My Autopano Pro forum

Another way to do this is to set the enter point on an object at tabout the same height as the camera, then level the pano.  You may then have to rotate it 90 degrees.

Right now, the nadir is so close to the centerpoint that rotating it 90 degrees just rotates it around a point near the tripod.

Last edited by hankkarl (2008-07-20 20:02:22)

thanks to you all!

it worked this way:
1. setting the center point to the nadir (tripod)
2. setting the center point to the top of all...
3. optimising the horizon

i just wondered, why this happened... i had this issue never before...
next week i will try it with a strange combination: d300/peleng... (just waiting for the adapter ring...)

GURL wrote:

Machart  wrote:

I just wondered, why this happened... I had this issue never before...

In any camera there is a special device used to record the camera orientation during exposure. It works fine when the camera is not oriented toward the zenith or the nadir (or about.) When camera pitch value is next to 90°, positive or negative, the recorded orientation is often (but not always) wrong and APP is confused, lost and disorientate...

Nikon D300 have a special feature known as "Virtual horizon" and this implies it can use a second hardware device to detect roll variations and display them: the perfect camera for panorama would record not only yaw (difficult, GPS needed) but pitch and roll. Nikon made a little move in the right direction: roll is fine but pitch would help, too !

A GPS receiver only records location, so it wouldn't help with yaw.  You would need a compass.

A GPS needs to be moving to tell direction.  Even walking is fast enough. 

Car GPSes may have a magnetic compas built-in, or may have gyroscopes (the solid state kind probably) or accellerometers to tell when and how far they move.

machart  wrote:

its a (compromising) 18-200mm nikon. the fisheye-like distortion with 18mm (27mm) doesn't fit with the fact, that its no fisheye...

The stitched image resulting from a rather large number of 28mm equivalent lens source images is identical to the stitched image resulting from a smaller number of fisheye source images. To stitch them properly APP must know whether a standard lens or a fisheye was used.

Though I often use a fisheye (8mm Olympus for 4/3 cameras) I recently made some attempts at using the 14mm setting (28mm equivalent) of my standard zoom and a NN3 panohead. The resulting image is much larger (for example 16000 x 8000 pix rather than 8000 x 4000 when using a fisheye) so that more details are visible. The corresponding disadvantage is that more shots are needed (I found 30 images being easy stitch using APP, this corresponds to a 45% global overlap.)

Using a motorized panohead like the modified Merlin/Orion should make this way of producing 360 x 180 panos adequate for quiet subjects...