I have just started doing some stitching, and plan to do more panoramic landscape photography with an Alpa camera. My questions is this; if I am doing a three shot panorama, and also bracketing the exposures, for a total of nine images, what will happen to clouds in the photos when I merge it all together?
There must be some mechanism in the software that adjusts for this to some degree in Photoshop or other programs. I am thinking it could take up to one minute or more to take all the pictures, during which the clouds could move quite a bit. Anyone have any pointers in this regard?
Hi,
A bit of planning to get the area involving the clouds done without delays helps. Then additionally, the stitching software should be capable of intelligently blending the overlapping areas. Software like PTGUI Pro also offers a manual masking capability, and allows to use alternative blending engines, e.g. Smartblend.
As for the planning part, if you shoot multi-row panos, try to shoot the clouds row as a single row, that reduces the time required and it reduces the number of overlapping seams. Shooting the camera in portrait orientation helps with the vertical coverage, and using a proper overlap, say at least 30% can also help. If there is a lot of wind with distinct clouds, it may also help to shoot the sequence in the opposite direction of the cloud travel direction, to avoid having the same clouds turning up in multiple frames. Using a pano rig with a rotation through the entrance pupil of the lens should allow to take images in a short time, especially if you use a click-stop indexing rotator, thus also reducing the risk of visible ghosting.
As for the software, dedicated pano stitcher applications usually have good blending engines that can manage to avoid clearly visible issues, and offer some masking capability for manual intervention in case it's needed.
I am pondering this because I am debating which wide angle lens to obtain, and the wider the angle lens, the fewer shots I would have to merge together; two images versus three images.
If you take single row panos at daytime, it shouldn't take a minute to record but rather a certain number of seconds depending on the number of images, even if HDR bracketing is involved. I would let the scene and the required output filesize dictate the pano requirements (single/multi-row and focal length). But if you only take a single row (which simplifies the pano-gear requirements a lot), then the required vertical Field of View coverage is leading, and using portrait orientation will help. On full-frame 35mm sensors, a 24mm will give you a lot of coverage (73.7 degrees vertical in portrait orientation). So depending on the sensor size on the Alpa, you could aim for something with a similar angle of view.
Cheers,
Bart