Re: [hugin-ptx] Re: Hugin Experience
kfj wrote: On 19 Nov., 08:44, Gnome Nomad wrote: After trying many times to level a handheld beach pano using horizontal lines, here's what I did that finally succeeded: 1. Set up horizontal lines from the first frame to each of the other photos, connecting the left edge point on the horizon to the left edge point in each case. 2. Set up similar horizontal lines for the right edge points. 3. Optimized everything INCLUDING translation. Like magic, the horizon straightened itself out. I think it was the translation that made it work. Optimizing without translation didn't straighten out the horizon. I think you probably performed some magic, rather ;-) Could be, maybe the phase of the moon. Your method sounds odd to me, even though coercing the translation parameter into use for a strip panorama might sometimes work, it'll certainly fail in a 360X180. I'm not that ambitious. I did a 4x4 interior panorama of a local cathedral, also handheld, that one aligned quite nicely without any effort on my part. Here's what I'd do: 1. pick out an image which is near the center of your pano and shows a good length of horizon. Set a horizontal line control point on it picking two horizon points as far apart as possible and only optimize roll for this single image - you may have to adapt pitch manually to have the horizon at the right height. 2. start with the leftmost image showing the horizon. Pick two points on the horizon and create a new line (not horizontal or vertical, just a line control point.) Carry on by adding two horizon points from each other image showing the horizon to that same line. 3. Now, with the image chose in 1. as your position anchor, optimize for position. The horizon should be level because of 1. and 2. should should bring all the other images in line. Thanks, I'll have to try that. If your horizon isn't level enough, you can add more horizontal line control points - now try and put these with one point on the leftmost horizon image and one on the rightmost. Finally, keep in mind that your other CPs will likely be from points on the beach, and since the pano is handheld, there will be parallactic errors. Using these CPs will result in your images being aligned by features on the beach, while your horizon goes awry. Try and delete as many of these CPs as possible - with the horizon defined by 1-2-3, you might even get away with one CP per pair (providing your lens is well-calibrated) Have never calibrated any of my lenses. While I'm on the topic I'd like to hint at a technique I sometimes use when I fix horizons: I've made an image in 2:1 format with a degree pattern (30X30 degree checkerboard, translates to, like, 30X30 pixel checkerboard on a 360X180 pixel image) and include this image into the panorama as being equirectangular with 360 degreed hfov. The grid has a clearly defined horizon and I can now 'glue' line CPs to this line. The grid image makes a good anchor, then - and for the stitching I just switch it off in the preview. Now that's an interesting idea! Will have to try that out! -- Gnome Nomad gnomeno...@gmail.com wandering the landscape of god http://www.cafepress.com/otherend/ -- 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 hugin-ptx@googlegroups.com To unsubscribe from this group, send email to hugin-ptx+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/hugin-ptx
Re: [hugin-ptx] Re: Hugin Experience
On Tue 22-Nov-2011 at 14:00 +0100, Rogier Wolff wrote: On Tue, Nov 22, 2011 at 11:46:01AM +, Bruno Postle wrote: Nope, the horizontal and vertical points are evaluated in the output canvas, so the output projection is critical. This is much simpler conceptually, as far as the optimiser is concerned they are the same thing. Do you mean that the control point matching happens in ouput projection space? It is for horizontal and vertical control points. I can't remember if it still is for 'normal' points, I seem to remember this might have changed at some point - You need to look in the pano13 code. A friend shoots "all around" panoramas. As output projection she needs a projection onto a cube around the pano-sphere. So if I understand things correctly, she will set the output projection to equirectangular, stitch an output image, rotate the viewpoint by 90 degrees and stitch another face of the cube until all 6 faces are done. This would be a bad idea since there is no guarantee that the enblend seams would continue over the edges between tiles. Definitely better to stitch an equirectangular and then split it to cubefaces as a subsequent step. -- Bruno -- 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 hugin-ptx@googlegroups.com To unsubscribe from this group, send email to hugin-ptx+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/hugin-ptx
