I had a look at the files you uploaded. Here's a PTO that has them all aligned.
I added the images into a fresh Hugin project. Did not assign stacks. I ran Hugin's CPFind on each exposure set (same position, different exposures). I then ran CPFind on the middle exposures. Optimized position, got rid of outliers, optimized again. You can now export each exposure plane (same exposures, different positions) separately and enfuse. I had the same trouble with align_image_stack with DSC_8625.tif. I thought it was because the images aren't very sharp but even with downsized images align_image_stack was failing to find suitable control points. It may be as you suspect, it's too translated with respect to the other two images. PS You may want to capture a greater range of exposures, not just +/-1EV. On Sunday, 6 August 2017 18:58:12 UTC+1, mh00h wrote: > > I'm still having problems getting this to stitch. > > The main suggestion that I hear from dkloi is to run align_image_stack at > the beginning of the workflow regardless of whether I plan to do > exposure fusion from stacks OR exposure fusion from any arrangement. Does > this not defeat the purpose of exposure fusion from any arrangement? Here > is what I mean: > > Given: Images were shot like this. > > > > <https://lh3.googleusercontent.com/-jD7-LKCWCMk/WYdOQ3vpZ-I/AAAAAAAANWQ/Kccs-zLxBXw4bfcnAUKzXeW2YE9jazfTwCL4CGAYYCw/s1600/unnamed0.jpg> > > > > What is being proposed: make stacks, even though images were not taken as > stacks per flow diagram above. Like this, repeated for all the "stacks": > > > > <https://lh3.googleusercontent.com/-Px_y8ZExwaA/WYdQ7zYzArI/AAAAAAAANWc/kQsL9CeepbQfuV16WLfXFaZYiVP2XK6RQCL4CGAYYCw/s1600/unnamed1.jpg> > > > Then I have two options: > > a) exposure fusion from any arrangement (as if moving from ais images, > DOWN) > b) exposure fused from stacks (as if moving from ais images to RIGHT) > > > > > > <https://lh3.googleusercontent.com/-P-0rtmCwLMY/WYdTYzSlJnI/AAAAAAAANWs/WwtldDzQaKgSvfHID-RHiSeg8eBGq6yBQCL4CGAYYCw/s1600/unnamed2.jpg> > > > The problem with this approach is that align_image_stack cannot identify > control points for the image DSC_8625.tif which match with the rest of the > "stack" (DSC_8632.tif, DSC_8633.tif, DSC_8625.tif). I had originally > intended this fusion to happen via any arrangement. I didn't think that > individual images had to be ligned up. > > > I've tried playing with the parameters of align_image_stack but it's not > able to make control points DSC_8625.tif no matter what I use. E.g. > > > align_image_stack -v -d -i -x -y -z -C --use-given-order -g 40 -a ais > ../DSC_8633.tif ........ > > > I suspect that DSC_8625.tif does not lign up because the image is > translated too much. > > > T. Modes, have I interpret the *exposure fused from any arrangement *workflow > wrong? How would this problem be dealt with? > -- A list of frequently asked questions is available at: http://wiki.panotools.org/Hugin_FAQ --- You received this message because you are subscribed to the Google Groups "hugin and other free panoramic software" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/hugin-ptx/8da35dc8-2005-4ee5-93fe-b8cbecfcb044%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
DSC_8625 - DSC_8636_A.pto
Description: Binary data
