diff --git a/nerfstudio/cameras/cameras.py b/nerfstudio/cameras/cameras.py
index 67e1ac4b7..5644e2723 100644
--- a/nerfstudio/cameras/cameras.py
+++ b/nerfstudio/cameras/cameras.py
@@ -244,7 +244,7 @@ class Cameras(TensorDataclass):
             if times.ndim == 0 or times.shape[-1] != 1:
                 times = times.unsqueeze(-1).to(self.device)
         else:
-            raise ValueError(f"times must be a tensor, got {type(times)}")
+            raise ValueError(f"times must be None or a tensor, got {type(times)}")
 
         return times
 
diff --git a/nerfstudio/viewer/server/viewer_utils.py b/nerfstudio/viewer/server/viewer_utils.py
index 12b138b5d..25a90579a 100644
--- a/nerfstudio/viewer/server/viewer_utils.py
+++ b/nerfstudio/viewer/server/viewer_utils.py
@@ -325,7 +325,7 @@ class ViewerState:
         # set the initial state whether to train or not
         self.vis["renderingState/isTraining"].write(start_train)
 
-        self.vis["renderingState/render_time"].write(str(0))
+        # self.vis["renderingState/render_time"].write(str(0))
 
         # set the properties of the camera
         # self.vis["renderingState/camera"].write(json_)
@@ -756,8 +756,9 @@ class ViewerState:
             dim=0,
         )
 
-        times = float(self.vis["renderingState/render_time"].read())
-        times = torch.tensor([times])
+        times = self.vis["renderingState/render_time"].read()
+        if times:
+            times = torch.tensor([float(times)])
 
         camera = Cameras(
             fx=intrinsics_matrix[0, 0],