Kay: On Thursday, December 29, 2011 4:21:52 AM UTC-7, kfj wrote:
On 29 Dez., 01:48, Calvin McDonald <[email protected]> wrote: > > > Result was good other than it didn't fix the misalignment of the stacked > > images. I'm getting 20 or so pixels of edge ghosting on the sky > boundary > > still. I experimented with many CPs on the sky boundary - didn't help. > I > > experimented with "Exposure fused from any arrangement" - changed things > a > > little but didn't improve the result. > > > > Ideas? > > If your CPs don't have an effect on the alignment, are you sure to > have run the optimizer after placing the CPs? CPs per se don't have > any effect whatsoever. > > Why do you bother with handmade control points? Just run cpfind on > your image set, this should give you much better alignment than just > using four handmade points. > Sorry, my description was poor. I did use cpfind, I meant that afterwards I placed a bunch of CPs on the sky boundary by hand to see if that would help. I use optimization also - actually quite a few times, as I tune the cpfind CPs. There is something very odd about this particular image set. In Hugin I completely started over from scratch and followed all your guidance, checked all the cpfind generated CPs (no manual ones this time) very closely for misplaced ones. After several optimization runs I could only get the error down to 5.04 with the max error being 22.4 - I've never had a pano converge to such a poor fit. Most of the CPs at the top of the error list are between the pairs of stacked images. Final result - no change - still bad. >From my inexperienced observation - it simply appears that the tools are unable to align the top portions of the top row of 12 stacked images. Everything looks fine from about the middle (going up/down) of this row of images and down. But from the middle going up they progressively get more what looks like misaligned until the sky boundary where it looks the worst. It's likely the worst at the top in the sky but the errors all get lost in the sky/clouds. Now this has been going on for quite a while and I feel I might help > you more effectively if I see > > a) your images > b) the pto you use for them > > do you have a place where you can put them online so I can have a go > at them (doesn't have to be the full set, and can be JPEG to save > space)? > Given I'm really just trying to find a method to fix a mistake I made upstream (didn't take the stacked images back-to-back) I'm starting to feel like I'm wasting both our times. I've fought this sufficiently to force me never to forget to shoot my exposure bracketing shots back-to-back. :) I'll forgo troubling you with handing off a test-case to you, unless you want it. I do have a place I could drop it online. One last question. Back on the thought that something very unusual is wrong.... What would happen if the focal length of a zoom lens were to change just a bit between a pair of otherwise identical exposure-stacked images? I ask because I did shoot these images with a zoom lens, a Nikkor 18-200mm VR-II. I shoot all my panos with this lens, with it zoom-locked at 18mm. It was definitely locked at 18mm when I took the images in question but I've noticed that even when it's locked the zoom barrel can still move a little. This is probably a long-shot, but what if the focal length got bumped slightly between the two sets of stacked images. Could that account for the trouble I'm having? I'm having a hard time believing this could because I've never had this problem before in the 100+ other panos I've done with this same lens. Thanks Calvin -- You received this message because you are subscribed to the Google Groups "Hugin and other free panoramic software" group. A list of frequently asked questions is available at: http://wiki.panotools.org/Hugin_FAQ To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/hugin-ptx
