diff --git a/bin/resources/shaders/dx11/tfx.fx b/bin/resources/shaders/dx11/tfx.fx
index 0e9167720b..04eb4b3f7b 100644
--- a/bin/resources/shaders/dx11/tfx.fx
+++ b/bin/resources/shaders/dx11/tfx.fx
@@ -766,7 +766,7 @@ float4 ps_color(PS_INPUT input)
 	float4 T = sample_color(st, input.t.w);
 #endif
 
-	if ((SW_BLEND || PS_TFX != 1) && PS_SHUFFLE && !PS_SHUFFLE_SAME && !PS_READ16_SRC && (PS_SHUFFLE_ACROSS || PS_PROCESS_BA == SHUFFLE_READWRITE || PS_PROCESS_RG == SHUFFLE_READWRITE))
+	if (PS_SHUFFLE && !PS_SHUFFLE_SAME && !PS_READ16_SRC)
 	{
 		uint4 denorm_c_before = uint4(T);
 		if (PS_PROCESS_BA & SHUFFLE_READ)
@@ -783,6 +783,8 @@ float4 ps_color(PS_INPUT input)
 			T.b = float((denorm_c_before.g << 1) & 0xF8);
 			T.a = float(denorm_c_before.g & 0x80);
 		}
+		
+		T.a = (T.a >= 127.5f ? TA.y : !PS_AEM || any(int3(T.rgb) & 0xF8) ? TA.x : 0) * 255.0f;
 	}
 
 	float4 C = tfx(T, input.c);
@@ -1057,7 +1059,7 @@ PS_OUTPUT ps_main(PS_INPUT input)
 
 	if (PS_SHUFFLE)
 	{
-		if ((SW_BLEND || PS_TFX != 1) && PS_SHUFFLE && !PS_SHUFFLE_SAME && !PS_READ16_SRC && (PS_SHUFFLE_ACROSS || PS_PROCESS_BA == SHUFFLE_READWRITE || PS_PROCESS_RG == SHUFFLE_READWRITE))
+		if (!PS_SHUFFLE_SAME && !PS_READ16_SRC)
 		{
 			uint4 denorm_c_after = uint4(C);
 			if (PS_PROCESS_BA & SHUFFLE_READ)
@@ -1072,12 +1074,12 @@ PS_OUTPUT ps_main(PS_INPUT input)
 			}
 		}
 
-		uint4 denorm_c = uint4(C);
-		uint2 denorm_TA = uint2(float2(TA.xy) * 255.0f + 0.5f);
 
