On October 31, 2010 04:47:02 pm Bernd Hohmann wrote: > Pardon me: you sound like my daughter in her younger years by mentioning > "hundreds". Why not thousand or a million shots for a pano?
Not pardoned. Your daughter in her younger year was probably brighter and wiser than you. What kind of glasses do you wear? ever seen a tele-lens? you seem to have a very very narrow vision, with the FoV of a 400mm lens. Unfortunately you don't seem to see that far. > 6H 1Z 1N is common, 6H 3Z 1N usual, 6H 6Z 1N and 6N 6Z also. > > No 100 shots. 10-12 shots. Maybe 18. There is a guy in the room who has never heard of the gigapan? not only short-sighted, but also deaf? > >> Ok. And what is your idea how to pre-distribute it? > > > > first: assume there is a shooting patter - if the pictures are all over > > the place, only manual positioning helps (and your 0/1/2 scheme is an > > incomplete system for manual positioning) > > Well, if you're shooting your panos while doing the "Jumping Jack" I > understand your concern. Me (and likely most of the others) is shooting > 360° around then zeniths and nadir. before speaking for "most of the others", you may want to do some research, for example on Flickr. Most panoramas are shoot with very simple point and shoot cameras, in the range of 28mm to 200mm, not with 8mm fisheye. I leave it to you to figure out the math of the rows and columns @ 28mm, you seem to be so smart. > Yes, manual input. That was my proposal. So what is yours? Don't assume that there is a zenith or a nadir. It is a needless complication. Most of the time there is neither zenith nor nadir. Also for sphericals. I have done most of my sphericals with either three around or six around. No zenith/nadir. At this very moment I am processing a spherical - 15mm fisheye on a full frame camera. 6 around looking down 15° plus one zenith and there is no need for a nadir. I still don't see the purpose of your request other than complicating something that does not need complication. The way Hugin works now is to assume that *if* there is a zenith and a nadir, they are the first/last shoot in the sequence. It works pretty well. For everything else, you can place the images manually on the fast preview canvas and at some point position information will be passed to the CP detector. > >> Sure, I'm using this mode for my ASIF wrapper. Either something is > >> wrong in Thomas implementation or I did something wrong: it simply > >> counts file_0+file_1, file_1+file_2 and so on. > >> > >> Didn't found any special multirow feature. > > > > I would dare to hint at something on your end. Nothing wrong, you just > > overlooked that this sequence is the "snake" going from zenith to nadir > > or the other way around. It makes a simple and very powerful > > assumption: pictures that are next to each other in the shooting > > sequence are also next to each other in the panorama. good enough for > > CP detection. > > Unfortunately I don't make my panoramas in the order "column 1, column > 2..." but "row 1, row 2". you are seeing the snake through dirty glasses. An overflying pigeon made a poop on them. You tried to clean them with your handkerchief but you forgot that you just blew your nose with it. A snake going from zenith to nadir or the other way around does not necessarily do so in vertical movements. picture a Z, not a N. You are currently imagining a sequence of N next to each other, which you obviously know it is the wrong way since you know how a panorama head looks like. > If you ever used a panorama adapter / panorama head you know where the > "row order" comes from. At the moment I happen to have half a dozen such adapters, Over time I had many more. I even built a few of my own, and I think you need to make some homework and understand what you are talking about before making wrong assumptions, wrong statements, and offensive comments. > > So again my question to you: what for do you need to define the shooting > > pattern for? > > Because your workflow is crackbrained? Or your brain crackflowd? Yuv
signature.asc
Description: This is a digitally signed message part.
