diff --git a/src/r_draw_pal.cpp b/src/r_draw_pal.cpp
index 5a6c94327..33c482934 100644
--- a/src/r_draw_pal.cpp
+++ b/src/r_draw_pal.cpp
@@ -314,17 +314,10 @@ namespace swrenderer
 			uint8_t pix = source[frac >> bits];
 			if (pix != 0)
 			{
-#ifdef NO_RGB666
-				uint32_t fg = fg2rgb[colormap[pix]];
-				uint32_t bg = bg2rgb[*dest];
-				fg = (fg + bg) | 0x1f07c1f;
-				*dest = RGB32k.All[fg & (fg >> 15)];
-#else
 				uint32_t r = MIN(GPalette.BaseColors[colormap[pix]].r + GPalette.BaseColors[*dest].r, 255);
 				uint32_t g = MIN(GPalette.BaseColors[colormap[pix]].g + GPalette.BaseColors[*dest].g, 255);
 				uint32_t b = MIN(GPalette.BaseColors[colormap[pix]].b + GPalette.BaseColors[*dest].b, 255);
 				*dest = RGB256k.RGB[r>>2][g>>2][b>>2];
-#endif
 			}
 			frac += fracstep;
 			dest += pitch;
@@ -364,17 +357,10 @@ namespace swrenderer
 				uint8_t pix = _bufplce[i][vplce[i] >> bits];
 				if (pix != 0)
 				{
-#ifdef NO_RGB666
-					uint32_t fg = fg2rgb[_palookupoffse[i][pix]];
-					uint32_t bg = bg2rgb[dest[i]];
-					fg = (fg + bg) | 0x1f07c1f;
-					dest[i] = RGB32k.All[fg & (fg >> 15)];
-#else
 					uint32_t r = MIN(GPalette.BaseColors[_palookupoffse[i][pix]].r + GPalette.BaseColors[dest[i]].r, 255);
 					uint32_t g = MIN(GPalette.BaseColors[_palookupoffse[i][pix]].g + GPalette.BaseColors[dest[i]].g, 255);
 					uint32_t b = MIN(GPalette.BaseColors[_palookupoffse[i][pix]].b + GPalette.BaseColors[dest[i]].b, 255);
 					dest[i] = RGB256k.RGB[r>>2][g>>2][b>>2];
-#endif
 				}
 				vplce[i] += vince[i];
 			}
@@ -617,21 +603,10 @@ namespace swrenderer
 				uint8_t pix = _bufplce[i][vplce[i] >> bits];
 				if (pix != 0)
 				{
-#ifdef NO_RGB666
-					uint32_t a = (bg2rgb[dest[i]] | 0x40100400) - fg2rgb[_palookupoffse[i][pix]];
-					uint32_t b = a;
-
-					b &= 0x40100400;
-					b = b - (b >> 5);
-					a &= b;
-					a |= 0x01f07c1f;
-					dest[i] = RGB32k.All[a & (a >> 15)];
-#else
 					uint32_t r = clamp(GPalette.BaseColors[_palookupoffse[i][pix]].r - GPalette.BaseColors[dest[i]].r, 0, 255);
 					uint32_t g = clamp(GPalette.BaseColors[_palookupoffse[i][pix]].g - GPalette.BaseColors[dest[i]].g, 0, 255);
 					uint32_t b = clamp(GPalette.BaseColors[_palookupoffse[i][pix]].b - GPalette.BaseColors[dest[i]].b, 0, 255);
 					dest[i] = RGB256k.RGB[r>>2][g>>2][b>>2];
-#endif
 				}
 				vplce[i] += vince[i];
 			}
@@ -715,11 +690,7 @@ namespace swrenderer
 				c_red = (c_red * alpha_bottom + solid_bottom_r * inv_alpha_bottom) >> 8;
 				c_green = (c_green * alpha_bottom + solid_bottom_g * inv_alpha_bottom) >> 8;
 				c_blue = (c_blue * alpha_bottom + solid_bottom_b * inv_alpha_bottom) >> 8;
-#ifdef NO_RGB666
-				*dest = RGB32k.RGB[(c_red >> 3)][(c_green >> 3)][(c_blue >> 3)];
-#else
 				*dest = RGB256k.RGB[(c_red >> 2)][(c_green >> 2)][(c_blue >> 2)];
-#endif
 			}
 