 		// Special case for 32bit input and 16bit output, shuffle used by The Godfather
 		if (PS_SHUFFLE_SAME)
 		{
+			uint4 denorm_c = uint4(C);
+			
 			if (PS_PROCESS_BA & SHUFFLE_READ)
 				C = (float4)(float((denorm_c.b & 0x7Fu) | (denorm_c.a & 0x80u)));
 			else
@@ -1086,6 +1088,8 @@ PS_OUTPUT ps_main(PS_INPUT input)
 		// Copy of a 16bit source in to this target
 		else if (PS_READ16_SRC)
 		{
+			uint4 denorm_c = uint4(C);
+			uint2 denorm_TA = uint2(float2(TA.xy) * 255.0f + 0.5f);
 			C.rb = (float2)float((denorm_c.r >> 3) | (((denorm_c.g >> 3) & 0x7u) << 5));
 			if (denorm_c.a & 0x80u)
 				C.ga = (float2)float((denorm_c.g >> 6) | ((denorm_c.b >> 3) << 2) | (denorm_TA.y & 0x80u));
@@ -1097,44 +1101,22 @@ PS_OUTPUT ps_main(PS_INPUT input)
 			if (PS_PROCESS_BA == SHUFFLE_READWRITE && PS_PROCESS_RG == SHUFFLE_READWRITE)
 			{
 				C.rb = C.br;
-				if ((denorm_c.a & 0x80u) != 0u)
-					C.g = float((denorm_c.a & 0x7Fu) | (denorm_TA.y & 0x80u));
-				else
-					C.g = float((denorm_c.a & 0x7Fu) | (denorm_TA.x & 0x80u));
-					
-				if ((denorm_c.g & 0x80u) != 0u)
-					C.a = float((denorm_c.g & 0x7Fu) | (denorm_TA.y & 0x80u));
-				else
-					C.a = float((denorm_c.g & 0x7Fu) | (denorm_TA.x & 0x80u));
+				float g_temp = C.g;
+				
+				C.g = C.a;
+				C.a = g_temp;
 			}
 			else if(PS_PROCESS_BA & SHUFFLE_READ)
 			{
 				C.rb = C.bb;
-				if ((denorm_c.a & 0x80u) != 0u)
-					C.ga =  (float2)(float((denorm_c.a & 0x7Fu) | (denorm_TA.y & 0x80u)));
-				else
-					C.ga =  (float2)(float((denorm_c.a & 0x7Fu) | (denorm_TA.x & 0x80u)));
+				C.ga = C.aa;
 			}
 			else
 			{
 				C.rb = C.rr;
-				if ((denorm_c.g & 0x80u) != 0u)
-					C.ga =  (float2)(float((denorm_c.g & 0x7Fu) | (denorm_TA.y & 0x80u)));
-				else
-					C.ga =  (float2)(float((denorm_c.g & 0x7Fu) | (denorm_TA.x & 0x80u)));
+				C.ga = C.gg;
 			}
 		}
-		else // Basically a direct copy but a shuffle of both pairs of channels, so green and alpha get modified by TEXA
-		{
-			if ((denorm_c.g & 0x80u) != 0u)
-				C.g = float((denorm_c.g & 0x7Fu) | (denorm_TA.y & 0x80u));
-			else
-				C.g = float((denorm_c.g & 0x7Fu) | (denorm_TA.x & 0x80u));
-			if ((denorm_c.a & 0x80u) != 0u)
-				C.a = float((denorm_c.a & 0x7Fu) | (denorm_TA.y & 0x80u));
-			else
-				C.a = float((denorm_c.a & 0x7Fu) | (denorm_TA.x & 0x80u));
-		}
 	}
 
 	ps_dither(C.rgb, alpha_blend.a, input.p.xy);
diff --git a/bin/resources/shaders/opengl/tfx_fs.glsl b/bin/resources/shaders/opengl/tfx_fs.glsl
index 4154e5b57e..a502a7faa7 100644
--- a/bin/resources/shaders/opengl/tfx_fs.glsl
+++ b/bin/resources/shaders/opengl/tfx_fs.glsl
@@ -686,7 +686,7 @@ vec4 ps_color()
 	vec4 T = sample_color(st);
 #endif
 
-	#if (SW_BLEND || PS_TFX != 1) && PS_SHUFFLE && !PS_SHUFFLE_SAME && !PS_READ16_SRC && (PS_SHUFFLE_ACROSS || PS_PROCESS_BA == SHUFFLE_READWRITE || PS_PROCESS_RG == SHUFFLE_READWRITE)
+	#if PS_SHUFFLE && !PS_READ16_SRC && !PS_SHUFFLE_SAME
 		uvec4 denorm_c_before = uvec4(T);
 		#if (PS_PROCESS_BA & SHUFFLE_READ)
 			T.r = float((denorm_c_before.b << 3) & 0xF8);
@@ -699,6 +699,8 @@ vec4 ps_color()
 			T.b = float((denorm_c_before.g << 1) & 0xF8);
 			T.a = float(denorm_c_before.g & 0x80);
 		#endif
+		
+		T.a = ((T.a >= 127.5f) ? TA.y : ((PS_AEM == 0 || any(bvec3(ivec3(T.rgb) & ivec3(0xF8)))) ? TA.x : 0.0f)) * 255.0f;
 	#endif
 	
 	vec4 C = tfx(T, PSin.c);
@@ -1042,7 +1044,7 @@ void ps_main()
 
