Composite Captures

Facebooktwittergoogle_pluslinkedin

Today’s Question: When shooting multiple vertical and horizontal photos for detailed panoramas, is it better to tilt the camera for the rows, or to raise/lower the tripod head?

Tim’s Quick Answer: For a typical composite panorama where you are photographing from a reasonable distance, either approach would be fine (though tilting would generally be easier). For closeup work, physically moving the camera would be preferred over tilting (or even panning) the camera.

More Detail: The key issue here is parallax and other distortion issues, which can create significant challenges when assembling the multiple captures into a composite panoramic image.

In general I recommend capturing the frames for composite panoramas at a lens focal length of 100mm or greater if at all possible. The reason is primarily to minimize distortion and parallax issues in the captures, in order to help ensure the individual frames go together into the final composite as smoothly and accurately as possible.

When capturing images at a focal length of around 100mm or greater, distortion will be minimized provided there are not any key subjects particularly close to the lens. If there are elements of the scene that are close to the lens, most of the parallax issues can be minimized by ensuring that the camera is rotating around the nodal point of the lens.

When you will be photographing a scene relatively close to the lens, it is best to actually move the camera across the scene for each capture, and move the camera up and down as needed for each row of the capture. This will help to minimize distortion overall, but it is also important to be sure that you are overlapping more than you otherwise would for this type of scenario.

When capturing a “typical” composite panorama, it is generally adequate to overlap each frame by about 20%. When distortion is a concern due to parallax issues, a shorter than optimal lens focal length, a close distance to the subject, or other related issues, it is a good idea to increase that overlap to about 50% of the frame for each capture.