Re: [hugin-ptx] Re: Hugin Experience
On Tue, Nov 22, 2011 at 11:46:01AM +, Bruno Postle wrote: > On 22 Nov 2011 08:15, "Rogier Wolff" wrote: > > > > I thought that for a horizontal controlpoint pair, the lattitude > > simply doesn't count. So all that the optimization step cares about > > is the that they line up horizontally. > > > > Similarly for the vertical control lines. There the horizontal position, > > or longitude is not taken into account. > > This is effectively true with equirectangular, or any of the other > cylindrical output projections. > > > I thought that all this was independent of the projection > > being used for the final result. > > Nope, the horizontal and vertical points are evaluated in the output > canvas, so the output projection is critical. > > This is much simpler conceptually, as far as the optimiser is concerned > they are the same thing. Do you mean that the control point matching happens in ouput projection space? i.e. a controlpoint in Image1 at X1, Y1, and in Image 2 at X2,Y2 is transformed using the parameters for Image1 (i.e. roll1, pitch1) to a roll/pitch coordinate pair in the pano-sphere and then onto the output canvas using the output transformation? The same is then done for the X2, Y2, and the difference is optimized. This would mean that for instance a mercator projection that has a distortion near the poles will favor the lattitude of controlpoints near the pole being "perfect" sacrificing all other overlaps. It would also mean that after changing the output projection, you need to optimize again. A friend shoots "all around" panoramas. As output projection she needs a projection onto a cube around the pano-sphere. So if I understand things correctly, she will set the output projection to equirectangular, stitch an output image, rotate the viewpoint by 90 degrees and stitch another face of the cube until all 6 faces are done. Now optimization is probably done with the output projection set to one of the faces. Now all control points that lie outside the face of the cube are distorted and optimized in weird ways that do not reflect their role in the final output. Of course something can be said for doing it this way: if there is a minute difference in the projection of the layout of two images near the center of the output image, and the same minute difference in degrees on the panosphere expands to several tens of pixels near the edge of the output image, it might be good to "fix" that controlpoint near the edge, and tolerate a slightly larger error on the one in the middle. But I would prefer to optimize in panosphere coordinates. Doing it the other way introduces errors based on the assumption that all controlpoints are perfect. They are not. And the lens parameters are not perfect. I think we'd get a much better fit (in a mathematical sense, on the panosphere) if we'd just use the panoshpere coordinates Once we have that, we're ready to optimize lens parameters etc etc, to get the final errors out. And then a "list the controlpoints starting with the largest error" allows you to find the controlpoints that really have errors in placement. But again... It's entirely possible that I'm misunderstanding how hugin actually works. (I'm reading "cooking for geeks" and the book has explained one simple thing to me and something I didn't manage before (again and again) worked first time, simply because now I understand the underlying chemistry. Similarly I want to know how hugin works to be able to better control it.) Roger. -- ** r.e.wo...@bitwizard.nl ** http://www.BitWizard.nl/ ** +31-15-2600998 ** **Delftechpark 26 2628 XH Delft, The Netherlands. KVK: 27239233** *-- BitWizard writes Linux device drivers for any device you may have! --* The plan was simple, like my brother-in-law Phil. But unlike Phil, this plan just might work. -- 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 hugin-ptx@googlegroups.com To unsubscribe from this group, send email to hugin-ptx+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/hugin-ptx
Re: [hugin-ptx] Re: Hugin Experience
On 22 Nov 2011 08:15, "Rogier Wolff" wrote: > > I thought that for a horizontal controlpoint pair, the lattitude > simply doesn't count. So all that the optimization step cares about > is the that they line up horizontally. > > Similarly for the vertical control lines. There the horizontal position, > or longitude is not taken into account. This is effectively true with equirectangular, or any of the other cylindrical output projections. > I thought that all this was independent of the projection > being used for the final result. Nope, the horizontal and vertical points are evaluated in the output canvas, so the output projection is critical. This is much simpler conceptually, as far as the optimiser is concerned they are the same thing. -- Bruno -- 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 hugin-ptx@googlegroups.com To unsubscribe from this group, send email to hugin-ptx+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/hugin-ptx
