S_NUM_MAX) {
-    return ESP_ERR_INVALID_ARG;
-  }
-	
-   // Not implemented
-   return ESP_OK;
-}
-
-
- esp_err_t i2s_parallel_send_dma(i2s_port_t port, lldesc_t* dma_descriptor) {
-  if(port < I2S_NUM_0 || port >= I2S_NUM_MAX) {
-    return ESP_ERR_INVALID_ARG;
-  }
-  
-   // Not implemented
-  return ESP_OK;
-}
-
-i2s_dev_t* i2s_parallel_get_dev(i2s_port_t port) {
-  if(port < I2S_NUM_0 || port >= I2S_NUM_MAX) {
-    return NULL;
-  }
-  return I2S; // HARCODE THIS TO RETURN &I2S0
-}
-
-// Double buffering flipping
-// Flip to a buffer: 0 for bufa, 1 for bufb
-void i2s_parallel_flip_to_buffer(i2s_port_t port, int buffer_id) {
-
-    if (i2s_state == NULL) {
-      return; // :-()
-    }
-
-    lldesc_t *active_dma_chain;
-    if (buffer_id == 0) {
-        active_dma_chain=(lldesc_t*)&i2s_state->dmadesc_a[0];
-    } else {
-        active_dma_chain=(lldesc_t*)&i2s_state->dmadesc_b[0];
-    }
-
-    // setup linked list to refresh from new buffer (continuously) when the end of the current list has been reached
-    i2s_state->dmadesc_a[i2s_state->desccount_a-1].qe.stqe_next = active_dma_chain;
-    i2s_state->dmadesc_b[i2s_state->desccount_b-1].qe.stqe_next = active_dma_chain;
-
-    // we're still shifting out the buffer, so it shouldn't be written to yet.
-    previousBufferFree = false;
-}
-
-bool i2s_parallel_is_previous_buffer_free() {
-    return previousBufferFree;
-}
-
-// End ESP32 original / S2, S3 check
-#endif
\ No newline at end of file
diff --git a/esp32_i2s_parallel_mcu_def.h b/esp32_i2s_parallel_mcu_def.h
index 4ab524e..dd276da 100644
--- a/esp32_i2s_parallel_mcu_def.h
+++ b/esp32_i2s_parallel_mcu_def.h
@@ -10,6 +10,7 @@
 #if CONFIG_IDF_TARGET_ESP32S2 || CONFIG_IDF_TARGET_ESP32S3
 
 	#define ESP32_SXXX 1
+	#define ESP32_I2S_DEVICE            I2S_NUM_0	
 		
 	#define I2S_PARALLEL_CLOCK_HZ 160000000L
 	#define DMA_MAX (4096-4)
@@ -18,6 +19,7 @@
 
 	// 2016 model that started it all, and this library. The best.
 	#define ESP32_ORIG 1	
+	#define ESP32_I2S_DEVICE            I2S_NUM_0	
 	
 	#define I2S_PARALLEL_CLOCK_HZ 80000000L
 	#define DMA_MAX (4096-4)
diff --git a/examples/ChainedPanelsScreenBuffer/ChainedPanelsScreenBuffer.ino b/examples/ChainedPanelsScreenBuffer/ChainedPanelsScreenBuffer.ino
new file mode 100644
index 0000000..b5de3c3
--- /dev/null
+++ b/examples/ChainedPanelsScreenBuffer/ChainedPanelsScreenBuffer.ino
@@ -0,0 +1,146 @@
+/*************************************************************************
+ *             IMPORANT PLEASE READ THE INFORMATION BELOW!
+ * 
+ * This example implements a 'pixel buffer' which is essentally an 
+ * off-screen copy of what is intended to be sent to output (LED panels)
+ *
+ * This essentially means DOUBLE THE AMOUNT OF MEMORY is required to 
+ * to store the off-screen image/pixel/display buffer WITH a similar 
+ * amount of memory used for the DMA output buffer for the physical panels.
+ *
+ * This means the practical resolution you will be able to output with the
+ * ESP32 will be CUT IN HALF. Do not try to run huge chains of 
+ * LED Matrix Panels using this buffer, you will run out of memory.
+ *
+ * Please DO NOT raise issues @ github about running out of memory,
+ * we can't do anything about it. It's an ESP32, not a Raspberry Pi!