 			frac += fracstep;
@@ -747,13 +718,8 @@ namespace swrenderer
 		int solid_bottom_r = RPART(solid_bottom);
 		int solid_bottom_g = GPART(solid_bottom);
 		int solid_bottom_b = BPART(solid_bottom);
-#ifdef NO_RGB666
-		uint32_t solid_top_fill = RGB32k.RGB[(solid_top_r >> 3)][(solid_top_g >> 3)][(solid_top_b >> 3)];
-		uint32_t solid_bottom_fill = RGB32k.RGB[(solid_bottom_r >> 3)][(solid_bottom_g >> 3)][(solid_bottom_b >> 3)];
-#else
 		uint32_t solid_top_fill = RGB256k.RGB[(solid_top_r >> 2)][(solid_top_g >> 2)][(solid_top_b >> 2)];
 		uint32_t solid_bottom_fill = RGB256k.RGB[(solid_bottom_r >> 2)][(solid_bottom_g >> 2)][(solid_bottom_b >> 2)];
-#endif
 		solid_top_fill = (solid_top_fill << 24) | (solid_top_fill << 16) | (solid_top_fill << 8) | solid_top_fill;
 		solid_bottom_fill = (solid_bottom_fill << 24) | (solid_bottom_fill << 16) | (solid_bottom_fill << 8) | solid_bottom_fill;
 
@@ -813,11 +779,7 @@ namespace swrenderer
 				c_red = (c_red * alpha_top + solid_top_r * inv_alpha_top) >> 8;
 				c_green = (c_green * alpha_top + solid_top_g * inv_alpha_top) >> 8;
 				c_blue = (c_blue * alpha_top + solid_top_b * inv_alpha_top) >> 8;
-#ifdef NO_RGB666
-				output[col] = RGB32k.RGB[(c_red >> 3)][(c_green >> 3)][(c_blue >> 3)];
-#else
 				output[col] = RGB256k.RGB[(c_red >> 2)][(c_green >> 2)][(c_blue >> 2)];
-#endif
 				frac[col] += fracstep[col];
 			}
 			*((uint32_t*)dest) = *((uint32_t*)output);
@@ -858,11 +820,7 @@ namespace swrenderer
 				c_red = (c_red * alpha_bottom + solid_bottom_r * inv_alpha_bottom) >> 8;
 				c_green = (c_green * alpha_bottom + solid_bottom_g * inv_alpha_bottom) >> 8;
 				c_blue = (c_blue * alpha_bottom + solid_bottom_b * inv_alpha_bottom) >> 8;
-#ifdef NO_RGB666
-				output[col] = RGB32k.RGB[(c_red >> 3)][(c_green >> 3)][(c_blue >> 3)];
-#else
 				output[col] = RGB256k.RGB[(c_red >> 2)][(c_green >> 2)][(c_blue >> 2)];
-#endif
 
 				frac[col] += fracstep[col];
 			}
@@ -944,11 +902,7 @@ namespace swrenderer
 				c_red = (c_red * alpha_bottom + solid_bottom_r * inv_alpha_bottom) >> 8;
 				c_green = (c_green * alpha_bottom + solid_bottom_g * inv_alpha_bottom) >> 8;
 				c_blue = (c_blue * alpha_bottom + solid_bottom_b * inv_alpha_bottom) >> 8;
-#ifdef NO_RGB666
-				*dest = RGB32k.RGB[(c_red >> 3)][(c_green >> 3)][(c_blue >> 3)];
-#else
 				*dest = RGB256k.RGB[(c_red >> 2)][(c_green >> 2)][(c_blue >> 2)];
-#endif
 			}
 
