diff --git a/src/lerobot/cameras/opencv/camera_opencv.py b/src/lerobot/cameras/opencv/camera_opencv.py
index 50e55f0c22..9f860b1827 100644
--- a/src/lerobot/cameras/opencv/camera_opencv.py
+++ b/src/lerobot/cameras/opencv/camera_opencv.py
@@ -233,15 +233,25 @@ def _validate_width_and_height(self) -> None:
 
         actual_width = int(round(self.videocapture.get(cv2.CAP_PROP_FRAME_WIDTH)))
         if not width_success or self.capture_width != actual_width:
-            raise RuntimeError(
-                f"{self} failed to set capture_width={self.capture_width} ({actual_width=}, {width_success=})."
+            logger.warning(
+                f"{self} failed to set capture_width={self.capture_width} ({actual_width=}, {width_success=}). "
+                f"Using camera's actual width={actual_width}."
             )
+            self.capture_width = actual_width
 
         actual_height = int(round(self.videocapture.get(cv2.CAP_PROP_FRAME_HEIGHT)))
         if not height_success or self.capture_height != actual_height:
-            raise RuntimeError(
-                f"{self} failed to set capture_height={self.capture_height} ({actual_height=}, {height_success=})."
+            logger.warning(
+                f"{self} failed to set capture_height={self.capture_height} ({actual_height=}, {height_success=}). "
+                f"Using camera's actual height={actual_height}."
             )
+            self.capture_height = actual_height
+
+        # Update output dimensions based on actual capture dimensions and rotation
+        if self.rotation in [cv2.ROTATE_90_CLOCKWISE, cv2.ROTATE_90_COUNTERCLOCKWISE]:
+            self.width, self.height = self.capture_height, self.capture_width
+        else:
+            self.width, self.height = self.capture_width, self.capture_height
 
     @staticmethod
     def find_cameras() -> list[dict[str, Any]]:
diff --git a/src/lerobot/cameras/utils.py b/src/lerobot/cameras/utils.py
index aa6ff98b48..48cfec638d 100644
--- a/src/lerobot/cameras/utils.py
+++ b/src/lerobot/cameras/utils.py
@@ -69,7 +69,7 @@ def get_cv2_backend() -> int:
     import cv2
 
     if platform.system() == "Windows":
-        return cv2.CAP_MSMF  # Use MSMF for Windows instead of AVFOUNDATION
+        return cv2.CAP_DSHOW  # Use DirectShow for Windows (more compatible than MSMF)
     # elif platform.system() == "Darwin":  # macOS
     #     return cv2.CAP_AVFOUNDATION
     else:  # Linux and others
diff --git a/src/lerobot/robots/so101_follower/so101_follower.py b/src/lerobot/robots/so101_follower/so101_follower.py
index acfd4bd114..a1395db7d9 100644
--- a/src/lerobot/robots/so101_follower/so101_follower.py
+++ b/src/lerobot/robots/so101_follower/so101_follower.py
@@ -176,7 +176,7 @@ def get_observation(self) -> dict[str, Any]:
 
         # Read arm position
         start = time.perf_counter()
-        obs_dict = self.bus.sync_read("Present_Position")
+        obs_dict = self.bus.sync_read("Present_Position", num_retry=5)
         obs_dict = {f"{motor}.pos": val for motor, val in obs_dict.items()}
         dt_ms = (time.perf_counter() - start) * 1e3
         logger.debug(f"{self} read state: {dt_ms:.1f}ms")
diff --git a/src/lerobot/scripts/lerobot_record.py b/src/lerobot/scripts/lerobot_record.py
index 55846ff63a..dd30655463 100644
--- a/src/lerobot/scripts/lerobot_record.py
+++ b/src/lerobot/scripts/lerobot_record.py
@@ -381,6 +381,19 @@ def record(cfg: RecordConfig) -> LeRobotDataset:
     robot = make_robot_from_config(cfg.robot)
     teleop = make_teleoperator_from_config(cfg.teleop) if cfg.teleop is not None else None
 
+    # Connect robot early to get actual camera dimensions
+    robot.connect()
+    if teleop is not None:
+        teleop.connect()
+
+    # Update config with actual camera dimensions after connection
+    if hasattr(robot, "cameras") and robot.cameras:
+        for cam_name, cam in robot.cameras.items():
+            if cam_name in cfg.robot.cameras:
+                # Update config with actual camera dimensions for dataset creation
+                cfg.robot.cameras[cam_name].width = cam.width
+                cfg.robot.cameras[cam_name].height = cam.height
+
     teleop_action_processor, robot_action_processor, robot_observation_processor = make_default_processors()
 
     dataset_features = combine_feature_dicts(
@@ -441,10 +454,6 @@ def record(cfg: RecordConfig) -> LeRobotDataset:
             },
         )
 
-    robot.connect()
-    if teleop is not None:
-        teleop.connect()
-
     listener, events = init_keyboard_listener()
 
     with VideoEncodingManager(dataset):
diff --git a/src/lerobot/teleoperators/so101_leader/so101_leader.py b/src/lerobot/teleoperators/so101_leader/so101_leader.py
index be804bf702..2180aca667 100644
--- a/src/lerobot/teleoperators/so101_leader/so101_leader.py
+++ b/src/lerobot/teleoperators/so101_leader/so101_leader.py
@@ -138,7 +138,7 @@ def setup_motors(self) -> None:
 
     def get_action(self) -> dict[str, float]:
         start = time.perf_counter()
-        action = self.bus.sync_read("Present_Position")
+        action = self.bus.sync_read("Present_Position", num_retry=5)
         action = {f"{motor}.pos": val for motor, val in action.items()}
         dt_ms = (time.perf_counter() - start) * 1e3
         logger.debug(f"{self} read action: {dt_ms:.1f}ms")