+ *
+ *************************************************************************/ 
+
+/* Use the FastLED_Pixel_Buffer class to handle panel chaining
+ * (it's based on the VirtualMatrixPanel class) AND also create an 
+ * off-screen CRGB FastLED pixel buffer.
+ */
+#include "FastLED_Pixel_Buffer.h" 
+
+  // Panel configuration
+  #define PANEL_RES_X 64 // Number of pixels wide of each INDIVIDUAL panel module. 
+  #define PANEL_RES_Y 32 // Number of pixels tall of each INDIVIDUAL panel module.
+   
+  #define NUM_ROWS 1 // Number of rows of chained INDIVIDUAL PANELS
+  #define NUM_COLS 2 // Number of INDIVIDUAL PANELS per ROW
+      
+  // Change this to your needs, for details please read the PDF in 
+  // the 'ChainedPanels'example folder!
+  #define SERPENT true
+  #define TOPDOWN false
+  
+  // placeholder for the matrix object
+  MatrixPanel_I2S_DMA *dma_display = nullptr;
+
+  // placeholder for the virtual display object
+  VirtualMatrixPanel_FastLED_Pixel_Buffer  *FastLED_Pixel_Buff = nullptr;
+  
+  /******************************************************************************
+   * Setup!
+   ******************************************************************************/
+  void setup()
+  {
+    delay(250);
+   
+    Serial.begin(115200);
+    Serial.println(""); Serial.println(""); Serial.println("");
+    Serial.println("*****************************************************");
+    Serial.println("*        FastLED Pixel BufferDemonstration          *");
+    Serial.println("*****************************************************");
+  
+/* 
+     // 62x32 1/8 Scan Panels don't have a D and E pin!
+     
+     HUB75_I2S_CFG::i2s_pins _pins = {
+      R1_PIN, G1_PIN, B1_PIN, R2_PIN, G2_PIN, B2_PIN, 
+      A_PIN, B_PIN, C_PIN, D_PIN, E_PIN, 
+      LAT_PIN, OE_PIN, CLK_PIN
+     };
+*/
+    HUB75_I2S_CFG mxconfig(
+                PANEL_RES_X,              // DO NOT CHANGE THIS
+                PANEL_RES_Y,              // DO NOT CHANGE THIS
+                NUM_ROWS*NUM_COLS           // DO NOT CHANGE THIS
+                //,_pins            // Uncomment to enable custom pins
+    );
+    
+    mxconfig.clkphase = false; // Change this if you see pixels showing up shifted wrongly by one column the left or right.    
+    //mxconfig.driver   = HUB75_I2S_CFG::FM6126A;     // in case that we use panels based on FM6126A chip, we can set it here before creating MatrixPanel_I2S_DMA object
+	
+	// Do NOT use mxconfig.double_buffer when using this pixel buffer. 
+  
+    // OK, now we can create our matrix object
+    dma_display = new MatrixPanel_I2S_DMA(mxconfig);
+  
+    // let's adjust default physical panel brightness to about 75%
+    dma_display->setBrightness8(96);    // range is 0-255, 0 - 0%, 255 - 100%
+  
+    // Allocate memory and start DMA electrical output to physical panels
+    if( not dma_display->begin() )
+      Serial.println("****** !KABOOM! I2S memory allocation failed ***********");
+   
+    dma_display->clearScreen();
+    delay(500);
+    
+    // NOW, create the 'Virtual Matrix Panel' class with a FastLED Pixel Buffer! Pass it a dma_display hardware library pointer to use.
+    FastLED_Pixel_Buff = new VirtualMatrixPanel_FastLED_Pixel_Buffer((*dma_display), NUM_ROWS, NUM_COLS, PANEL_RES_X, PANEL_RES_Y, SERPENT, TOPDOWN);
+
+   if( not FastLED_Pixel_Buff->allocateMemory() )
+      Serial.println("****** !KABOOM! Unable to find enough memory for the FastLED pixel buffer! ***********");
+	  
+  }
+  
+   
+ // Borrowed from the SimpleTextShapes example.
+ uint16_t colorWheel(uint8_t pos) {
+   if(pos < 85) {
+     return dma_display->color565(pos * 3, 255 - pos * 3, 0);
+   } else if(pos < 170) {
+     pos -= 85;
+     return dma_display->color565(255 - pos * 3, 0, pos * 3);
+   } else {
+     pos -= 170;
+     return dma_display->color565(0, pos * 3, 255 - pos * 3);
+   }
+ }
+ 
+ /* A crap demonstration of using the pixel buffer.
