I believe so: two horizontal layers of 9 images each over a sweep of about 180 degrees horizontally and 20 degrees vertically should be enough, surely?
Thanks for the pointer to the AP tutorial - I hadn’t seen this before, so I must follow it up.
I believe so: two horizontal layers of 9 images each over a sweep of about 180 degrees horizontally and 20 degrees vertically should be enough, surely?
Thanks for the pointer to the AP tutorial - I hadn’t seen this before, so I must follow it up.
Why not try and stitch part of the panorama first? Just let Hugin pick the control points (careful with sky/clouds) and align the images.
And if you used a 21mm FF equivalent focal lenght, you should have a FoV per image of over 80°. So you may actually have too much overlap: 4 images should cover the full 180° sweep…
If you have clouds, those may result in false control points being detected. See Hugin Preferences - PanoTools.org Wiki
Try the option Control Point Detector Programs = Hugin’s CPFind + Celeste
Hmm, good point. I’ll try that.
I think I have a bigger problem than this: having searched for a couple of hours I am unable to find any method for getting an editable file (such as a jpeg, tiff, png or even psd) out of Hugin. It’s obviously there but so well hidden that it means the human factors design is compromised such that I cannot justify spending any more time on it today without first trying a google search or trying to get the information out of the disjointed set of notes that constitute the Hugin manual, on panotools.
I don’t get this, sorry. If the interface is set to the simple or beginner (whatever it is called) mode, you have 3 buttons, which you have to press in order:
Yeah, I can understand your feeling of incredulity. I had been looking for some sort of ‘save this stitched output’ button, option, icon, file-menu entry or whatever - but there is nothing. What I didn’t realise is that the output I wanted is automatically written as a result of running the ‘create panorama’ step. I have just discovered this, this morning, entirely by accident, when doing a back-up. And this was after considerable more reading and experimentation. I make no further comment about usability here…
However, I still have not found a way to stitch images which are in a vertical alignment with other images which are in a horizontal alignment - as I typically have when I make two passes across an architectural scene when I cannot capture all the vertical field of view, even with a reasonable (21 mm) wide angle lens, in a single pass.