 			frac += fracstep;
@@ -978,13 +932,8 @@ namespace swrenderer
 		int solid_bottom_r = RPART(solid_bottom);
 		int solid_bottom_g = GPART(solid_bottom);
 		int solid_bottom_b = BPART(solid_bottom);
-#ifdef NO_RGB666
-		uint32_t solid_top_fill = RGB32k.RGB[(solid_top_r >> 3)][(solid_top_g >> 3)][(solid_top_b >> 3)];
-		uint32_t solid_bottom_fill = RGB32k.RGB[(solid_bottom_r >> 3)][(solid_bottom_g >> 3)][(solid_bottom_b >> 3)];
-#else
 		uint32_t solid_top_fill = RGB256k.RGB[(solid_top_r >> 2)][(solid_top_g >> 2)][(solid_top_b >> 2)];
 		uint32_t solid_bottom_fill = RGB256k.RGB[(solid_bottom_r >> 2)][(solid_bottom_g >> 2)][(solid_bottom_b >> 2)];
-#endif
 		solid_top_fill = (solid_top_fill << 24) | (solid_top_fill << 16) | (solid_top_fill << 8) | solid_top_fill;
 		solid_bottom_fill = (solid_bottom_fill << 24) | (solid_bottom_fill << 16) | (solid_bottom_fill << 8) | solid_bottom_fill;
 
@@ -1050,11 +999,7 @@ namespace swrenderer
 				c_red = (c_red * alpha_top + solid_top_r * inv_alpha_top) >> 8;
 				c_green = (c_green * alpha_top + solid_top_g * inv_alpha_top) >> 8;
 				c_blue = (c_blue * alpha_top + solid_top_b * inv_alpha_top) >> 8;
-#ifdef NO_RGB666
-				output[col] = RGB32k.RGB[(c_red >> 3)][(c_green >> 3)][(c_blue >> 3)];
-#else
 				output[col] = RGB256k.RGB[(c_red >> 2)][(c_green >> 2)][(c_blue >> 2)];
-#endif
 
 				frac[col] += fracstep[col];
 			}
@@ -1108,11 +1053,7 @@ namespace swrenderer
 				c_red = (c_red * alpha_bottom + solid_bottom_r * inv_alpha_bottom) >> 8;
 				c_green = (c_green * alpha_bottom + solid_bottom_g * inv_alpha_bottom) >> 8;
 				c_blue = (c_blue * alpha_bottom + solid_bottom_b * inv_alpha_bottom) >> 8;
-#ifdef NO_RGB666
-				output[col] = RGB32k.RGB[(c_red >> 3)][(c_green >> 3)][(c_blue >> 3)];
-#else
 				output[col] = RGB256k.RGB[(c_red >> 2)][(c_green >> 2)][(c_blue >> 2)];
-#endif
 
 				frac[col] += fracstep[col];
 			}
@@ -1244,9 +1185,12 @@ namespace swrenderer
 
 		do
 		{
-			uint32_t bg;
-			bg = (fg + bg2rgb[*dest]) | 0x1f07c1f;
-			*dest = RGB32k.All[bg & (bg >> 15)];
+			const PalEntry* pal = GPalette.BaseColors;
+			// *** [SP] this is incomplete, not sure what to do here.
+			/*int r = clamp((int)pal[_srccolor].r, 0, 255) >> 2;
+			int g = clamp((int)pal[_srccolor].g, 0, 255) >> 2;
+			int b = clamp((int)pal[_srccolor].b, 0, 255) >> 2;
+			*dest = RGB256k.RGB[r][g][b];*/
 			dest += pitch;
 		} while (--count);
 
