diff --git a/invokeai/app/invocations/latent.py b/invokeai/app/invocations/latent.py
index f0ad850bb3..35e05934ff 100644
--- a/invokeai/app/invocations/latent.py
+++ b/invokeai/app/invocations/latent.py
@@ -366,6 +366,10 @@ class TextToLatentsInvocation(BaseInvocation):
             )
         )
 
+        latents_shape = noise.shape
+        control_height_resize = latents_shape[2] * 8
+        control_width_resize = latents_shape[3] * 8
+
         # copied from old backend/txt2img.py
         # FIXME: still need to test with different widths, heights, devices, dtypes
         #        and add in batch_size, num_images_per_prompt?
@@ -375,10 +379,8 @@ class TextToLatentsInvocation(BaseInvocation):
                     image=control_image,
                     # do_classifier_free_guidance=do_classifier_free_guidance,
                     do_classifier_free_guidance=True,
-                    # width=width,
-                    # height=height,
-                    width=512,
-                    height=512,
+                    width=control_width_resize,
+                    height=control_height_resize,
                     # batch_size=batch_size * num_images_per_prompt,
                     # num_images_per_prompt=num_images_per_prompt,
                     device=control_model.device,
@@ -391,10 +393,8 @@ class TextToLatentsInvocation(BaseInvocation):
                         image=image_,
                         # do_classifier_free_guidance=do_classifier_free_guidance,
                         do_classifier_free_guidance=True,
-                        # width=width,
-                        # height=height,
-                        width=512,
-                        height=512,
+                        width=control_width_resize,
+                        height=control_height_resize,
                         # batch_size=batch_size * num_images_per_prompt,
                         # num_images_per_prompt=num_images_per_prompt,
                         device=control_model.device,
@@ -403,8 +403,6 @@ class TextToLatentsInvocation(BaseInvocation):
                     images.append(image_)
                 control_image = images
 
-
-
         # TODO: Verify the noise is the right size
         result_latents, result_attention_map_saver = model.latents_from_embeddings(
             latents=torch.zeros_like(noise, dtype=torch_dtype(model.device)),
diff --git a/invokeai/backend/stable_diffusion/diffusers_pipeline.py b/invokeai/backend/stable_diffusion/diffusers_pipeline.py
index 1fe20e6361..9656fe7eee 100644
--- a/invokeai/backend/stable_diffusion/diffusers_pipeline.py
+++ b/invokeai/backend/stable_diffusion/diffusers_pipeline.py
@@ -1030,6 +1030,7 @@ class StableDiffusionGeneratorPipeline(StableDiffusionPipeline):
         dtype=torch.float16,
         do_classifier_free_guidance=True,
     ):
+
         if not isinstance(image, torch.Tensor):
             if isinstance(image, PIL.Image.Image):
                 image = [image]