Re: [hugin-ptx] Re: Hugin Experience
Bruno, It cannot possibly be the case that I'm right and you're wrong. You know much more about this than I do. So please tell me where my understanding is wrong. On Mon, Nov 21, 2011 at 10:59:46PM +, Bruno Postle wrote: > On Mon 21-Nov-2011 at 13:56 -0800, JohnPW wrote: > >Please clarify this for me as I want to make sure I understand (and it > >may be helpful to other newer Panorama makers like myself.) > >These are my assumptions: > > >1.) Only the actual horizon should be assigned as a "horizontal > >line" (unless you just want some line, or the average of some lines, > >to be straight and at the horizontal center (equator) of the panorama) > >because the horizon line is the only latitudinal line that lies upon a > >great circle line (the equator.) > > Yes, for spherical panoramas. Horizontal lines can also be useful > for removing perspective from façades of buildings, but only when > you are using rectilinear projection for the output. In that case, they should be called horizon-lines instead of horizontal lines. I thought that horizontal and vertical control points matter to the optimization step. Normally, I thought the control points are all transformed into the spherical coordinates, and for each pair both the longitude and lattitude are compared. In fact the distance is calculated and optimized. I thought that for a horizontal controlpoint pair, the lattitude simply doesn't count. So all that the optimization step cares about is the that they line up horizontally. Similarly for the vertical control lines. There the horizontal position, or longitude is not taken into account. I thought that all this was independent of the projection being used for the final result. Roger. -- ** r.e.wo...@bitwizard.nl ** http://www.BitWizard.nl/ ** +31-15-2600998 ** **Delftechpark 26 2628 XH Delft, The Netherlands. KVK: 27239233** *-- BitWizard writes Linux device drivers for any device you may have! --* The plan was simple, like my brother-in-law Phil. But unlike Phil, this plan just might work. -- 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 hugin-ptx@googlegroups.com To unsubscribe from this group, send email to hugin-ptx+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/hugin-ptx
Re: [hugin-ptx] Re: Hugin Experience
On Mon 21-Nov-2011 at 18:37 -0500, Robert Krawitz wrote: On Mon, 21 Nov 2011 22:59:46 +, Bruno Postle wrote: 1.) Only the actual horizon should be assigned as a "horizontal line" (unless you just want some line, or the average of some lines, to be straight and at the horizontal center (equator) of the panorama) because the horizon line is the only latitudinal line that lies upon a great circle line (the equator.) Yes, for spherical panoramas. Horizontal lines can also be useful for removing perspective from façades of buildings, but only when you are using rectilinear projection for the output. What about equirectangular or cylindrical (or Mercator)? In these projections the only features in the scene that will be horizontal in the output image are: the horizon at sea, or features of circular buildings (so long as you are standing in the exact centre of the building). -- Bruno -- 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 hugin-ptx@googlegroups.com To unsubscribe from this group, send email to hugin-ptx+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/hugin-ptx
Re: [hugin-ptx] Re: Hugin Experience
On Mon, 21 Nov 2011 22:59:46 +, Bruno Postle wrote: > On Mon 21-Nov-2011 at 13:56 -0800, JohnPW wrote: >>Please clarify this for me as I want to make sure I understand (and it >>may be helpful to other newer Panorama makers like myself.) >>These are my assumptions: > >>1.) Only the actual horizon should be assigned as a "horizontal >>line" (unless you just want some line, or the average of some lines, >>to be straight and at the horizontal center (equator) of the panorama) >>because the horizon line is the only latitudinal line that lies upon a >>great circle line (the equator.) > > Yes, for spherical panoramas. Horizontal lines can also be useful for > removing perspective from façades of buildings, but only when you are using > rectilinear projection for the output. What about equirectangular or cylindrical (or Mercator)? -- Robert Krawitz Tall Clubs International -- http://www.tall.org/ or 1-888-IM-TALL-2 Member of the League for Programming Freedom -- http://ProgFree.org Project lead for Gutenprint --http://gimp-print.sourceforge.net "Linux doesn't dictate how I work, I dictate how Linux works." --Eric Crampton -- 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 hugin-ptx@googlegroups.com To unsubscribe from this group, send email to hugin-ptx+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/hugin-ptx