+  * 1) Draw text at an incrementing (going down) y coordinate
+  * 2) Move down a pixel row
+  * 3) Draw the text again, fade the 'old' pixels. Using the pixel buffer to update all pixels on screen.
+  * 4) 'show' (send) the pixel buffer to the DMA output.
+  * 5) LOOP
+  */   
+  
+  uint8_t y_coord     = 0;
+  uint8_t wheel = 0;
+  
+  void loop() 
+  {
+      // draw text with a rotating colour
+      FastLED_Pixel_Buff->dimAll(200); // Dim all pixels by 250/255
+	  
+      FastLED_Pixel_Buff->setTextSize(1);     // size 1 == 8 pixels high
+      FastLED_Pixel_Buff->setTextWrap(false); // Don't wrap at end of line - will do ourselves
+      
+      FastLED_Pixel_Buff->setCursor(FastLED_Pixel_Buff->width()/4, y_coord);    // start at top left, with 8 pixel of spacing     
+      FastLED_Pixel_Buff->setTextColor(colorWheel(wheel++));
+      
+      FastLED_Pixel_Buff->print("MythicalForce");
+      
+      FastLED_Pixel_Buff->show();  // IMPORTANT -> SEND Pixel Buffer to DMA / Panel Output!
+      
+      y_coord++;
+      
+      if ( y_coord >= FastLED_Pixel_Buff->height())
+              y_coord = 0;
+
+      delay(35);
+      
+  } // end loop
diff --git a/examples/ChainedPanelsScreenBuffer/FastLED_Pixel_Buffer.cpp b/examples/ChainedPanelsScreenBuffer/FastLED_Pixel_Buffer.cpp
new file mode 100644
index 0000000..79e19c9
--- /dev/null
+++ b/examples/ChainedPanelsScreenBuffer/FastLED_Pixel_Buffer.cpp
@@ -0,0 +1,116 @@
+/**
+ * Experimental layer class to do play with pixel in an off-screen buffer before painting to the DMA 
+ *
+ * Requires FastLED
+ *
+ * Faptastic 2020-2021
+ **/
+
+#include "FastLED_Pixel_Buffer.h"
+
+/**
+ *  The one for 256+ matrices
+ *  otherwise this:
+ *    for (uint8_t i = 0; i < MATRIX_WIDTH; i++) {}
+ *  turns into an infinite loop
+ */
+inline uint16_t VirtualMatrixPanel_FastLED_Pixel_Buffer::XY16( uint16_t x, uint16_t y) {
+
+    if (x >= virtualResX) return 0;
+    if (y >= virtualResY) return 0;
+    
+    return (y * virtualResX) + x + 1; // everything offset by one to compute out of bounds stuff - never displayed by ShowFrame()
+}
+
+// For adafruit
+void VirtualMatrixPanel_FastLED_Pixel_Buffer::drawPixel(int16_t x, int16_t y, uint16_t color) {
+
+  //Serial.println("calling our drawpixel!");
+
+    // 565 color conversion
+    uint8_t r = ((((color >> 11)  & 0x1F) * 527) + 23) >> 6;
+    uint8_t g = ((((color >> 5)   & 0x3F) * 259) + 33) >> 6;
+    uint8_t b = (((color & 0x1F)  * 527) + 23) >> 6;
+
+    this->drawPixel(x, y, CRGB(r,g,b));
+}
+
+void VirtualMatrixPanel_FastLED_Pixel_Buffer::drawPixel(int16_t x, int16_t y, int r, int g, int b) {
+    this->drawPixel(x, y, CRGB(r,g,b));
+}
+
+// We actually just draw to ourselves... to our buffer
+void VirtualMatrixPanel_FastLED_Pixel_Buffer::drawPixel(int16_t x, int16_t y, CRGB color) 
+{
+  //Serial.printf("updated x y : %d %d", x, y);
+    buffer[XY16(x,y)] = color;
+}
+
+CRGB VirtualMatrixPanel_FastLED_Pixel_Buffer::getPixel(int16_t x, int16_t y) 
+{
+    return buffer[XY16(x,y)];
+}
+
+/**
+ * Dim all the pixels on the layer.
