diff --git a/api/onnx_web/diffusers/run.py b/api/onnx_web/diffusers/run.py
index 1405456e..f69421e5 100644
--- a/api/onnx_web/diffusers/run.py
+++ b/api/onnx_web/diffusers/run.py
@@ -285,7 +285,11 @@ def run_img2img_pipeline(
             **pipe_params,
         )
 
-    for image, output in zip(result.images, outputs):
+    images = result.images
+    if source_filter is not None:
+        images.append(source)
+
+    for image, output in zip(images, outputs):
         image = run_upscale_correction(
             job,
             server,
diff --git a/api/onnx_web/image/source_filter.py b/api/onnx_web/image/source_filter.py
index 56f02df4..3885bfdf 100644
--- a/api/onnx_web/image/source_filter.py
+++ b/api/onnx_web/image/source_filter.py
@@ -183,6 +183,7 @@ def source_filter_canny(
 
     image = cv2.Canny(pil_to_cv2(source), low_threshold, high_threshold)
     image = Image.fromarray(image)
+    image = image.convert("RGB")
 
     return image
 
@@ -192,7 +193,7 @@ def source_filter_openpose(server: ServerContext, source: Image.Image) -> Image.
 
     model = OpenposeDetector.from_pretrained(
         "lllyasviel/ControlNet",
-        cache_dir=server.cache_dir,
+        cache_dir=server.cache_path,
     )
     image = model(source)
 
diff --git a/api/onnx_web/output.py b/api/onnx_web/output.py
index 5a9a7ed8..892ffd6a 100644
--- a/api/onnx_web/output.py
+++ b/api/onnx_web/output.py
@@ -69,7 +69,9 @@ def make_output_name(
     params: ImageParams,
     size: Size,
     extras: Optional[List[Optional[Param]]] = None,
+    count: Optional[int] = None,
 ) -> List[str]:
+    count = count or params.batch
     now = int(time())
     sha = sha256()
 
@@ -93,7 +95,7 @@ def make_output_name(
 
     return [
         f"{mode}_{params.seed}_{sha.hexdigest()}_{now}_{i}.{server.image_format}"
-        for i in range(params.batch)
+        for i in range(count)
     ]
 
 
diff --git a/api/onnx_web/server/api.py b/api/onnx_web/server/api.py
index 29585582..3d5247af 100644
--- a/api/onnx_web/server/api.py
+++ b/api/onnx_web/server/api.py
@@ -176,8 +176,12 @@ def img2img(server: ServerContext, pool: DevicePoolExecutor):
         get_config_value("strength", "min"),
     )
 
-    # TODO: add filtered source to outputs
-    output = make_output_name(server, "img2img", params, size, extras=[strength])
+    output_count = params.batch
+    if source_filter is not None:
+        output_count += 1
+
+    output = make_output_name(server, "img2img", params, size, extras=[strength], count=output_count)
+
     job_name = output[0]
     logger.info("img2img job queued for: %s", job_name)