Re: [hugin-ptx] Re: Hugin Experience
Erik Krause wrote: Am 08.04.2011 10:43, schrieb Yclept Nemo: By the way, whats the difference between vertical control point lines and horizontal control point lines? -> http://wiki.panotools.org/Horizontal_control_points -> http://wiki.panotools.org/Vertical_control_points -> http://wiki.panotools.org/Panotools_internals#Line_control_points > And is it useful to have these lines across different images Yes, it is useful. If you want to level your panorama using the horizon it is best to have horizontal CPs approximately 45° apart. After trying many times to level a handheld beach pano using horizontal lines, here's what I did that finally succeeded: 1. Set up horizontal lines from the first frame to each of the other photos, connecting the left edge point on the horizon to the left edge point in each case. 2. Set up similar horizontal lines for the right edge points. 3. Optimized everything INCLUDING translation. Like magic, the horizon straightened itself out. I think it was the translation that made it work. Optimizing without translation didn't straighten out the horizon. Now I need to go back and re-do some of my older panos that had similar issues. -- Gnome Nomad gnomeno...@gmail.com wandering the landscape of god http://www.cafepress.com/otherend/ -- 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 hugin-ptx@googlegroups.com To unsubscribe from this group, send email to hugin-ptx+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/hugin-ptx
Re: [hugin-ptx] Re: Hugin Experience
Yclept Nemo wrote: Is it possible to create a rectilinear panorama from images (shot from one location) ranging (tip-to-tip) 67 degrees horizontally and 50 degrees vertically such that the output image plane does not appear to "lean" forwards or backwards, but flat as in http://www.flickr.com/photos/acmace/5239601953/sizes/l/ ? I'm only able to create a semi-decent projection from equirectangular and even then the top flanges outwards, but perhaps this is since the image was shot quite close (100ft) to the base of the buildings and there is substantial amounts of perspective distortion. You can only retro-correct perspective if the subject is flat (2D); the facades of many building are (close to) 2D, so this is feasible. If you alter the perspective (implicitly altering the point where the image was taken), and there is genuine perspective in the image (lots of 3D objects ate various distances) it all goes wrong. I used perspective correction to "square up" this shot of some gothic carving, and the strongly 3D carving looks a bit odd. http://galootcentral.com/components/cpgalbums/userpics/10152/gothic.JPG BugBear -- 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 hugin-ptx@googlegroups.com To unsubscribe from this group, send email to hugin-ptx+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/hugin-ptx
Re: [hugin-ptx] Re: Hugin Experience
> Add several vertical line control points and optimize ypr of all images > except the yaw of one center image. > The horizon should end up going through the center of the pano. > The pano can then be cropped to show just the image. That is what I did ... I guess because the picture was literally taken leaning backwards and looking upwards and the horizon line really runs through the bottom portion of the image, it is not possible to flatten the rectilinear version. I'm satisfied with mercator however. -- 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 hugin-ptx@googlegroups.com To unsubscribe from this group, send email to hugin-ptx+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/hugin-ptx
Re: [hugin-ptx] Re: Hugin Experience
> Btw, this is an awesome image! Yeah, isn't it! I pulled it off the hugin flickriver stream. Ok, I have an HDR EXR image that was successfully and completely written to by Hugin but which is still broken... it's 10MB if anyone wants it. I also posted a smaller PNG version at http://imagebin.org/147952 Notice the black areas, as well as the unnaturally noisy white area in the bottom left. I am having a problem with the corresponding tif fused output version. There are two perfectly horizontal 15px-wide dark (darker EV) or burned bands running directly across the image, not near any seams. Within these bands are three additional dodged or lightened bands. While the top band is running across the bottom quarter of the image where there is no overlap, it is still too high to crop - I would lose 1/8 of my pano. -- 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 hugin-ptx@googlegroups.com To unsubscribe from this group, send email to hugin-ptx+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/hugin-ptx
