Methods, apparatus, and computer-readable media are provided for processing image data captured by a first device for display on a second device. For example, a range of predicted orientations of the second device can be determined. A predicted field-of-view of the second device can then be determined. The predicted field-of-view corresponds to the range of predicted orientations of the second device. The predicted field-of-view can be transmitted to the first device. Cropped image data may then be received from the first device, which includes image data cropped to include the predicted field-of-view. An updated orientation of the second device can be determined, and an updated field-of-view within the cropped image data can be determined that corresponds to the updated orientation of the second device.