 
 #if PS_SHUFFLE
-	#if (SW_BLEND || PS_TFX != 1) &&  PS_SHUFFLE && !PS_SHUFFLE_SAME && !PS_READ16_SRC && (PS_SHUFFLE_ACROSS || PS_PROCESS_BA == SHUFFLE_READWRITE || PS_PROCESS_RG == SHUFFLE_READWRITE)
+	#if !PS_READ16_SRC && !PS_SHUFFLE_SAME
 		uvec4 denorm_c_after = uvec4(C);
 		#if (PS_PROCESS_BA & SHUFFLE_READ)
 			C.b = float(((denorm_c_after.r >> 3) & 0x1F) | ((denorm_c_after.g << 2) & 0xE0));
@@ -1052,60 +1054,40 @@ void ps_main()
 			C.g = float(((denorm_c_after.g >> 6) & 0x3) | ((denorm_c_after.b >> 1) & 0x7C) | (denorm_c_after.a & 0x80));
 		#endif
 	#endif
-	
-	uvec4 denorm_c = uvec4(C);
-	uvec2 denorm_TA = uvec2(vec2(TA.xy) * 255.0f + 0.5f);
 
-// Special case for 32bit input and 16bit output, shuffle used by The Godfather
-#if PS_SHUFFLE_SAME
-#if (PS_PROCESS_BA & SHUFFLE_READ)
-	C = vec4(float((denorm_c.b & 0x7Fu) | (denorm_c.a & 0x80u)));
-#else
-	C.ga = C.rg;
-#endif
-// Copy of a 16bit source in to this target
-#elif PS_READ16_SRC
-	C.rb = vec2(float((denorm_c.r >> 3) | (((denorm_c.g >> 3) & 0x7u) << 5)));
-	if (bool(denorm_c.a & 0x80u))
-		C.ga = vec2(float((denorm_c.g >> 6) | ((denorm_c.b >> 3) << 2) | (denorm_TA.y & 0x80u)));
-	else
-		C.ga = vec2(float((denorm_c.g >> 6) | ((denorm_c.b >> 3) << 2) | (denorm_TA.x & 0x80u)));
-#elif PS_SHUFFLE_ACROSS
-	#if(PS_PROCESS_BA == SHUFFLE_READWRITE && PS_PROCESS_RG == SHUFFLE_READWRITE)
-		C.rb = C.br;
-		if ((denorm_c.a & 0x80u) != 0u)
-			C.g = float((denorm_c.a & 0x7Fu) | (denorm_TA.y & 0x80u));
-		else
-			C.g = float((denorm_c.a & 0x7Fu) | (denorm_TA.x & 0x80u));
-			
-		if ((denorm_c.g & 0x80u) != 0u)
-			C.a = float((denorm_c.g & 0x7Fu) | (denorm_TA.y & 0x80u));
-		else
-			C.a = float((denorm_c.g & 0x7Fu) | (denorm_TA.x & 0x80u));
-		
-	#elif(PS_PROCESS_BA & SHUFFLE_READ)
-		C.rb = C.bb;
-		if ((denorm_c.a & 0x80u) != 0u)
-			C.ga = vec2(float((denorm_c.a & 0x7Fu) | (denorm_TA.y & 0x80u)));
-		else
-			C.ga = vec2(float((denorm_c.a & 0x7Fu) | (denorm_TA.x & 0x80u)));
+	// Special case for 32bit input and 16bit output, shuffle used by The Godfather
+	#if PS_SHUFFLE_SAME
+		uvec4 denorm_c = uvec4(C);
+	#if (PS_PROCESS_BA & SHUFFLE_READ)
+		C = vec4(float((denorm_c.b & 0x7Fu) | (denorm_c.a & 0x80u)));
 	#else
-		C.rb = C.rr;
-		if ((denorm_c.g & 0x80u) != 0u)
-			C.ga = vec2(float((denorm_c.g & 0x7Fu) | (denorm_TA.y & 0x80u)));
+		C.ga = C.rg;
+	#endif
+	// Copy of a 16bit source in to this target
+	#elif PS_READ16_SRC
+		uvec4 denorm_c = uvec4(C);
+		uvec2 denorm_TA = uvec2(vec2(TA.xy) * 255.0f + 0.5f);
+		
+		C.rb = vec2(float((denorm_c.r >> 3) | (((denorm_c.g >> 3) & 0x7u) << 5)));
+		if (bool(denorm_c.a & 0x80u))
+			C.ga = vec2(float((denorm_c.g >> 6) | ((denorm_c.b >> 3) << 2) | (denorm_TA.y & 0x80u)));
 		else
-			C.ga = vec2(float((denorm_c.g & 0x7Fu) | (denorm_TA.x & 0x80u)));
-	#endif // PS_PROCESS_BA
-#else // PS_SHUFFLE_ACROSS
-	if ((denorm_c.g & 0x80u) != 0u)
-		C.g = float((denorm_c.g & 0x7Fu) | (denorm_TA.y & 0x80u));
-	else
-		C.g = float((denorm_c.g & 0x7Fu) | (denorm_TA.x & 0x80u));
-	if ((denorm_c.a & 0x80u) != 0u)
-		C.a = float((denorm_c.a & 0x7Fu) | (denorm_TA.y & 0x80u));
-	else
-		C.a = float((denorm_c.a & 0x7Fu) | (denorm_TA.x & 0x80u));
-#endif // PS_SHUFFLE_ACROSS
+			C.ga = vec2(float((denorm_c.g >> 6) | ((denorm_c.b >> 3) << 2) | (denorm_TA.x & 0x80u)));
+	#elif PS_SHUFFLE_ACROSS
+		#if(PS_PROCESS_BA == SHUFFLE_READWRITE && PS_PROCESS_RG == SHUFFLE_READWRITE)
+			C.rb = C.br;
+			float g_temp = C.g;
+			
+			C.g = C.a;
+			C.a = g_temp;
+		#elif(PS_PROCESS_BA & SHUFFLE_READ)
+			C.rb = C.bb;
+			C.ga = C.aa;
+		#else
+			C.rb = C.rr;
+			C.ga = C.gg;
+		#endif // PS_PROCESS_BA
+	#endif // PS_SHUFFLE_ACROSS
 #endif // PS_SHUFFLE
 