@@ -1276,15 +1220,11 @@ namespace swrenderer
 
 		do
 		{
-			uint32_t a = fg + bg2rgb[*dest];
-			uint32_t b = a;
-
-			a |= 0x01f07c1f;
-			b &= 0x40100400;
-			a &= 0x3fffffff;
-			b = b - (b >> 5);
-			a |= b;
-			*dest = RGB32k.All[a & (a >> 15)];
+			const PalEntry* pal = GPalette.BaseColors;
+			int r = clamp(pal[*dest].r + pal[fg].r, 0, 255) >> 2;
+			int g = clamp(pal[*dest].g + pal[fg].g, 0, 255) >> 2;
+			int b = clamp(pal[*dest].b + pal[fg].b, 0, 255) >> 2;
+			*dest = RGB256k.RGB[r][g][b];
 			dest += pitch;
 		} while (--count);
 	}
diff --git a/src/r_drawt_pal.cpp b/src/r_drawt_pal.cpp
index 16d24ecaf..156e97ba6 100644
--- a/src/r_drawt_pal.cpp
+++ b/src/r_drawt_pal.cpp
@@ -446,14 +446,15 @@ namespace swrenderer
 		source = &thread->dc_temp[thread->temp_line_for_thread(yl)*4 + hx];
 		pitch = _pitch * thread->num_cores;
 		colormap = _colormap;
+		const PalEntry *palette = GPalette.BaseColors;
 
 		do {
 			uint32_t fg = colormap[*source];
 			uint32_t bg = *dest;
 
-			int r = MIN((GPalette.BaseColors[fg].r * _srcalpha + GPalette.BaseColors[bg].r * _destalpha)>>18, 63);
-			int g = MIN((GPalette.BaseColors[fg].g * _srcalpha + GPalette.BaseColors[bg].g * _destalpha)>>18, 63);
-			int b = MIN((GPalette.BaseColors[fg].b * _srcalpha + GPalette.BaseColors[bg].b * _destalpha)>>18, 63);
+			int r = MIN((palette[fg].r * _srcalpha + palette[bg].r * _destalpha)>>18, 63);
+			int g = MIN((palette[fg].g * _srcalpha + palette[bg].g * _destalpha)>>18, 63);
+			int b = MIN((palette[fg].b * _srcalpha + palette[bg].b * _destalpha)>>18, 63);
 			*dest = RGB256k.RGB[r][g][b];
 			source += 4;
 			dest += pitch;
@@ -476,15 +477,16 @@ namespace swrenderer
 		source = &thread->dc_temp[thread->temp_line_for_thread(yl)*4];
 		pitch = _pitch * thread->num_cores;
 		colormap = _colormap;
+		const PalEntry *palette = GPalette.BaseColors;
 
 		do {
 			for (int ks = 0; ks < 4; ks++)
 			{	// [SP] this 4col function was a block of copy-pasted code. 4 times. I regret nothing.
 				uint32_t fg = colormap[source[ks]];
 				uint32_t bg = dest[ks];
-				int r = MIN((GPalette.BaseColors[fg].r * _srcalpha + GPalette.BaseColors[bg].r * _destalpha)>>18, 63);
-				int g = MIN((GPalette.BaseColors[fg].g * _srcalpha + GPalette.BaseColors[bg].g * _destalpha)>>18, 63);
-				int b = MIN((GPalette.BaseColors[fg].b * _srcalpha + GPalette.BaseColors[bg].b * _destalpha)>>18, 63);
+				int r = MIN((palette[fg].r * _srcalpha + palette[bg].r * _destalpha)>>18, 63);
+				int g = MIN((palette[fg].g * _srcalpha + palette[bg].g * _destalpha)>>18, 63);
+				int b = MIN((palette[fg].b * _srcalpha + palette[bg].b * _destalpha)>>18, 63);
 				dest[ks] = RGB256k.RGB[r][g][b];
 			}
 
@@ -511,13 +513,14 @@ namespace swrenderer
 		dest = ylookup[yl + thread->skipped_by_thread(yl)] + sx + _destorg;
 		source = &thread->dc_temp[thread->temp_line_for_thread(yl)*4 + hx];
 		pitch = _pitch * thread->num_cores;
+		const PalEntry *palette = GPalette.BaseColors;
 
