Raw or JPEG?
When shooting the spherical panorama different parts of the future image can be illuminated differently and, correspondingly, strongly differ from each other in light and color. Therefore, to ensure the natural color reproduction, the original frames, before being stitched together, are being strongly processed. It is the well-known fact, that the images in the JPEG format in comparison with those in the RAW format are very limited. That’s why it’s preferable to take the original frames for the 360 panorama production in the RAW format. The main advantage of the filming in RAW format is the possibility of the conversion parameters adjustment on computer in the RAW converter whereas when shooting in JPEG format all the parameters must be set on the camera. Because of that, unlike RAW files, in JPEG files one can’t change the white balance, exposure or reconstruct the information in the overexposed areas. Besides, the artifacts arising from the JPEG compression can increase during the process of spherical panorama panorama stitching.
When filming the virtual 360 panoramas everything should be in focus – both the background and the objects on the forefront (which include the floor underfoot as well). In order to achieve the maximum depth of the sharply represented space – from ~1,5m minimum and to infinity, it is recommended to direct the focus, taking the hyperfocal distance into an account. To determine the hyperfocal distance of your lens, you can use one of the online calculators of the maximum depth of the sharply represented space, for example, this one: - http://www.dofmaster.com/doftable.html
The focusing must be done in the manual mode. The usage of the automatic focusing when filming the 3D panorama may cause the separate frames to have a different focus, or some of the frames will be out of focus, which is even worse, because then it will be imprssible to band the panorama together.
The exposure must be adjusted in the manual mode, too; all the frames of the future spherical panorama must be taken with the same stutter speed and aperture. Possible small differences in the object’s illumination can be compensated by the change of exposure in RAW converter (within +/- 1 EV) or by the exposure correction of the separate frames when shooting. If the contrast level of the scene was too high, film the frames with the exposure bracketing and in the process of post processing combine them so that the resulting image contains all the details of all the source images, as the extreme shadows and maximum lights (using the HDR techniques or its analogs).
The initial frames shooting for the 360 spherical panorama production.
As it was already mentioned in the article about the lenses, the amount of frames, necessary for the full spherical panorama production depends mainly on the focal distance of the lens you’re using. The single-row spherical panoramas can be shot only with the Fisheye type of lens, and when using the other types of lenses, you’d have to take 2 or more rows of photographs. The important factor to estimate the number of photos is a degree of overlap of the adjacent frames. Usually it’s enough to shoot with 20-30% overlapping, but the percent of it must be increased in the following cases:
- When the difference in the exposure of adjacent frames is present – drop of the brightness can be aligned with a more natural result.
- When shooting surfaces with insufficient detail (for example, the bare walls) – the area needed for the placement of control points for panorama stitching will increase.
- When having the moving objects in the picture (people, cars) – for easy retouching of the stitched panorama, in which the certain objects can be reflected multiple times or only partially (for example, the half of a man).
The table below shows the approximate number of photographs needed for the spherical panorama production with different focal distances.
|Focal lenght||Sensor format||Number of photographs*|
The number of digits indicates the number of rows, the value of digits indicates the number of pictures in each row.
Z – the zenith frame (top)
N – the nadir frame (bottom)
It’s not crucial to follow these table’s schemes, they are not the only applicable ones. Depending on what will be the inclination angle (angles) during the process of shooting, the different schemes may be compiled.
Nadir frame shooting
The separate nadir frame (of that very place, where the tripod was standing) is used when stitching the panorama together or postprocessing for the shooting equipment retouching, which remained visible after the panorama stitching. Sometimes if the nadir space has a homogeneous structure without relief, the equipment can be removed with the standard retouching tools and not to shoot the nadir frame, however, it never hurts to make an extra shoot.
There are a lot of methods of nadir frame shooting. The most simple and fast way is to shoot a frame while holding the camera in your hands. Using this technique, you can remove the photo camera from the tripod (1.) or leave it there, (2.) holding it at arm’s length. It’s very important that the nodal point of the lens was in the same position (or as close as possible to it), in which it was when the camera was mounted on the tripod. If we shoot at the slow shutter speed or if the maximum positioning accuracy is required, the camera must be somehow fixed and often the shooting is being made from the “crane” (3.) or tilted stand (4.).
I’ll add, that in case, when our footprints can be seen at the place of the shooting (on the snow, sand, lawn, etc.), the nadir frame should be shot at the first place and after that frame is done, we shoot the other ones.
Shooting the nadir
This is how the original frames will look like obtained at different focal lengths:
Full frame Fisheye: 6+Z+N
18mm: 8+8+Z+N, both rows were shot with a slope of +/- 30° from the horizon
28mm: 12+12+12+Z+N, the rows were shot with a slope of +45°, 0°, - 45° from the horizon