How is a QTVR-panorama photographed?
The QTVR technique is based on overlapping photos linked together with a special software. The photo equipment is small and easy to carry. The only real demand is that there is a clear, stabile spot for the tripod. I prefer to only use the existing light on the location, partly because this provides natural shadows (the view is depicted as it is in reality) and partly because it removes the need for space demanding flash equipment. When the tripod is positioned and the camera is properly adjusted, the actual photographing can start. The camera is turned to an appropriate starting view and the first photo is taken. Then the camera is turned a few degrees to the side and the next photo is taken. Then you continue with turning and exposures until the whole lap has been captured and the camera has returned to it's initial position.

An ordinary QTVR Panorama require only one "lap" of photos around the tripod. A Cubic QTVR on the other hand, also require a second and a third lap where the camera is tilted upwards 45°, respectively downwards 45°. The Cubic QTVR is then completed with one photo straight up (sky/ceiling) and a final photo straight down (ground/floor). An ordinary QTVR Panorama normally consists of 12 to 20 overlapping photos (depending on the focal length of the camera's lens). A Cubic QTVR, which also is depicting the floor and the ceiling, may need up to 34 photos in order to fully cover the whole view.

When the photo session is completed, the photos are moved to a computer where the patience demanding process of linking and rendering begins.

Movement
In order for the software to be able to link the images together, it is important that the scenery hasn't changed/moved during the photo session. As an example, it would be difficult to capture a huge wave rolling to the shore or a town square filled with people moving around. Both the wave and the people would have had time to move themselves while the camera was turned for the next photo in the session. Of course it is possible to create a QTVR with a lot of people moving around the camera, but the result may include "halved" persons walking in or out of the picture.

The Parallax Problem
This may sound a bit complicated but it is a really vital detail in QTVR photography. The camera needs to be moved backwards on the tripod. The front of the lens should be placed almost straight above the tripod's center point. The parallax problem is easiest spotted when details closer to the camera change their alignment with the background while the camera is turned sideways.
An example: Imagine you are in a park and you have aligned your camera with a line of lampposts. See the illustration below. 1. Here the camera is shown from above where the red dot is indicating the tripod's center point. The four black dots show the four lampposts also seen from above. On the far right you see what the camera's viewer would show in the respective case. The four lampposts are placed in a perfect row so the first one does more or less completely hide the other ones. 2. When you turn the camera counter clockwise, the motif is pushed to the right in the viewer. But then the row of lampposts are no longer seen from straight ahead, but rather slightly from the side. That means the closest lamp does not longer completely hide the other ones. 3. The same thing happens when you turn the camera clockwise. The previously hidden lampposts becomes visible right next to the one closest to the camera. So, the motif is changing it's appearance depending on if it is placed in the middle of the viewer, or closer to the sides. This will create grave problems later when the pictures are linked together and the details in the motif doesn't match.

4. This parallax problem can be completely eliminated if the camera is moved backwards on the tripod. Preferably with some kind of a camera sleigh. Then when the camera is turned counter clockwise, the row of lampposts will be pushed to the right in the viewer but they will remain aligned with the camera's lens. The closest lamppost will still hide the other ones. The motif is left intact. 5. The same thing happens when the camera is turned to the other side. The row of lamps is unchanged no matter where it is placed in the viewer.
Picture 4 and 5 would match perfectly if they were to be used in a QTVR production. Picture 2 and 3 on the other hand, would leave you with semi transparent shadows on either side of the closest lamppost.
It is nearly impossible to produce a completely perfect QTVR, but if you eliminate the parallax problem then you are well on your way.
How far back you should move the camera is depending on what lens you are using. Like with so many other things, you need to experiment a bit to find the optimal distance.

Return to the QTVR Main Page