Re: [hugin-ptx] Re: Hugin Experience
On 2011-04-12 11:34 PM, Yclept Nemo wrote: Is it possible to create a rectilinear panorama from images such that the output image plane does not appear to "lean" forwards or backwards, but flat? Add several vertical line control points and optimize ypr of all images except the yaw of one center image. The horizon should end up going through the center of the pano. The pano can then be cropped to show just the image. -- Jim Watters http://photocreations.ca -- 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 hugin-ptx@googlegroups.com To unsubscribe from this group, send email to hugin-ptx+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/hugin-ptx
Re: [hugin-ptx] Re: Hugin Experience
Erik Krause wrote: Am 12.04.2011 10:41, schrieb paul womack: > Horizontal lines and vertical lines can have only one pair each. Ah - hah. So for a multi-point horizon I should make a straight line which happens to be horizontal? No, just use multiple pairs of horizontal control points. You can't force a straight line to be horizontal. How about a combination: Put a horizontal pair on A-E and a straight line across A-B-C-D-E ? BugBear -- 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 hugin-ptx@googlegroups.com To unsubscribe from this group, send email to hugin-ptx+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/hugin-ptx
RE: [hugin-ptx] Re: Hugin Experience
Btw, this is an awesome image! > Date: Tue, 12 Apr 2011 22:34:45 -0400 > Subject: Re: [hugin-ptx] Re: Hugin Experience > From: orbisvi...@gmail.com > To: hugin-ptx@googlegroups.com > > Is it possible to create a rectilinear panorama from images (shot from > one location) ranging (tip-to-tip) 67 degrees horizontally and 50 > degrees vertically such that the output image plane does not appear to > "lean" forwards or backwards, but flat as in > http://www.flickr.com/photos/acmace/5239601953/sizes/l/ ? > > I'm only able to create a semi-decent projection from equirectangular -- 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 hugin-ptx@googlegroups.com To unsubscribe from this group, send email to hugin-ptx+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/hugin-ptx
Re: [hugin-ptx] Re: Hugin Experience
Mm sorry FOV ranges from 142 degrees horizontal to 144 degrees vertical according to the sticther tab... whereas I calculated FOV from the images tab as (yaw group0 - yaw group1) * (# groups) + [2-(mod2 #groups)/2](yaw group1 - yaw group1). What is the difference ? -- 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 hugin-ptx@googlegroups.com To unsubscribe from this group, send email to hugin-ptx+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/hugin-ptx
Re: [hugin-ptx] Re: Hugin Experience
Is it possible to create a rectilinear panorama from images (shot from one location) ranging (tip-to-tip) 67 degrees horizontally and 50 degrees vertically such that the output image plane does not appear to "lean" forwards or backwards, but flat as in http://www.flickr.com/photos/acmace/5239601953/sizes/l/ ? I'm only able to create a semi-decent projection from equirectangular and even then the top flanges outwards, but perhaps this is since the image was shot quite close (100ft) to the base of the buildings and there is substantial amounts of perspective distortion. -- 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 hugin-ptx@googlegroups.com To unsubscribe from this group, send email to hugin-ptx+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/hugin-ptx
Re: [hugin-ptx] Re: Hugin Experience
I ask because enblend has been running for 18 hours so far -- 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 hugin-ptx@googlegroups.com To unsubscribe from this group, send email to hugin-ptx+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/hugin-ptx
Re: [hugin-ptx] Re: Hugin Experience
I mean to ask, is this known? Is there a workaround? I see no new versions of enblend... -- 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 hugin-ptx@googlegroups.com To unsubscribe from this group, send email to hugin-ptx+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/hugin-ptx
Re: [hugin-ptx] Re: Hugin Experience
I should probably start a new thread for this, but whatever...
