diff --git a/src/ImageSharp/Processing/Processors/Convolution/Convolution2PassProcessor{TPixel}.cs b/src/ImageSharp/Processing/Processors/Convolution/Convolution2PassProcessor{TPixel}.cs
index 6d1d7c23c2..38b6cab3f8 100644
--- a/src/ImageSharp/Processing/Processors/Convolution/Convolution2PassProcessor{TPixel}.cs
+++ b/src/ImageSharp/Processing/Processors/Convolution/Convolution2PassProcessor{TPixel}.cs
@@ -123,15 +123,10 @@ internal class Convolution2PassProcessor<TPixel> : ImageProcessor<TPixel>
             this.Configuration,
             this.PreserveAlpha);
 
-        using (IMemoryOwner<Vector4> hBuffer = allocator.Allocate<Vector4>(horizontalOperation.GetRequiredBufferLength(interest)))
-        {
-            Span<Vector4> hSpan = hBuffer.Memory.Span;
-
-            for (int y = interest.Top; y < interest.Bottom; y++)
-            {
-                horizontalOperation.Invoke(y, hSpan);
-            }
-        }
+        ParallelRowIterator.IterateRows<HorizontalConvolutionRowOperation, Vector4>(
+                    this.Configuration,
+                    interest,
+                    in horizontalOperation);
 
         // Vertical convolution
         VerticalConvolutionRowOperation verticalOperation = new(
@@ -143,15 +138,10 @@ internal class Convolution2PassProcessor<TPixel> : ImageProcessor<TPixel>
             this.Configuration,
             this.PreserveAlpha);
 
-        using (IMemoryOwner<Vector4> vBuffer = allocator.Allocate<Vector4>(verticalOperation.GetRequiredBufferLength(interest)))
-        {
-            Span<Vector4> vSpan = vBuffer.Memory.Span;
-
-            for (int y = interest.Top; y < interest.Bottom; y++)
-            {
-                verticalOperation.Invoke(y, vSpan);
-            }
-        }
+        ParallelRowIterator.IterateRows<VerticalConvolutionRowOperation, Vector4>(
+            this.Configuration,
+            interest,
+            in verticalOperation);
     }
 
     /// <summary>
diff --git a/src/ImageSharp/Processing/Processors/Convolution/ConvolutionProcessor{TPixel}.cs b/src/ImageSharp/Processing/Processors/Convolution/ConvolutionProcessor{TPixel}.cs
index b704f556fa..69d72b3cc0 100644
--- a/src/ImageSharp/Processing/Processors/Convolution/ConvolutionProcessor{TPixel}.cs
+++ b/src/ImageSharp/Processing/Processors/Convolution/ConvolutionProcessor{TPixel}.cs
@@ -97,17 +97,10 @@ internal class ConvolutionProcessor<TPixel> : ImageProcessor<TPixel>
             map.BuildSamplingOffsetMap(this.KernelXY.Rows, this.KernelXY.Columns, interest, this.BorderWrapModeX, this.BorderWrapModeY);
 
             RowOperation operation = new(interest, targetPixels, source.PixelBuffer, map, this.KernelXY, this.Configuration, this.PreserveAlpha);
-
-            // Convolution is memory-bandwidth-bound with low arithmetic intensity.
-            // Parallelization degrades performance due to cache line contention from
-            // overlapping source row reads. See #3111.
-            using IMemoryOwner<Vector4> buffer = allocator.Allocate<Vector4>(operation.GetRequiredBufferLength(interest));
-            Span<Vector4> span = buffer.Memory.Span;
-
-            for (int y = interest.Top; y < interest.Bottom; y++)
-            {
-                operation.Invoke(y, span);
-            }
+            ParallelRowIterator.IterateRows<RowOperation, Vector4>(
+               this.Configuration,
+               interest,
+               in operation);
         }
 
         Buffer2D<TPixel>.SwapOrCopyContent(source.PixelBuffer, targetPixels);
diff --git a/src/ImageSharp/Processing/Processors/Convolution/EdgeDetectorCompassProcessor{TPixel}.cs b/src/ImageSharp/Processing/Processors/Convolution/EdgeDetectorCompassProcessor{TPixel}.cs
index f1edb75a20..eae7481661 100644
--- a/src/ImageSharp/Processing/Processors/Convolution/EdgeDetectorCompassProcessor{TPixel}.cs
+++ b/src/ImageSharp/Processing/Processors/Convolution/EdgeDetectorCompassProcessor{TPixel}.cs
@@ -83,15 +83,11 @@ internal class EdgeDetectorCompassProcessor<TPixel> : ImageProcessor<TPixel>
                 processor.Apply(pass);
             }
 
-            // Convolution is memory-bandwidth-bound with low arithmetic intensity.
-            // Parallelization degrades performance due to cache line contention from
-            // overlapping source row reads. See #3111.
             RowOperation operation = new(source.PixelBuffer, pass.PixelBuffer, interest);
-
-            for (int y = interest.Top; y < interest.Bottom; y++)
-            {
-                operation.Invoke(y);
-            }
+            ParallelRowIterator.IterateRows(
+                this.Configuration,
+                interest,
+                in operation);
         }
     }