Expanding Size of Pano (detection excessively overlaps)  

Share your tips and tricks here or get help with any Autopano Pro / Giga problem!
No bug reports (of any kind) in this forum!
no avatar
jeradg
Member
 
Topic author
Posts: 99
Likes: 0 post
Liked in: 0 post
Joined: Tue Jan 17, 2012 6:31 pm
Info

Expanding Size of Pano (detection excessively overlaps)

by jeradg » Thu May 17, 2012 7:53 pm

If I detect a pano, and some of the images are overlapping more than they should be.

I can't figure out a way to make the canvas size bigger so I can lay the images in the proper places.

Any suggestions/help would be appreciated.

Thank you,

Jerad

User avatar
AlexandreJ
Kolor Team
 
Posts: 5987
Likes: 7 posts
Liked in: 10 posts
Joined: Mon Nov 14, 2005 4:56 pm
Location: Francin, France
Info

by AlexandreJ » Mon May 21, 2012 8:20 am

I do not understand your question : can you rephrase it or even better, do you an illustration of the issue ?

no avatar
jeradg
Member
 
Topic author
Posts: 99
Likes: 0 post
Liked in: 0 post
Joined: Tue Jan 17, 2012 6:31 pm
Info

by jeradg » Mon May 21, 2012 3:55 pm

I drag the images I want into APP 2.6, and click detect. The result is APP linking the images overlapping at say 50%, but the images only really have 30% overlap. If this happens for many links, then the result of the width of the entire pano is much less than it should be.

I am looking for a way to make the pano preview (as seen when attempting to make corrections prior to rendering) not fit the width of the screen. I want to manually move the images to the correct position, but there is not enough room on the background to spread them out.

I am currently attempting to recreate my problem and I'll post a screenshot.

no avatar
mediavets
Moderator
 
Posts: 16415
Likes: 2 posts
Liked in: 130 posts
Joined: Wed Nov 14, 2007 2:12 pm
Location: Isleham, Cambridgeshire, UK.
Info

by mediavets » Mon May 21, 2012 4:42 pm

jeradg wrote:I drag the images I want into APP 2.6, and click detect. The result is APP linking the images overlapping at say 50%, but the images only really have 30% overlap. If this happens for many links, then the result of the width of the entire pano is much less than it should be.

I am looking for a way to make the pano preview (as seen when attempting to make corrections prior to rendering) not fit the width of the screen. I want to manually move the images to the correct position, but there is not enough room on the background to spread them out.

I am currently attempting to recreate my problem and I'll post a screenshot.

Looking back through some of your previous posts I se that you use a Nikon D5100 witha Bower (Samyang-clone) 8mm fisheye lens on a NN3 pano yhead.

If you have stitching anomalies when shooting with such a setup then I feel that either you are using a weird shooting pattern or the nature of the scene means you are getting spurious links detected with automatic control point detection.

In the latter case it shoudl be easy to identify and remove those false links.

What shooting pattern are you using?

If you are shooting spherical panos always with the setup described above then you can force the focal length and lens type if your images lack EXIF data; and also force the projection type and pano FOV.




Andrew Stephens
Many different Nodal Ninja and Agnos pano heads. Merlin/Panogear mount with Papywizard on Nokia Internet tablets.
Nikon D5100 and D40, Sigma 8mm f3.5 FE, Nikon 10.5mm FE, 35mm, 50mm, 18-55mm, 70-210mm. Promote control.

no avatar
jeradg
Member
 
Topic author
Posts: 99
Likes: 0 post
Liked in: 0 post
Joined: Tue Jan 17, 2012 6:31 pm
Info

by jeradg » Mon May 21, 2012 4:56 pm

The first one is the way the initial detection came out, but I moved image 5 up a little to show the little vertical ability there is for expansion.
The second one is the pictures arranged next to each other. There is not enough space on either side to lay them properly side by side.
The third is the result of many deleted links, local optimizations, full optimization and manual control pt adding.


Just a side note, the initial detection today was worse for some reason than it was a few days ago when I actually worked to the end result shown in the third image. If the initial detection was as bad as it is today, idk if I would be able to complete it.

I know that shooting in such close-quarters as a hallway with 2 walls with no distinguishable features is ill-advised (and difficult), and that is why I added numbers to the walls that I will get rid of later. The numbers didn't seem to help APP all that much though, so idk if I will be trying that again in the future.







User avatar
klausesser
Member
 
Posts: 8836
Likes: 5 posts
Liked in: 64 posts
Joined: Mon May 22, 2006 12:18 am
Location: Duesseldorf, Germany
Info

by klausesser » Mon May 21, 2012 5:15 pm

jeradg wrote:Just a side note, the initial detection today was worse for some reason than it was a few days ago when I actually worked to the end result shown in the third image. If the initial detection was as bad as it is today, idk if I would be able to complete it.

