-
[quote user="bryanr"]The S-shape happens because your panorama is projected into the inside of a virtual sphere. If it is not projected "level," then when it is unrolled onto the screen, it will have an S-shape because some parts of the pano are higher up the side of the sphere than others. Hope this description helps.[/quote] This is the direction I am also thinking. The S-shape has to do with the wide field of view (about 240 degrees). If I rotate the picture one way or the
-
You are right. I used both straight arrows, which appear when the cursor is in the frame not the curved one outside. And as you predicted, the latter one allows to correct the S-shape. You will find the result on the same site as the preceding ones ( http://rhadorn.zenfolio.com/p664872946/h23371cb0#h23371cb0 ). Now, the problem is that the horizon is not horizontal. I am missing a rotation function. Or did I just overlook it? I have prepared a set of the original pictures in reduced format (90% quality
-
One of my panos shows a horizontal S shape. Actually, the two panos taken on the same spot with the same subject show the same distortion. It shows in both horizontal sphere and cylinder projections. I understand in some way a bow-like distortion, which can be corrected; the S-shape is tricky because one curve is accentuated when trying to correct the othe one. The pictures were taken with handheld camera. Reductions can be found on following address: http://rhadorn.zenfolio.com/p664872946 How to
-
I am surprised this was a bug, even if it is very encouraging to hear of corrected bugs solving problems! My hypothesis was that the difficulty to stitch had to do with the mismatch between front and back scene when you superpose pictures 3 and 4. I myself shoot panoramas without the aid of a dedicated pano head, so I get similar mismatches between adjascent pictures, when the front scene is too close to the camera. And I actually got into trouble as a started stitching such panoramas with Autopano