+ */
+void VirtualMatrixPanel_FastLED_Pixel_Buffer::dimAll(byte value) {   
+
+        //Serial.println("performing dimall");
+        // nscale8 max value is 255, or it'll flip back to 0 
+        // (documentation is wrong when it says x/256), it's actually x/255
+        /*
+        for (int y = 0; y < LAYER_HEIGHT; y++) {
+            for (int x = 0; x < LAYER_WIDTH; x++) {
+                pixels->data[y][x].nscale8(value);
+        }}
+        */
+        dimRect(0,0, virtualResX, virtualResY, value);
+}
+
+/**
+ * Dim all the pixels in a rectangular option of the layer the layer.
+ */
+void VirtualMatrixPanel_FastLED_Pixel_Buffer::dimRect(int16_t x, int16_t y, int16_t w, int16_t h, byte value) {
+        for (int16_t i = x; i < x + w; i++)
+        {
+            for (int16_t j = y; j < y + h; j++)
+            {
+                buffer[XY16(i,j)].nscale8(value);
+            }
+        }
+}
+
+void VirtualMatrixPanel_FastLED_Pixel_Buffer::clear() {
+    memset(buffer, CRGB(0,0,0), (virtualResX * virtualResY) );
+}
+
+/**
+ * Actually Send the CRGB FastLED buffer to the DMA engine / Physical Panels!
+ * Do this via the underlying 'VirtualMatrixPanel' that does all the pixel-remapping for 
+ * all sorts of chained panels, and panel scan types.
+ */
+void VirtualMatrixPanel_FastLED_Pixel_Buffer::show() { 
+
+    //Serial.println("Doing Show");
+    
+    CRGB _pixel = 0;
+    for (int16_t y = 0; y < virtualResY; y++) {
+        for (int16_t x = 0; x < virtualResX; x++)
+        {
+            //VirtualMatrixPanel::getCoords(x, y); // call to base to update coords for chaining approach           
+            _pixel = buffer[XY16(x,y)];
+            drawPixelRGB888( x, y, _pixel.r, _pixel.g, _pixel.b); // call VirtualMatrixPanel::drawPixelRGB888(...)
+            //drawPixelRGB888( x, y, 0, 0, 128); // call VirtualMatrixPanel::drawPixelRGB888(...)
+        } // end loop to copy fast led to the dma matrix
+    }
+
+} // show
+
+/**
+ * Cleanup should we delete this buffer class. Unlikely during runtime.
+ */
+VirtualMatrixPanel_FastLED_Pixel_Buffer::~VirtualMatrixPanel_FastLED_Pixel_Buffer(void)
+{
+  delete(buffer);
+}
\ No newline at end of file
diff --git a/examples/ChainedPanelsScreenBuffer/FastLED_Pixel_Buffer.h b/examples/ChainedPanelsScreenBuffer/FastLED_Pixel_Buffer.h
new file mode 100644
index 0000000..28db7d5
--- /dev/null
+++ b/examples/ChainedPanelsScreenBuffer/FastLED_Pixel_Buffer.h
@@ -0,0 +1,51 @@
+#ifndef VIRTUAL_MATRIX_PANEL_FASTLED_LAYER
+#define VIRTUAL_MATRIX_PANEL_FASTLED_LAYER
+
+#include <ESP32-VirtualMatrixPanel-I2S-DMA.h>
+#include <FastLED.h>
+
+class VirtualMatrixPanel_FastLED_Pixel_Buffer : public VirtualMatrixPanel 
+{
+    public:
+        using VirtualMatrixPanel::VirtualMatrixPanel; // perform VirtualMatrixPanel class constructor
+
+        bool allocateMemory() // allocate memory
+        {   
+                // https://www.geeksforgeeks.org/how-to-declare-a-2d-array-dynamically-in-c-using-new-operator/ 
+                buffer = new CRGB[virtualResX * virtualResY]; // These are defined in the underliny  
+                
+                if (!buffer) {  return false; }
+                
+                Serial.printf("Allocated %d bytes of memory for pixel buffer.\r\n", sizeof(CRGB)*((virtualResX * virtualResY)+1));
+                this->clear();
+                
+                return true;
+                
+        } // end Buffer
+
+        virtual void drawPixel(int16_t x, int16_t y, uint16_t color);           // overwrite adafruit implementation
+        void drawPixel(int16_t x, int16_t y, int r, int g, int b);      // Buffer implementation        
+        void drawPixel(int16_t x, int16_t y, CRGB color);               // Buffer implementation        
+        CRGB getPixel(int16_t x, int16_t y);                            // Returns a pixel value from the buffer.
