at my gigapixel test panorama i got with APG stitching errors, which i couldn understand ( one example: see first picture as a small extract of the picture). The global RMS in this picture was 2.93. The control points at the pairs of pitures were arranged over a large part and the RMS between this pictures was around 2.5, and no control points were worst then 5. I didn't know what to do.
For a test i exported the project as a PTGui file. Then i loaded the file to PTGui and let PTGui optimize it once. No control points were were shifted. Then i stitched it by PTGui (second picture). You can see PTGui didn't make this stitching errors. As far as i can see (it is not easy at a picture with more than 2 Gigapixel) i didn't found any stitching errors.
Btw i had the feeling that the picture from PTGui is more sharp especially at the overlap between pictures. Both pictures are an extract of the original PSB-file which comes direct out of the programs.
div lenses from 8 to 2000 mm div cameras div panoheads div comuters