 	ps_dither(C.rgb, alpha_blend.a);
diff --git a/bin/resources/shaders/vulkan/tfx.glsl b/bin/resources/shaders/vulkan/tfx.glsl
index 5f9c8f6a4f..ab109e4a5e 100644
--- a/bin/resources/shaders/vulkan/tfx.glsl
+++ b/bin/resources/shaders/vulkan/tfx.glsl
@@ -953,7 +953,7 @@ vec4 ps_color()
 	vec4 T = sample_color(st);
 #endif
 
-	#if (SW_BLEND || PS_TFX != 1) && PS_SHUFFLE && !PS_SHUFFLE_SAME && !PS_READ16_SRC && (PS_SHUFFLE_ACROSS || PS_PROCESS_BA == SHUFFLE_READWRITE || PS_PROCESS_RG == SHUFFLE_READWRITE)
+	#if PS_SHUFFLE && !PS_READ16_SRC && !PS_SHUFFLE_SAME
 		uvec4 denorm_c_before = uvec4(T);
 		#if (PS_PROCESS_BA & SHUFFLE_READ)
 			T.r = float((denorm_c_before.b << 3) & 0xF8);
@@ -966,6 +966,8 @@ vec4 ps_color()
 			T.b = float((denorm_c_before.g << 1) & 0xF8);
 			T.a = float(denorm_c_before.g & 0x80);
 		#endif
+		
+		T.a = ((T.a >= 127.5f) ? TA.y : ((PS_AEM == 0 || any(bvec3(ivec3(T.rgb) & ivec3(0xF8)))) ? TA.x : 0.0f)) * 255.0f;
 	#endif
 	
 	vec4 C = tfx(T, vsIn.c);
@@ -1307,7 +1309,7 @@ void main()
 	ps_blend(C, alpha_blend);
 
 #if PS_SHUFFLE
-		#if (SW_BLEND || PS_TFX != 1) && !PS_SHUFFLE_SAME && !PS_READ16_SRC && (PS_SHUFFLE_ACROSS || PS_PROCESS_BA == SHUFFLE_READWRITE || PS_PROCESS_RG == SHUFFLE_READWRITE)
+		#if !PS_READ16_SRC && !PS_SHUFFLE_SAME
 			uvec4 denorm_c_after = uvec4(C);
 			#if (PS_PROCESS_BA & SHUFFLE_READ)
 				C.b = float(((denorm_c_after.r >> 3) & 0x1F) | ((denorm_c_after.g << 2) & 0xE0));
@@ -1318,59 +1320,37 @@ void main()
 			#endif
 		#endif
 