Hi!

Would you tell us about your setup?

best, Klaus
Simplicity is the keynote of all true elegance. Coco Chanel

no avatar
jeradg
Member
 
Topic author
Posts: 99
Likes: 0 post
Liked in: 0 post
Joined: Tue Jan 17, 2012 6:31 pm
Info

by jeradg » Mon May 21, 2012 6:15 pm

mediavets wrote:
jeradg wrote:I drag the images I want into APP 2.6, and click detect. The result is APP linking the images overlapping at say 50%, but the images only really have 30% overlap. If this happens for many links, then the result of the width of the entire pano is much less than it should be.

I am looking for a way to make the pano preview (as seen when attempting to make corrections prior to rendering) not fit the width of the screen. I want to manually move the images to the correct position, but there is not enough room on the background to spread them out.

I am currently attempting to recreate my problem and I'll post a screenshot.

Looking back through some of your previous posts I se that you use a Nikon D5100 witha Bower (Samyang-clone) 8mm fisheye lens on a NN3 pano yhead.

If you have stitching anomalies when shooting with such a setup then I feel that either you are using a weird shooting pattern or the nature of the scene means you are getting spurious links detected with automatic control point detection.

In the latter case it shoudl be easy to identify and remove those false links.

What shooting pattern are you using?

If you are shooting spherical panos always with the setup described above then you can force the focal length and lens type if your images lack EXIF data; and also force the projection type and pano FOV.

I switched last week to stop using fisheye because there is much less distortion and greater clarity without it. I am using the Nikon D5100 with the lens that I believe came with the camera. It says AF-5 Nikkor 18-55 mm 1:3.5-5.6G.

Using my Nodal Ninja I shoot at 15 degrees up all the way around, then 15 degrees down all the way around.

User avatar
klausesser
Member
 
Posts: 8836
Likes: 5 posts
Liked in: 64 posts
Joined: Mon May 22, 2006 12:18 am
Location: Duesseldorf, Germany
Info

by klausesser » Mon May 21, 2012 6:40 pm

jeradg wrote:I switched last week to stop using fisheye because there is much less distortion and greater clarity without it.

Hi!

That´s a big misunderstanding. Fisheye-distortion is completely eliminated in a spherical pano.

There´s absolutely no difference in terms of distortion between a sphere shot with a fisheye and a sphere shot with, let´s say, 35mm or 85mm.

The only difference is resolution: you get about 3 GPx with a 85mm (264 images) lens on a 5D2 - and you get about 112 MPx on the same camera having 6 images using a fisheye.

best, Klaus
Last edited by klausesser on Mon May 21, 2012 6:41 pm, edited 1 time in total.
Simplicity is the keynote of all true elegance. Coco Chanel

no avatar
mediavets
Moderator
 
Posts: 16415
Likes: 2 posts
Liked in: 130 posts
Joined: Wed Nov 14, 2007 2:12 pm
Location: Isleham, Cambridgeshire, UK.
Info

by mediavets » Mon May 21, 2012 6:47 pm

jeradg wrote:I switched last week to stop using fisheye because there is much less distortion and greater clarity without it.

I'm puzzled by your statement as this fisheye lens, and the various other branded clones of the Samyang, has a reputation for optical excellence and it's highly rated by many pano shooters.

What aperture did you use and how were you focusing the lens?

If shooting panos then the fisheye distortion is accounted for when projecting the stitched pano image; so your comment about distortion doesn't really make any sense to me.

I am using the Nikon D5100 with the lens that I believe came with the camera. It says AF-5 Nikkor 18-55 mm 1:3.5-5.6G.

Using my Nodal Ninja I shoot at 15 degrees up all the way around, then 15 degrees down all the way around.

Assuming you are using the 'kit' zoom lens at 18mm focal length then your shooting a pattern is quite inadequate. This is the minimum suggested for 360x180 coverage with your setup as listed here:

http://www.vrwave.com/panoramic-lens-database/nikon/

Minimum (1.5x): N(adir), 10 images every 36° at -45° pitch, 10 images every 36° at 0° pitch, 10 images every 36° at +45° pitch, Z(enith)

.............

Your problems stem from a number of factors I feel:

1. The nature of the scene. It would be so much easier to shoot this scene using your fisheye lens.

2. You chosen shooting pattern.

3. A potential issue with this 'kit' zoom lens of zoom and focus creep at higher +/- pitch angles. At least that what I found with mine on a Merlin robotic head.