Hugin HDR panorama output (2010.4.0) is broken. Whether outputting the
final output as EXR or TIFF, the intermediate stacks are in the EXR
format. For example, ProjectName_stack_hdr_[[:digit:]]{4}.exr. These
intermediate stacks are broken. They are unable to be recognized or
opened by geeqie, gthumb, gimp. Cinepaint will open these files but
display an empty panorama-sized canvas. The reference viewer
exrdisplay displays a dithered mess of red/cyan/yellow/magenta pixels
against a white brackground. Convert from imagemagick and exrtopng
from exrtools both segfault while attempting to convert these files to
a more common format. exrtopng claims to have finished the conversion
before segfaulting, however the resulting png is nothing but a
pure-white stack-sized image curved from barrel distortion against a
transparent rectangular background.
Enblend, whether writing the final output as EXR or TIFF, will
continue writing forever. When TIFF, outputs a base 600MB image within
the first minute, then adds 100MB ever hour nonstop... 1.6GB to go and
no end in site - the fused version is 800MB. When EXR, outputs a base
2.2MB image within the first image, then adds 500K per hour.
--
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 hugin-ptx@googlegroups.com
To unsubscribe from this group, send email to
hugin-ptx+unsubscr...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/hugin-ptx
Re: [hugin-ptx] Re: Hugin Experience
> Ah - hah. So for a multi-point horizon I should make a straight line > which happens to be horizontal? I found vertical lines work better than straight lines which happen to be vertical. The images between the horizontal/vertical lines should be oriented by overlapping normal control points, imo. -- 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 hugin-ptx@googlegroups.com To unsubscribe from this group, send email to hugin-ptx+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/hugin-ptx
Re: [hugin-ptx] Re: Hugin Experience
Erik Krause wrote: Am 11.04.2011 10:48, schrieb paul womack: If I understand you right, a line control can have more than two points associated with it. Only straight lines can have more than two points. They are designated by the same t-number (>2). Hugin allows for this, just select the same line number under "mode". Horizontal lines and vertical lines can have only one pair each. Ah - hah. So for a multi-point horizon I should make a straight line which happens to be horizontal? BugBear -- 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 hugin-ptx@googlegroups.com To unsubscribe from this group, send email to hugin-ptx+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/hugin-ptx
Re: [hugin-ptx] Re: Hugin Experience
Thanks for all the points, I've aligned specific stacks and created stack masks, now despite a mundane landscape the fused version nonetheless looks impressive. My panorama is outputting to "exposure fused from stacks as TIFF" and "HDR as EXR", in both cases the blend process has taken about two hours. However the fused TIFF file was written to in under 15 minutes, whereas enblend has been "writing final output" of the EXR file for the past eight hours. The file size is 2.6MB, six hours ago it was 2.2MB (compare to 870MB TIFF). I'm not familiar enough in terms of size, compression, and processing power required by the OpenEXR format to understand if this is a bug in 2010.4.0. Should I backup the individual stacks, cancel the current process, and attempt manually stitching to TIFF instead? -- 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 hugin-ptx@googlegroups.com To unsubscribe from this group, send email to hugin-ptx+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/hugin-ptx
Re: [hugin-ptx] Re: Hugin Experience