-		uvec4 denorm_c = uvec4(C);
-		uvec2 denorm_TA = uvec2(vec2(TA.xy) * 255.0f + 0.5f);
+		
 		
 		// Special case for 32bit input and 16bit output, shuffle used by The Godfather
 		#if PS_SHUFFLE_SAME
 			#if (PS_PROCESS_BA & SHUFFLE_READ)
+				uvec4 denorm_c = uvec4(C);
 				C = vec4(float((denorm_c.b & 0x7Fu) | (denorm_c.a & 0x80u)));
 			#else
 				C.ga = C.rg;
 			#endif
 		// Copy of a 16bit source in to this target
 		#elif PS_READ16_SRC
+			uvec4 denorm_c = uvec4(C);
+			uvec2 denorm_TA = uvec2(vec2(TA.xy) * 255.0f + 0.5f);
 			C.rb = vec2(float((denorm_c.r >> 3) | (((denorm_c.g >> 3) & 0x7u) << 5)));
-			if ((denorm_c.a & 0x80u) != 0u)
-				C.ga = vec2(float((denorm_c.g >> 6) | ((denorm_c.b >> 3) << 2) | (denorm_TA.y & 0x80u)));
-			else
-				C.ga = vec2(float((denorm_c.g >> 6) | ((denorm_c.b >> 3) << 2) | (denorm_TA.x & 0x80u)));
+			C.ga = vec2(float((denorm_c.g >> 6) | ((denorm_c.b >> 3) << 2) | (denorm_TA.x & 0x80u)));
 		// Write RB part. Mask will take care of the correct destination
 		#elif PS_SHUFFLE_ACROSS
 			#if(PS_PROCESS_BA == SHUFFLE_READWRITE && PS_PROCESS_RG == SHUFFLE_READWRITE)
 				C.rb = C.br;
-				if ((denorm_c.a & 0x80u) != 0u)
-					C.g = float((denorm_c.a & 0x7Fu) | (denorm_TA.y & 0x80u));
-				else
-					C.g = float((denorm_c.a & 0x7Fu) | (denorm_TA.x & 0x80u));
-					
-				if ((denorm_c.g & 0x80u) != 0u)
-					C.a = float((denorm_c.g & 0x7Fu) | (denorm_TA.y & 0x80u));
-				else
-					C.a = float((denorm_c.g & 0x7Fu) | (denorm_TA.x & 0x80u));
+				float g_temp = C.g;
 				
+				C.g = C.a;
+				C.a = g_temp;
 			#elif(PS_PROCESS_BA & SHUFFLE_READ)
 				C.rb = C.bb;
-				if ((denorm_c.a & 0x80u) != 0u)
-					C.ga = vec2(float((denorm_c.a & 0x7Fu) | (denorm_TA.y & 0x80u)));
-				else
-					C.ga = vec2(float((denorm_c.a & 0x7Fu) | (denorm_TA.x & 0x80u)));
+				C.ga = C.aa;
 			#else
 				C.rb = C.rr;
-				if ((denorm_c.g & 0x80u) != 0u)
-					C.ga = vec2(float((denorm_c.g & 0x7Fu) | (denorm_TA.y & 0x80u)));
-				else
-					C.ga = vec2(float((denorm_c.g & 0x7Fu) | (denorm_TA.x & 0x80u)));
+				C.ga = C.gg;
 			#endif // PS_PROCESS_BA
-		#else // PS_SHUFFLE_ACROSS
-			if ((denorm_c.g & 0x80u) != 0u)
-				C.g = float((denorm_c.g & 0x7Fu) | (denorm_TA.y & 0x80u));
-			else
-				C.g = float((denorm_c.g & 0x7Fu) | (denorm_TA.x & 0x80u));
-			if ((denorm_c.a & 0x80u) != 0u)
-				C.a = float((denorm_c.a & 0x7Fu) | (denorm_TA.y & 0x80u));
-			else
-				C.a = float((denorm_c.a & 0x7Fu) | (denorm_TA.x & 0x80u));
 		#endif // PS_SHUFFLE_ACROSS
 	#endif // PS_SHUFFLE
 