 		do {
 			uint32_t val = *source;
-			int r = (GPalette.BaseColors[*dest].r * (63-val) + GPalette.BaseColors[_color].r * val) >> 8;
-			int g = (GPalette.BaseColors[*dest].g * (63-val) + GPalette.BaseColors[_color].g * val) >> 8;
-			int b = (GPalette.BaseColors[*dest].b * (63-val) + GPalette.BaseColors[_color].b * val) >> 8;
-			*dest = RGB256k.RGB[MIN(r,63)][MIN(g,63)][MIN(b,63)];
+			int r = (palette[*dest].r * (255-val)) >> 10;
+			int g = (palette[*dest].g * (255-val)) >> 10;
+			int b = (palette[*dest].b * (255-val)) >> 10;
+			*dest = RGB256k.RGB[clamp(r,0,63)][clamp(g,0,63)][clamp(b,0,63)];
 			source += 4;
 			dest += pitch;
 		} while (--count);
@@ -541,6 +544,7 @@ namespace swrenderer
 		dest = ylookup[yl + thread->skipped_by_thread(yl)] + sx + _destorg;
 		source = &thread->dc_temp[thread->temp_line_for_thread(yl)*4];
 		pitch = _pitch * thread->num_cores;
+		const PalEntry *palette = GPalette.BaseColors;
 
 		do {
 			uint32_t val;
@@ -548,10 +552,10 @@ namespace swrenderer
 			for (int ks = 0; ks < 4; ks++)
 			{
 				val = source[ks];
-				int r = (GPalette.BaseColors[dest[ks]].r * (63-val) + GPalette.BaseColors[_color].r * val) >> 8;
-				int g = (GPalette.BaseColors[dest[ks]].g * (63-val) + GPalette.BaseColors[_color].g * val) >> 8;
-				int b = (GPalette.BaseColors[dest[ks]].b * (63-val) + GPalette.BaseColors[_color].b * val) >> 8;
-				dest[ks] = RGB256k.RGB[MIN(r,63)][MIN(g,63)][MIN(b,63)];
+				int r = (palette[dest[ks]].r * (255-val)) >> 10;
+				int g = (palette[dest[ks]].g * (255-val)) >> 10;
+				int b = (palette[dest[ks]].b * (255-val)) >> 10;
+				dest[ks] = RGB256k.RGB[clamp(r,0,63)][clamp(g,0,63)][clamp(b,0,63)];
 			}
 
 			source += 4;
@@ -575,13 +579,14 @@ namespace swrenderer
 		source = &thread->dc_temp[thread->temp_line_for_thread(yl)*4 + hx];
 		pitch = _pitch * thread->num_cores;
 		colormap = _colormap;
+		const PalEntry *palette = GPalette.BaseColors;
 
 		do {
 			int fg = *source;
 			int bg = *dest;
-			int r = MIN((GPalette.BaseColors[fg].r * _srcalpha + GPalette.BaseColors[bg].r * _destalpha)>>18, 63);
-			int g = MIN((GPalette.BaseColors[fg].g * _srcalpha + GPalette.BaseColors[bg].g * _destalpha)>>18, 63);
-			int b = MIN((GPalette.BaseColors[fg].b * _srcalpha + GPalette.BaseColors[bg].b * _destalpha)>>18, 63);
+			int r = MIN((palette[fg].r * _srcalpha + palette[bg].r * _destalpha)>>18, 63);
+			int g = MIN((palette[fg].g * _srcalpha + palette[bg].g * _destalpha)>>18, 63);
+			int b = MIN((palette[fg].b * _srcalpha + palette[bg].b * _destalpha)>>18, 63);
 			*dest = RGB256k.RGB[r][g][b];
 			source += 4;
 			dest += pitch;
@@ -604,15 +609,16 @@ namespace swrenderer
 		source = &thread->dc_temp[thread->temp_line_for_thread(yl)*4];
 		pitch = _pitch * thread->num_cores;
 		colormap = _colormap;
+		const PalEntry *palette = GPalette.BaseColors;
 
 		do {
 			for (int ks = 0; ks < 4; ks++)
 			{
 				int fg = source[ks];
 				int bg = dest[ks];
-				int r = MIN((GPalette.BaseColors[fg].r * _srcalpha + GPalette.BaseColors[bg].r * _destalpha)>>18, 63);
-				int g = MIN((GPalette.BaseColors[fg].g * _srcalpha + GPalette.BaseColors[bg].g * _destalpha)>>18, 63);
-				int b = MIN((GPalette.BaseColors[fg].b * _srcalpha + GPalette.BaseColors[bg].b * _destalpha)>>18, 63);
+				int r = MIN((palette[fg].r * _srcalpha + palette[bg].r * _destalpha)>>18, 63);
+				int g = MIN((palette[fg].g * _srcalpha + palette[bg].g * _destalpha)>>18, 63);
+				int b = MIN((palette[fg].b * _srcalpha + palette[bg].b * _destalpha)>>18, 63);
 				dest[ks] = RGB256k.RGB[r][g][b];
 			}
 
