Vertical Hugin Panorama of Building has Discontinuity on sides

[scott092707]

New problem:

I have a two-photo vertical panorama of a building (with a neat bunch of sculptures at the top), and despite indicating the entire sides of the buildings in each photo of the same lines - bottom to top - the building suddenly becomes wider  on each side.

I have done further attempts, by extending the lines as far as possible, but the only thing that changes is that the discontinuity became almost at the same level on each side of the building.

The photos: https://imgur.com/a/Kwvoyfu

I will attach the .pto file

Any hints you can give would be greatly appreciated...

-Scott

[johnfi...@gmail.com]

On Sunday, February 20, 2022 at 4:52:32 PM UTC-5 scott092707 wrote:

New problem:

I have a two-photo vertical panorama of a building (with a neat bunch of sculptures at the top), and despite indicating the entire sides of the buildings in each photo of the same lines - bottom to top - the building suddenly becomes wider  on each side.

The answer you seem to need is the same as last time:  You need to optimize with additional parameters.  The discontinuities go away with a better fit.  The pop up after each optimize attempt tells you a lot about how well it did.

I only tried a few combinations and found the two extra parameters that make the most difference.  I didn't test whether adding just those two to the basic yaw, pitch and roll would be enough.  These and a few others were definitely enough and the following two were most of the improvement.

The lens b parameter:  I think I understand why, but I'm not sure enough to post that where others here understand that better.

The translation y parameter:  When you changed the pitch (upward angle) between photos, you did not rotate on exactly the perfect axis for your lens, so in effect you changed the vertical position of the camera in addition to the vertical angle.  As long as the important parts of the subjects where they cross the boundary are all the same distance away, that deviation from perfect picture taking is fixed by optimizing the translation y.

[scott092707]

Well, I'll look into that.

I actually did bring up the Optimizer Tab as before (since it solved the last issue), but the only optimizations presented to me this time were yaw/pitch/roll.

I'll try again tomorrow (?) and see where one can ADD parameters, and do the 'translation y'.

I'll also look at the lens section (didn't do so before), and look to selecting the 'lens b'.

Thank you for answering so quickly.

-Scott

[Bruno Postle]

On Mon, 21 Feb 2022 at 02:31, scott092707 wrote:

I actually did bring up the Optimizer Tab as before (since it solved the last issue), but the only optimizations presented to me this time were yaw/pitch/roll.

I'll try again tomorrow (?) and see where one can ADD parameters, and do the 'translation y'.

I'll also look at the lens section (didn't do so before), and look to selecting the 'lens b'.

The XYZ Translation parameters are initially hidden as they are not needed for normal usage, they are made available by selecting Menu -> Interface -> Expert.

--

Bruno

[johnfi...@gmail.com]

On Monday, February 21, 2022 at 5:52:13 AM UTC-5 bruno...@gmail.com wrote:

The XYZ Translation parameters are initially hidden as they are not needed for normal usage, they are made available by selecting Menu -> Interface -> Expert.

That is one of the reasons I think only the "Expert" interface is usable.

I understand the concept that the alignment doesn't need to be great:  The blending tools can cover up for poor alignment.  But I've never seen that work for any of my panoramas.  I have sometimes used detailed masks to cover up for poor alignment.  But otherwise, poor alignment gives a terrible result.

I understand that correct use of a tripod with a nodal slide can give you a collection of photos in which there is zero Translation.  But I haven't managed that yet, even with careful adjustment of my nodal slide.  I also understand that this issue is much less significant for panoramas in which the subject is very far away from the camera (500 meters or more).  But I think the more common case for ordinary photographers has enough translation to require those in the optimization.

As for the lens, I noticed the slight barrel shape of the parallel vertical lines in the original two photos of this thread.  I don't know whether to consider that a lens characteristic different from the default "rectilinear":  The part of such parallel lines that are perpendicular to the point of view are wider apart in viewing angle, and the viewing angle between the lines goes down with the distance from that perpendicular.  But those two photos both had the camera pointing up relative to the face of the building.  The perpendicular point is at the bottom of the lower picture.  So the visible slight bulge in parallel vertical lines looks to me like it is closer to the vertical center of the image, which would definitely be a lens characteristic, as opposed to representing the actual point of widest viewing angle between the lines (which I don't know whether that would also be a lens characteristic different from "rectilinear").

[Bruno Postle]

On Mon, 21 Feb 2022 at 11:46, johnfi...@gmail.com <johnfi...@gmail.com> wrote:

On Monday, February 21, 2022 at 5:52:13 AM UTC-5 bruno...@gmail.com wrote:

The XYZ Translation parameters are initially hidden as they are not needed for normal usage, they are made available by selecting Menu -> Interface -> Expert.

That is one of the reasons I think only the "Expert" interface is usable.

I understand the concept that the alignment doesn't need to be great:  The blending tools can cover up for poor alignment.  But I've never seen that work for any of my panoramas.  I have sometimes used detailed masks to cover up for poor alignment.  But otherwise, poor alignment gives a terrible result.

The reason why it is hidden is that although it accounts for 3D movement of the camera, it is only really useful for assembling an image of a flat surface (like the facade of a building), it isn't generally useful for assembling a 'normal' panorama.

With a normal panorama you might find that occasionally adding the X, Y, Z parameters to the optimisation helps get a better fit, but really you are just substituting for the lack of morph to fit functionality, rather than calculating the 'actual' values for these parameters.

 

As for the lens, I noticed the slight barrel shape of the parallel vertical lines in the original two photos of this thread.  I don't know whether to consider that a lens characteristic different from the default "rectilinear":  The part of such parallel lines that are perpendicular to the point of view are wider apart in viewing angle, and the viewing angle between the lines goes down with the distance from that perpendicular.  But those two photos both had the camera pointing up relative to the face of the building.  The perpendicular point is at the bottom of the lower picture.  So the visible slight bulge in parallel vertical lines looks to me like it is closer to the vertical center of the image, which would definitely be a lens characteristic, as opposed to representing the actual point of widest viewing angle between the lines (which I don't know whether that would also be a lens characteristic different from "rectilinear").

The barrel distortion is normal with any lens, and is accounted-for in the a,b & c (and d & e) lens parameters. Though in general you should only need to use the 'b' parameter for a rectilinear lens.

--

Bruno

--

Bruno

[scott092707]

Before coming back here, and reading that I needed to set the interface to "Expert" to get the translation parameters,

I found no way to add them, and just did the Lens b parameter, and got a perfect result!

(see new image added to the imgur page)

(Maybe TOO perfect - I'm not sure that it might not look better slightly tilted away towards the top - after all, I AM looking

up at the top of the building from the bottom...  What do people think!  This is just aesthetics, of course...

I probably won't bother, unless the consensus is to tilt it.)