Jim Watters wrote: On 2011-04-08 10:35 AM, paul womack wrote: whats the difference between vertical control point lines and horizontal control point lines? -> http://wiki.panotools.org/Horizontal_control_points -> http://wiki.panotools.org/Vertical_control_points -> http://wiki.panotools.org/Panotools_internals#Line_control_points If I have 5 images A B C D E that I wish to line up, am I better making ? A-B, B-C, C-D, D-E or A-B, A-C, A-D, A-E or A-C, B-C, C-D, D-E pairs? BugBear "Line" control points are evaluated (optimized) by comparing the a line drawn by the first two points and the distance all other points are from that line. So I always place the first two points at the ends of the line. The rest of the points for that line makes no difference how they are added. So if I had multiple straight lines going though multiple images I would probably do. A-E, B-D, C-C If does not matter how the "line" points are added to B, C, D Important: Horizontal, vertical, and straight lines are evaluated on their output projection. If I understand you right, a line control can have more than two points associated with it. I'm using Version: 2010.5.0.b5a907b23b85 and do not see any obvious way to create anything other than a two point control (pair). BugBear -- 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 hugin-ptx@googlegroups.com To unsubscribe from this group, send email to hugin-ptx+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/hugin-ptx
Re: [hugin-ptx] Re: Hugin Experience
I ran a rough test (I don't have the tripod) and determined that my NPP was displaced by approximately 10cm - which caused about a 4px pixel error for mid-range objects. It's really neat that object distance can be estimated from pixel error. I've noticed that enblend is 99% of the stiching process; furthermore it seems severely limited by disk-io. If enblend is swapping to disk due to a memory limit then the enblend manually describes how to increase the limit; is this a good idea? I run a 4GB system of which 3.25GB are available to user processes; enblend uses 28%, 932MB very near to the 1GB limit. If I increase the limit to 2.5GB or 2.25GB will it create a noticeable speed-up? On the other hand, if enblend is simply writing to disk, then I'd like to know how much space on-disk it uses/requires - to where is it writing? Therefore maybe I can create a 1.5GB ramdisk.. which begs the question, if the ramdisk becomes full, will enblend fall back to another partition? -- 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 hugin-ptx@googlegroups.com To unsubscribe from this group, send email to hugin-ptx+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/hugin-ptx
Re: [hugin-ptx] Re: Hugin Experience
Yclept Nemo kirjoitti viestissään (lähetysaika lauantai, 9. huhtikuuta 2011): > Ah well! I have "Ø" symbol printed on the side of my Canon EOS > 350, I was told this was the NPP point... > Anyone know what this point marks? It marks the sensor plane location in the body. The no-parallax point is in the _lens_, not in the body. See the following documents: http://www.johnhpanos.com/epcalib.htm http://toothwalker.org/optics/cop.html#stitching http://www.janrik.net/PanoPostings/NoParallaxPoint/TheoryOfTheNoParallaxPoint.pdf -- Markku Kolkka markku.kol...@iki.fi -- 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 hugin-ptx@googlegroups.com To unsubscribe from this group, send email to hugin-ptx+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/hugin-ptx
Re: [hugin-ptx] Re: Hugin Experience
Ah well! I have "Ø" symbol printed on the side of my Canon EOS 350, I was told this was the NPP point... as you pointed out, likely not. http://www.dpreview.com/reviews/CanonEOS350D/Images/allroundview.jpg Top right view, above the strap slit. Anyone know what this point marks? -- 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 hugin-ptx@googlegroups.com To unsubscribe from this group, send email to hugin-ptx+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/hugin-ptx
Re: [hugin-ptx] Re: Hugin Experience
> Important: Horizontal, vertical, and straight lines are evaluated on their > output projection. Hm, so that's why my mercator projection + straight line @ ~25° was throwing off the alignment... so this means that: equirectangular: vertical lines only, plus horizon line Does this also apply to cylindrical, mercator, miller, architectural ? (i'm using mercator) That explains the GSoC idea of moving "line" control points to the preview windows. > Yes, it is useful. If you want to level your panorama using the horizon it is > best to have horizontal CPs approximately 45° apart. Only problem I see with horizon control lines, is that unlike straight lines they only specify two points. Horizontal CPs are more effective further apart, however all the intermediate images are not straightened: they can be wavy/squiggly. -- 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 hugin-ptx@googlegroups.com To unsubscribe from this group, send email to hugin-ptx+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/hugin-ptx