@@ -639,16 +645,15 @@ namespace swrenderer
 		source = &thread->dc_temp[thread->temp_line_for_thread(yl)*4 + hx];
 		pitch = _pitch * thread->num_cores;
 		colormap = _colormap;
+		const PalEntry *palette = GPalette.BaseColors;
 
 		do {
-			uint32_t a = (fg2rgb[colormap[*source]] | 0x40100400) - bg2rgb[*dest];
-			uint32_t b = a;
-
-			b &= 0x40100400;
-			b = b - (b >> 5);
-			a &= b;
-			a |= 0x01f07c1f;
-			*dest = RGB32k.All[(a>>15) & a];
+			int fg = *source;
+			int bg = *dest;
+			int r = MAX((palette[fg].r * _srcalpha - palette[bg].r * _destalpha)>>18, 0);
+			int g = MAX((palette[fg].g * _srcalpha - palette[bg].g * _destalpha)>>18, 0);
+			int b = MAX((palette[fg].b * _srcalpha - palette[bg].b * _destalpha)>>18, 0);
+			*dest = RGB256k.RGB[r][g][b];
 			source += 4;
 			dest += pitch;
 		} while (--count);
@@ -672,40 +677,18 @@ namespace swrenderer
 		source = &thread->dc_temp[thread->temp_line_for_thread(yl)*4];
 		pitch = _pitch * thread->num_cores;
 		colormap = _colormap;
+		const PalEntry *palette = GPalette.BaseColors;
 
 		do {
-			uint32_t a = (fg2rgb[colormap[source[0]]] | 0x40100400) - bg2rgb[dest[0]];
-			uint32_t b = a;
-
-			b &= 0x40100400;
-			b = b - (b >> 5);
-			a &= b;
-			a |= 0x01f07c1f;
-			dest[0] = RGB32k.All[(a>>15) & a];
-
-			a = (fg2rgb[colormap[source[1]]] | 0x40100400) - bg2rgb[dest[1]];
-			b = a;
-			b &= 0x40100400;
-			b = b - (b >> 5);
-			a &= b;
-			a |= 0x01f07c1f;
-			dest[1] = RGB32k.All[(a>>15) & a];
-
-			a = (fg2rgb[colormap[source[2]]] | 0x40100400) - bg2rgb[dest[2]];
-			b = a;
-			b &= 0x40100400;
-			b = b - (b >> 5);
-			a &= b;
-			a |= 0x01f07c1f;
-			dest[2] = RGB32k.All[(a>>15) & a];
-
-			a = (fg2rgb[colormap[source[3]]] | 0x40100400) - bg2rgb[dest[3]];
-			b = a;
-			b &= 0x40100400;
-			b = b - (b >> 5);
-			a &= b;
-			a |= 0x01f07c1f;
-			dest[3] = RGB32k.All[(a>>15) & a];
+			for (int ks = 0; ks < 4; ks++)
+			{
+				int fg = source[ks];
+				int bg = dest[ks];
+				int r = MAX((palette[fg].r * _srcalpha - palette[bg].r * _destalpha)>>18, 0);
+				int g = MAX((palette[fg].g * _srcalpha - palette[bg].g * _destalpha)>>18, 0);
+				int b = MAX((palette[fg].b * _srcalpha - palette[bg].b * _destalpha)>>18, 0);
+				dest[ks] = RGB256k.RGB[r][g][b];
+			}
 