diff --git a/pcsx2/GS/Renderers/HW/GSRendererHW.cpp b/pcsx2/GS/Renderers/HW/GSRendererHW.cpp
index bac0d9fa03..f265afef7b 100644
--- a/pcsx2/GS/Renderers/HW/GSRendererHW.cpp
+++ b/pcsx2/GS/Renderers/HW/GSRendererHW.cpp
@@ -4743,6 +4743,7 @@ __ri void GSRendererHW::EmulateTextureSampler(const GSTextureCache::Target* rt,
 		// m_ps_sel.tex_fmt = 0; // removed as an optimization
 
 		//ASSERT(tex->m_target);
+		m_conf.ps.aem = TEXA.AEM;
 
 		// Require a float conversion if the texure is a depth otherwise uses Integral scaling
 		if (psm.depth)
@@ -4753,17 +4754,11 @@ __ri void GSRendererHW::EmulateTextureSampler(const GSTextureCache::Target* rt,
 		// Shuffle is a 16 bits format, so aem is always required
 		if (m_cached_ctx.TEX0.TCC)
 		{
-			m_conf.ps.aem = TEXA.AEM;
 			GSVector4 ta(TEXA & GSVector4i::x000000ff());
 			ta /= 255.0f;
 			m_conf.cb_ps.TA_MaxDepth_Af.x = ta.x;
 			m_conf.cb_ps.TA_MaxDepth_Af.y = ta.y;
 		}
-		else
-		{
-			m_conf.cb_ps.TA_MaxDepth_Af.x = 0;
-			m_conf.cb_ps.TA_MaxDepth_Af.y = 1.0f;
-		}
 
 		// The purpose of texture shuffle is to move color channel. Extra interpolation is likely a bad idea.
 		bilinear &= m_vt.IsLinear();
diff --git a/pcsx2/GS/Renderers/Metal/tfx.metal b/pcsx2/GS/Renderers/Metal/tfx.metal
index 344f4fe7d4..c7632c21bb 100644
--- a/pcsx2/GS/Renderers/Metal/tfx.metal
+++ b/pcsx2/GS/Renderers/Metal/tfx.metal
@@ -831,7 +831,7 @@ struct PSMain
 		else
 			T = sample_color(st);
 
-		if ((SW_BLEND || PS_TFX != 1) && PS_SHUFFLE && !PS_SHUFFLE_SAME && !PS_READ16_SRC && (PS_SHUFFLE_ACROSS || PS_PROCESS_BA == SHUFFLE_READWRITE || PS_PROCESS_RG == SHUFFLE_READWRITE))
+		if (PS_SHUFFLE && !PS_SHUFFLE_SAME && !PS_READ16_SRC)
 		{
 			uint4 denorm_c_before = uint4(T);
 			if (PS_PROCESS_BA & SHUFFLE_READ)
@@ -848,6 +848,8 @@ struct PSMain
 				T.b = float((denorm_c_before.g << 1) & 0xF8);
 				T.a = float(denorm_c_before.g & 0x80);
 			}
+			
+			T.a = (T.a >= 127.5 ? cb.ta.y : !PS_AEM || any((int3(T.rgb) & 0xF8) != 0) ? cb.ta.x : 0.f) * 255.f;
 		}
 	
 		float4 C = tfx(T, IIP ? in.c : in.fc);
@@ -1125,7 +1127,7 @@ struct PSMain
 
 		if (PS_SHUFFLE)
 		{
-			if ((SW_BLEND || PS_TFX != 1) && PS_SHUFFLE && !PS_SHUFFLE_SAME && !PS_READ16_SRC && (PS_SHUFFLE_ACROSS || PS_PROCESS_BA == SHUFFLE_READWRITE || PS_PROCESS_RG == SHUFFLE_READWRITE))
+			if (!PS_SHUFFLE_SAME && !PS_READ16_SRC)
 			{
 				uint4 denorm_c_after = uint4(C);
 				if (PS_PROCESS_BA & SHUFFLE_READ)
@@ -1140,12 +1142,11 @@ struct PSMain
 				}
 			}
 
