XP SP3 - APG3b3 - Encountering strange optimisation results
Using APG3b3 to stitch 6 FE images together.
Initial stitch comes up with some links and then I go over each pair of images to increase the number of RMS<5.0 links to get a better, overall RMS.
What I am finding with one pano is that when I use "Add auto control points", "Quick optimisation", "clean bad control ponts" (<5.0), "full optimisation". I do this for each of links between images. i.e. one linking image 1 to 2, 2 to 3, 3 to 4 .... 6 to 1 etc.
What I am seeing that in 3b3 is that where I have previously had a good set of control points between two images, when working on another pailr and using the above workflow, those previously good links are either then losing ALL of these good control points or they are changing to +10 or +20 RMS control points when full optimisation is run!?!
Tried same set of images in 2.6.3 and have no issues with this.
Having trouble to replicate to show screen shots as when I encounter it, reverse the change to take the before screen shot and re-do it doesn't replicate!!??!
I'll carry on trying to catch it but any thoughts or unique to myself?
Re: XP SP3 - APG3b3 - Encountering strange optimisation results
Found an example.
First image I am linking images 1 & 2 together by auto-adding further control points. Quick optimisation gives the following good CPs. Then auto removed >5.0 RMS CPs. As you can just see, Img 4 is linked to 3 & 5 in the CP editor window.
Second image shows Img 4 with what CP's are linked.
Third image is after a full optimisation. as you can see Img 4 now has had all CPs deleted!!??!! WTF!?!