I am trying to adjust the calibration, alignment and mismatch of stereo cameras. It works great with regular data samples. However, I am trying to use dual cameras on the iPhone 7+, which have a different scale. The telephoto lens has a 2x zoom compared to a wide-angle camera. I ran images through an algorithm, and it succeeds, albeit with a high error rate. However, when I open the corrected images, they have a strange spherical appearance to the edges. The center looks great. I assume that this is due to the fact that cameras have different zoom levels. Is there anything special I need to do to handle this? Or do I just need to trim any output to a useful undistorted area? Here is what I see:
EDIT:
I tried to use the calibration result from these checkerboard images to correct the image of some objects, and the correction was turned off, not even closed. If I fix one of my checkerboard images, they will be in place. Any ideas why this is happening?
EDIT2:
This is what my input images look like this is the result of a spherical looking output image. They are both taken from the same position, the iPhone was mounted on a tripod, and I used a Bluetooth device to start the shutter so that the image does not shake, my code automatically receives one image with each lens. I took 19 such images from different angles, all images show a complete chessboard. A larger image is an image that is fixed on the upper spherical image. 
EDIT3:
Here is a mismatch map using the calibration I received.
source share