diff --git a/src/ImageSharp/Formats/Webp/Lossy/LossyUtils.cs b/src/ImageSharp/Formats/Webp/Lossy/LossyUtils.cs
index 7c262a30e..5b27af821 100644
--- a/src/ImageSharp/Formats/Webp/Lossy/LossyUtils.cs
+++ b/src/ImageSharp/Formats/Webp/Lossy/LossyUtils.cs
@@ -942,7 +942,7 @@ namespace SixLabors.ImageSharp.Formats.Webp.Lossy
             FilterLoop24(v, offsetPlus4, 1, stride, 8, thresh, ithresh, hevThresh);
         }
 
-        public static void Mean16x4(Span<byte> input, Span<uint> dc, Span<ushort> tmp)
+        public static void Mean16x4(Span<byte> input, Span<uint> dc)
         {
 #if SUPPORTS_RUNTIME_INTRINSICS
             if (Sse2.IsSupported)
@@ -966,13 +966,13 @@ namespace SixLabors.ImageSharp.Formats.Webp.Lossy
                 Vector128<int> e0 = Sse2.Add(d0, d1);
                 Vector128<int> e1 = Sse2.Add(d2, d3);
                 Vector128<int> f0 = Sse2.Add(e0, e1);
-                ref ushort outputRef = ref MemoryMarshal.GetReference(tmp);
-                Unsafe.As<ushort, Vector128<ushort>>(ref outputRef) = f0.AsUInt16();
+                Vector128<short> hadd = Ssse3.HorizontalAdd(f0.AsInt16(), f0.AsInt16());
+                Vector64<short> lower = hadd.GetLower();
 
-                dc[0] = (uint)(tmp[1] + tmp[0]);
-                dc[1] = (uint)(tmp[3] + tmp[2]);
-                dc[2] = (uint)(tmp[5] + tmp[4]);
-                dc[3] = (uint)(tmp[7] + tmp[6]);
+                dc[0] = (uint)lower.GetElement(0);
+                dc[1] = (uint)lower.GetElement(1);
+                dc[2] = (uint)lower.GetElement(2);
+                dc[3] = (uint)lower.GetElement(3);
             }
             else
 #endif
diff --git a/src/ImageSharp/Formats/Webp/Lossy/Vp8EncIterator.cs b/src/ImageSharp/Formats/Webp/Lossy/Vp8EncIterator.cs
index 57e18832e..6279aef65 100644
--- a/src/ImageSharp/Formats/Webp/Lossy/Vp8EncIterator.cs
+++ b/src/ImageSharp/Formats/Webp/Lossy/Vp8EncIterator.cs
@@ -363,7 +363,7 @@ namespace SixLabors.ImageSharp.Formats.Webp.Lossy
             uint m2;
             for (k = 0; k < 16; k += 4)
             {
-                LossyUtils.Mean16x4(this.YuvIn.AsSpan(YOffEnc + (k * WebpConstants.Bps)), dc.Slice(k, 4), tmp);
+                LossyUtils.Mean16x4(this.YuvIn.AsSpan(YOffEnc + (k * WebpConstants.Bps)), dc.Slice(k, 4));
             }
 
             for (m = 0, m2 = 0, k = 0; k < 16; k++)
diff --git a/tests/ImageSharp.Tests/Formats/WebP/LossyUtilsTests.cs b/tests/ImageSharp.Tests/Formats/WebP/LossyUtilsTests.cs
index 16b8e1166..09727293c 100644
--- a/tests/ImageSharp.Tests/Formats/WebP/LossyUtilsTests.cs
+++ b/tests/ImageSharp.Tests/Formats/WebP/LossyUtilsTests.cs
@@ -25,11 +25,10 @@ namespace SixLabors.ImageSharp.Tests.Formats.WebP
                 173, 175, 166, 155, 155, 159, 159, 158
             };
             uint[] dc = new uint[4];
-            ushort[] tmp = new ushort[8];
             uint[] expectedDc = { 1940, 2139, 2252, 1813 };
 
             // act
-            LossyUtils.Mean16x4(input, dc, tmp);
+            LossyUtils.Mean16x4(input, dc);
 
             // assert
             Assert.True(dc.SequenceEqual(expectedDc));
@@ -73,7 +72,7 @@ namespace SixLabors.ImageSharp.Tests.Formats.WebP
         public void Mean16x4_WithHardwareIntrinsics_Works() => FeatureTestRunner.RunWithHwIntrinsicsFeature(RunMean16x4Test, HwIntrinsics.AllowAll);
 
         [Fact]
-        public void Mean16x4_WithoutSSE2_Works() => FeatureTestRunner.RunWithHwIntrinsicsFeature(RunMean16x4Test, HwIntrinsics.DisableSSE2);
+        public void Mean16x4_WithoutHardwareIntrinsics_Works() => FeatureTestRunner.RunWithHwIntrinsicsFeature(RunMean16x4Test, HwIntrinsics.DisableHWIntrinsic);
 
         [Fact]
         public void HadamardTransform_WithHardwareIntrinsics_Works() => FeatureTestRunner.RunWithHwIntrinsicsFeature(RunHadamardTransformTest, HwIntrinsics.AllowAll);