+
+
+        void dimAll(byte value); 
+        void dimRect(int16_t x, int16_t y, int16_t w, int16_t h, byte value); 
+        
+        void clear();
+        
+        void show();  //    Send buffer to physical hardware / DMA engine.
+        
+        // Release Memory
+        ~VirtualMatrixPanel_FastLED_Pixel_Buffer(void); 
+        
+    protected:
+        uint16_t XY16( uint16_t x, uint16_t y);
+
+    private:
+        CRGB* buffer = nullptr;
+};
+
+
+
+#endif
\ No newline at end of file
diff --git a/examples/One_Eight_1_8_ScanPanel/README.md b/examples/One_Eight_1_8_ScanPanel/README.md
index a205c26..2fd55b9 100644
--- a/examples/One_Eight_1_8_ScanPanel/README.md
+++ b/examples/One_Eight_1_8_ScanPanel/README.md
@@ -4,10 +4,4 @@
 ESP32-HUB75-MatrixPanel-I2S-DMA library will not display output correctly with 1/8 scan panels such [as this](https://github.com/mrfaptastic/ESP32-HUB75-MatrixPanel-I2S-DMA/issues/154) by default.
 
 ## Solution
-It is possible to connect 1/8 scan panels to this library and 'trick' the output to work correctly on these panels by way of adjusting the pixel co-ordinates that are 'sent' to the ESP32-HUB75-MatrixPanel-I2S-DMA library (in this example, it is the 'dmaOutput' class).
-
-Creation of a '1_8_ScanPanel.h' class which sends an adjusted drawPixel() x,y co-ordinates to the underlying ESP32-HUB75-MatrixPanel-I2S-DMA library's drawPixel routine, to trick the output to look pixel perfect. 
-
-Refer to the '1_8_ScanPanel.h' logic which builds upon the library's core Virtual Display 'ESP32-VirtualMatrixPanel-I2S-DMA.h' to also support chaining of 1/8 Scan Panels as well. Refer to 'ChainedPanels' example on how to configure panel chaining to create bigger displays.
-
-
+It is possible to connect 1/8 scan panels to this library and 'trick' the output to work correctly on these panels by way of adjusting the pixel co-ordinates that are 'sent' to the underlying ESP32-HUB75-MatrixPanel-I2S-DMA library (in this example, it is the 'dmaOutput' class).
diff --git a/examples/README.md b/examples/README.md
index 2acbd79..5197413 100644
--- a/examples/README.md
+++ b/examples/README.md
@@ -6,8 +6,9 @@
 |AnimatedGIFPanel           |Using Larry Bank's GIF Decoder to display animated GIFs.                                             |
 |AuroraDemo                 |Simple example demonstrating various animated effects.                                                         |
 |BitmapIcons                |Simple example of how to display a bitmap image to the display.                                                        |
-|ChainedPanels              |Popular example on how to use the 'VirtualDisplay' class to chain multiple LED Matrix Panels to form a much bigger display! Refer to the README within this example's folder! |
+|ChainedPanels              |Popular example on how to use the 'VirtualMatrixPanel' class to chain multiple LED Matrix Panels to form a much bigger display! Refer to the README within this example's folder! |
 |ChainedPanelsAuroraDemo    |As above, but showing a large trippy plasma animation.                                                                 |
-|One_Quarter_1_4_ScanPanel  |Using this library with a 32w x 16h 1/4 Scan LED Matrix Panel. Custom co-ordinate remapping logic required.                               |
+|ChainedPanelsScreenBuffer  |Using the same 'VirtualMatrixPanel' class but also implementing a FastLED off-screen pixel buffer to do cool stuff. |
+|One_Quarter_1_4_ScanPanel  |Using this library with a 32w x 16h 1/4 Scan LED Matrix Panel. Custom co-ordinate remapping logic required.  NOT WORKING.                             |
 |One_Eighth_1_8_ScanPanel   |Using this library with a 64w x 32h 1/8 Scan LED Matrix Panel. Custom co-ordinate remapping logic required.
 |PIO_TestPatterns           |Non-Arduino example of how to display basic shapes.                                                                    |