Re: [hugin-ptx] Re: Hugin Experience
On 2011-04-08 10:35 AM, paul womack wrote: whats the difference between vertical control point lines and horizontal control point lines? -> http://wiki.panotools.org/Horizontal_control_points -> http://wiki.panotools.org/Vertical_control_points -> http://wiki.panotools.org/Panotools_internals#Line_control_points If I have 5 images A B C D E that I wish to line up, am I better making ? A-B, B-C, C-D, D-E or A-B, A-C, A-D, A-E or A-C, B-C, C-D, D-E pairs? BugBear "Line" control points are evaluated (optimized) by comparing the a line drawn by the first two points and the distance all other points are from that line. So I always place the first two points at the ends of the line. The rest of the points for that line makes no difference how they are added. So if I had multiple straight lines going though multiple images I would probably do. A-E, B-D, C-C If does not matter how the "line" points are added to B, C, D Important: Horizontal, vertical, and straight lines are evaluated on their output projection. -- Jim Watters http://photocreations.ca -- 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 hugin-ptx@googlegroups.com To unsubscribe from this group, send email to hugin-ptx+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/hugin-ptx
Re: [hugin-ptx] Re: Hugin Experience
thanks for the suggestions: Just to be clear, the problem is not in my control points. I have 24 stacks with 50% overlap, per overlap I've manually placed 20-40 high-correlation well-distributed accurate control points. After optimizing my average error is 0.4, rms error 0.6, max error 1.7. The problem is that despite such a high accuracy, hugin produces a panorama with so many and so noticeable artifacts. Furthermore, I doubt the problem can be attributed to accuracy: I aligned the camera's sensor plane with the panoramic head's center of rotation, and pretty much all objects are more than 35 feet distant. In any case, your suggestion was really very good, am I'm running a test stitch now. My zoom lens has slots around the barrel demarcating focal lengths, so therefore it was relatively easy to set a focal length of 28mm. The barrel is stiff enough to prevent the focal length from changing. Furthermore, the camera provides the focal length in the exif image data and it turns out this number is consistent across all images; this is what hugin uses to calculate FOV. Nonetheless I adopted the philosophy of "distort each individual image as non-physically as needed to minimize control-point distances". Even though I was pretty sure all lens parameters where correct (v) and consistent (a,b,c,d,e) across all images, I nonetheless unlinked and optimized v,a,b,c,d,e for each image: After optimizing my average error is 0.1, rms error 0.2 and maximum error distance 0.93. Like I said, I'm still stitching but hopefully this gets rid of the glitches once-and-for-all. I'm also going to attempt another test stitch with a similar strategy on the version with the straight lines. -- 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 hugin-ptx@googlegroups.com To unsubscribe from this group, send email to hugin-ptx+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/hugin-ptx
Re: [hugin-ptx] Re: Hugin Experience
Erik Krause wrote: Am 08.04.2011 10:43, schrieb Yclept Nemo: By the way, whats the difference between vertical control point lines and horizontal control point lines? -> http://wiki.panotools.org/Horizontal_control_points -> http://wiki.panotools.org/Vertical_control_points -> http://wiki.panotools.org/Panotools_internals#Line_control_points > And is it useful to have these lines across different images Yes, it is useful. If you want to level your panorama using the horizon it is best to have horizontal CPs approximately 45° apart. With regards to "line" control points. If I have 5 images A B C D E that I wish to line up, am I better making A-B, B-C, C-D, D-E or A-B, A-C, A-D, A-E or A-C, B-C, C-D, D-E pairs? BugBear -- 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 hugin-ptx@googlegroups.com To unsubscribe from this group, send email to hugin-ptx+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/hugin-ptx