-			uint4 denorm_c = uint4(C);
-			uint2 denorm_TA = uint2(cb.ta * 255.5f);
-
 			// Special case for 32bit input and 16bit output, shuffle used by The Godfather
 			if (PS_SHUFFLE_SAME)
 			{
+				uint4 denorm_c = uint4(C);
+				
 				if (PS_PROCESS_BA & SHUFFLE_READ)
 					C = (denorm_c.b & 0x7F) | (denorm_c.a & 0x80);
 				else
@@ -1154,6 +1155,9 @@ struct PSMain
 			// Copy of a 16bit source in to this target
 			else if (PS_READ16_SRC)
 			{
+				uint4 denorm_c = uint4(C);
+				uint2 denorm_TA = uint2(cb.ta * 255.5f);
+				
 				C.rb = (denorm_c.r >> 3) | (((denorm_c.g >> 3) & 0x7) << 5);
 				if (denorm_c.a & 0x80)
 					C.ga = (denorm_c.g >> 6) | ((denorm_c.b >> 3) << 2) | (denorm_TA.y & 0x80);
@@ -1165,44 +1169,22 @@ struct PSMain
 				if (PS_PROCESS_BA == SHUFFLE_READWRITE && PS_PROCESS_RG == SHUFFLE_READWRITE)
 				{
 					C.rb = C.br;
-					if ((denorm_c.a & 0x80) != 0)
-						C.g = (denorm_c.a & 0x7F) | (denorm_TA.y & 0x80);
-					else
-						C.g = (denorm_c.a & 0x7F) | (denorm_TA.x & 0x80);
-						
-					if ((denorm_c.g & 0x80) != 0)
-						C.a = (denorm_c.g & 0x7F) | (denorm_TA.y & 0x80);
-					else
-						C.a = (denorm_c.g & 0x7F) | (denorm_TA.x & 0x80);
+					float g_temp = C.g;
+					
+					C.g = C.a;
+					C.a = g_temp;
 				}
 				else if(PS_PROCESS_BA & SHUFFLE_READ)
 				{
 					C.rb = C.bb;
-					if ((denorm_c.a & 0x80) != 0)
-						C.ga =  (denorm_c.a & 0x7F) | (denorm_TA.y & 0x80);
-					else
-						C.ga =  (denorm_c.a & 0x7F) | (denorm_TA.x & 0x80);
+					C.ga = C.aa;
 				}
 				else
 				{
 					C.rb = C.rr;
-					if ((denorm_c.g & 0x80) != 0)
-						C.ga = (denorm_c.g & 0x7F) | (denorm_TA.y & 0x80);
-					else
-						C.ga = (denorm_c.g & 0x7F) | (denorm_TA.x & 0x80);
+					C.ga = C.gg;
 				}
 			}
-			else // Basically a direct copy but a shuffle of both pairs of channels, so green and alpha get modified by TEXA
-			{
-				if ((denorm_c.g & 0x80) != 0)
-					C.g = (denorm_c.g & 0x7F) | (denorm_TA.y & 0x80);
-				else
-					C.g = (denorm_c.g & 0x7F) | (denorm_TA.x & 0x80);
-				if ((denorm_c.a & 0x80) != 0)
-					C.a = (denorm_c.a & 0x7F) | (denorm_TA.y & 0x80);
-				else
-					C.a = (denorm_c.a & 0x7F) | (denorm_TA.x & 0x80);
-			}
 		}
 		
 		ps_dither(C, alpha_blend.a);
diff --git a/pcsx2/ShaderCacheVersion.h b/pcsx2/ShaderCacheVersion.h
index 6f482460b9..23060639d1 100644
--- a/pcsx2/ShaderCacheVersion.h
+++ b/pcsx2/ShaderCacheVersion.h
@@ -3,4 +3,4 @@
 
 /// Version number for GS and other shaders. Increment whenever any of the contents of the
 /// shaders change, to invalidate the cache.
-static constexpr u32 SHADER_CACHE_VERSION = 49;
+static constexpr u32 SHADER_CACHE_VERSION = 50;