diff --git a/wled00/FX.h b/wled00/FX.h
index ceb5dfc0..046b7d5b 100644
--- a/wled00/FX.h
+++ b/wled00/FX.h
@@ -598,7 +598,7 @@ typedef struct Segment {
     void setPixelColorXY(int x, int y, uint32_t c); // set relative pixel within segment with color
     void setPixelColorXY(int x, int y, byte r, byte g, byte b, byte w = 0) { setPixelColorXY(x, y, RGBW32(r,g,b,w)); } // automatically inline
     void setPixelColorXY(int x, int y, CRGB c)                             { setPixelColorXY(x, y, RGBW32(c.r,c.g,c.b,0)); } // automatically inline
-    void setPixelColorXY(float x, float y, uint32_t c, bool aa = true);
+    void setPixelColorXY(float x, float y, uint32_t c, bool aa = true, bool fast = true);
     void setPixelColorXY(float x, float y, byte r, byte g, byte b, byte w = 0, bool aa = true) { setPixelColorXY(x, y, RGBW32(r,g,b,w), aa); }
     void setPixelColorXY(float x, float y, CRGB c, bool aa = true)                             { setPixelColorXY(x, y, RGBW32(c.r,c.g,c.b,0), aa); }
     uint32_t getPixelColorXY(uint16_t x, uint16_t y);
diff --git a/wled00/FX_2Dfcn.cpp b/wled00/FX_2Dfcn.cpp
index d862632d..c5c77bb4 100644
--- a/wled00/FX_2Dfcn.cpp
+++ b/wled00/FX_2Dfcn.cpp
@@ -242,13 +242,13 @@ void IRAM_ATTR_YN Segment::setPixelColorXY(int x, int y, uint32_t col) //WLEDMM:
 }
 
 // anti-aliased version of setPixelColorXY()
-void Segment::setPixelColorXY(float x, float y, uint32_t col, bool aa)
+void Segment::setPixelColorXY(float x, float y, uint32_t col, bool aa, bool fast) // WLEDMM some speedups due to fast int and faster sqrt16
 {
   if (Segment::maxHeight==1) return; // not a matrix set-up
   if (x<0.0f || x>1.0f || y<0.0f || y>1.0f) return; // not normalized
 
-  const uint16_t cols = virtualWidth();
-  const uint16_t rows = virtualHeight();
+  const uint_fast16_t cols = virtualWidth();
+  const uint_fast16_t rows = virtualHeight();
 
   float fX = x * (cols-1);
   float fY = y * (rows-1);
@@ -267,10 +267,17 @@ void Segment::setPixelColorXY(float x, float y, uint32_t col, bool aa)
     uint32_t cXRYB = getPixelColorXY(xR, yB);
 
     if (xL!=xR && yT!=yB) {
-      setPixelColorXY(xL, yT, color_blend(col, cXLYT, uint8_t(sqrtf(dL*dT)*255.0f))); // blend TL pixel
-      setPixelColorXY(xR, yT, color_blend(col, cXRYT, uint8_t(sqrtf(dR*dT)*255.0f))); // blend TR pixel
-      setPixelColorXY(xL, yB, color_blend(col, cXLYB, uint8_t(sqrtf(dL*dB)*255.0f))); // blend BL pixel
-      setPixelColorXY(xR, yB, color_blend(col, cXRYB, uint8_t(sqrtf(dR*dB)*255.0f))); // blend BR pixel
+      if (!fast) {
+        setPixelColorXY(xL, yT, color_blend(col, cXLYT, uint8_t(sqrtf(dL*dT)*255.0f))); // blend TL pixel
+        setPixelColorXY(xR, yT, color_blend(col, cXRYT, uint8_t(sqrtf(dR*dT)*255.0f))); // blend TR pixel
+        setPixelColorXY(xL, yB, color_blend(col, cXLYB, uint8_t(sqrtf(dL*dB)*255.0f))); // blend BL pixel
+        setPixelColorXY(xR, yB, color_blend(col, cXRYB, uint8_t(sqrtf(dR*dB)*255.0f))); // blend BR pixel
+      } else {
+        setPixelColorXY(xL, yT, color_blend(col, cXLYT, uint8_t(sqrt16(dL*dT*65025.0f)))); // blend TL pixel     // WLEDMM: use faster sqrt16 for integer; perform multiplication by 255^2 before sqrt
+        setPixelColorXY(xR, yT, color_blend(col, cXRYT, uint8_t(sqrt16(dR*dT*65025.0f)))); // blend TR pixel     //         this is possible because sqrt(a) * sqrt(b)  =  sqrt(a * b)
+        setPixelColorXY(xL, yB, color_blend(col, cXLYB, uint8_t(sqrt16(dL*dB*65025.0f)))); // blend BL pixel
+        setPixelColorXY(xR, yB, color_blend(col, cXRYB, uint8_t(sqrt16(dR*dB*65025.0f)))); // blend BR pixel
+      }
     } else if (xR!=xL && yT==yB) {
       setPixelColorXY(xR, yT, color_blend(col, cXLYT, uint8_t(dL*255.0f))); // blend L pixel
       setPixelColorXY(xR, yT, color_blend(col, cXRYT, uint8_t(dR*255.0f))); // blend R pixel