 			source += 4;
 			dest += pitch;
@@ -730,16 +713,15 @@ namespace swrenderer
 		source = &thread->dc_temp[thread->temp_line_for_thread(yl)*4 + hx];
 		pitch = _pitch * thread->num_cores;
 		colormap = _colormap;
+		const PalEntry *palette = GPalette.BaseColors;
 
 		do {
-			uint32_t a = (bg2rgb[*dest] | 0x40100400) - fg2rgb[colormap[*source]];
-			uint32_t b = a;
-
-			b &= 0x40100400;
-			b = b - (b >> 5);
-			a &= b;
-			a |= 0x01f07c1f;
-			*dest = RGB32k.All[(a>>15) & a];
+			int fg = *source;
+			int bg = *dest;
+			int r = MAX((-palette[fg].r * _srcalpha + palette[bg].r * _destalpha)>>18, 0);
+			int g = MAX((-palette[fg].g * _srcalpha + palette[bg].g * _destalpha)>>18, 0);
+			int b = MAX((-palette[fg].b * _srcalpha + palette[bg].b * _destalpha)>>18, 0);
+			*dest = RGB256k.RGB[r][g][b];
 			source += 4;
 			dest += pitch;
 		} while (--count);
@@ -763,40 +745,18 @@ namespace swrenderer
 		source = &thread->dc_temp[thread->temp_line_for_thread(yl)*4];
 		pitch = _pitch * thread->num_cores;
 		colormap = _colormap;
+		const PalEntry *palette = GPalette.BaseColors;
 
 		do {
-			uint32_t a = (bg2rgb[dest[0]] | 0x40100400) - fg2rgb[colormap[source[0]]];
-			uint32_t b = a;
-
-			b &= 0x40100400;
-			b = b - (b >> 5);
-			a &= b;
-			a |= 0x01f07c1f;
-			dest[0] = RGB32k.All[(a>>15) & a];
-
-			a = (bg2rgb[dest[1]] | 0x40100400) - fg2rgb[colormap[source[1]]];
-			b = a;
-			b &= 0x40100400;
-			b = b - (b >> 5);
-			a &= b;
-			a |= 0x01f07c1f;
-			dest[1] = RGB32k.All[(a>>15) & a];
-
-			a = (bg2rgb[dest[2]] | 0x40100400) - fg2rgb[colormap[source[2]]];
-			b = a;
-			b &= 0x40100400;
-			b = b - (b >> 5);
-			a &= b;
-			a |= 0x01f07c1f;
-			dest[2] = RGB32k.All[(a>>15) & a];
-
-			a = (bg2rgb[dest[3]] | 0x40100400) - fg2rgb[colormap[source[3]]];
-			b = a;
-			b &= 0x40100400;
-			b = b - (b >> 5);
-			a &= b;
-			a |= 0x01f07c1f;
-			dest[3] = RGB32k.All[(a>>15) & a];
+			for (int ks = 0; ks < 4; ks++)
+			{
+				int fg = source[ks];
+				int bg = dest[ks];
+				int r = MAX((-palette[fg].r * _srcalpha + palette[bg].r * _destalpha)>>18, 0);
+				int g = MAX((-palette[fg].g * _srcalpha + palette[bg].g * _destalpha)>>18, 0);
+				int b = MAX((-palette[fg].b * _srcalpha + palette[bg].b * _destalpha)>>18, 0);
+				dest[ks] = RGB256k.RGB[r][g][b];
+			}
 
 			source += 4;
 			dest += pitch;
diff --git a/src/textures/pcxtexture.cpp b/src/textures/pcxtexture.cpp
index 0ec5d2933..dda431993 100644
--- a/src/textures/pcxtexture.cpp
+++ b/src/textures/pcxtexture.cpp
@@ -528,7 +528,7 @@ void FPCXTexture::MakeTexture()
 		{
 			for(int x=0; x < Width; x++)
 			{
-				Pixels[y+Height*x] = RGB32k.RGB[row[0]>>3][row[1]>>3][row[2]>>3];
+				Pixels[y+Height*x] = RGB256k.RGB[row[0]>>2][row[1]>>2][row[2]>>2];
 				row+=3;
 			}
 		}