.............

I really think you'd be better advised to stick with the fisheye - for shooting interior panos - and learn how to use it effectively. I suppose there's just a chance that you have a 'bad' one; there are reports of these lenses shipping with badly adjusted/calibrated focus settings.

..............

There is a Samyang branded version of this fisheye lens with a 'chip' to allow communication with the camera for metering:
http://www.dpreview.com/news/2011/11/25/samyang8nikon
Last edited by mediavets on Mon May 21, 2012 8:37 pm, edited 1 time in total.
Andrew Stephens
Many different Nodal Ninja and Agnos pano heads. Merlin/Panogear mount with Papywizard on Nokia Internet tablets.
Nikon D5100 and D40, Sigma 8mm f3.5 FE, Nikon 10.5mm FE, 35mm, 50mm, 18-55mm, 70-210mm. Promote control.

no avatar
jeradg
Member
 
Topic author
Posts: 99
Likes: 0 post
Liked in: 0 post
Joined: Tue Jan 17, 2012 6:31 pm
Info

by jeradg » Tue May 22, 2012 6:41 pm

klausesser wrote:
jeradg wrote:I switched last week to stop using fisheye because there is much less distortion and greater clarity without it.

Hi!

That´s a big misunderstanding. Fisheye-distortion is completely eliminated in a spherical pano.

There´s absolutely no difference in terms of distortion between a sphere shot with a fisheye and a sphere shot with, let´s say, 35mm or 85mm.

The only difference is resolution: you get about 3 GPx with a 85mm (264 images) lens on a 5D2 - and you get about 112 MPx on the same camera having 6 images using a fisheye.

best, Klaus

In my experience, and from my boss's perspective, I don't believe this is true. We have done side by side comparisons of the same rooms processed in the same way with APP and PTP. On the edges/corners of the viewable areas, there is still distortion.

You are correct about the resolution. This is an additional benefit, and actually the reason I switched away from the fisheye lens (because I read a post of yours a couple weeks ago pointing this out). With better resolution I can more precisely edit the image before adjusting the size of the final pano for the tour.

User avatar
klausesser
Member
 
Posts: 8836
Likes: 5 posts
Liked in: 64 posts
Joined: Mon May 22, 2006 12:18 am
Location: Duesseldorf, Germany
Info

by klausesser » Tue May 22, 2012 7:03 pm

jeradg wrote:
klausesser wrote:There´s absolutely no difference in terms of distortion between a sphere shot with a fisheye and a sphere shot with, let´s say, 35mm or 85mm.

In my experience, and from my boss's perspective, I don't believe this is true. We have done side by side comparisons of the same rooms processed in the same way with APP and PTP. On the edges/corners of the viewable areas, there is still distortion.

believe me: it definitely IS true. :cool: You always have distortion in the edges - that´s inavoidable and depends of the settings of viewing-angle. If you use 60° or 70° instead of 90° any distortions in the corners become fewer.

Remember: its up to you to choose the angle which provides the best over-all quality. 90° can´t do that - so use a lower angle setting in the editor or in the xml. That´s axactly like a superwide lens: in the outer areas distortions became extreme.
A lens with 90° distorts very heavily - and so does a viewer setting of 90°. The distortions of a 70° lens are much lower - and so it is with a viewersetting to 70°.

I did tours which have 70° instead of the default 90° constantly because the client wanted it less extreme.

See it here: http://www.sonne-frankenberg.de/de/13/home.aspx

jeradg wrote:You are correct about the resolution. This is an additional benefit, and actually the reason I switched away from the fisheye lens (because I read a post of yours a couple weeks ago pointing this out). With better resolution I can more precisely edit the image before adjusting the size of the final pano for the tour.

The resolution becomes higher - of course: you make more images . . :cool: - but in terms of distortions there´s no difference: a sphere is a sphere and remains a sphere (as a matter of fact you always see 6 cubefaces projected spherically) . . . . no matter wether you shoot is with a fisheye or a 300mm lens . . ;):cool: The amount of distortion which you see depends on how you set the viewing angle (FOV - field of view) in the xml.

best, Klaus
Simplicity is the keynote of all true elegance. Coco Chanel

no avatar
jeradg
Member
 
Topic author
Posts: 99
Likes: 0 post
Liked in: 0 post
Joined: Tue Jan 17, 2012 6:31 pm
Info

by jeradg » Tue May 22, 2012 7:07 pm

What aperture did you use and how were you focusing the lens?
5.0, using area auto focus and the camera's built in hdr in fine jpeg format.

If shooting panos then the fisheye distortion is accounted for when projecting the stitched pano image; so your comment about distortion doesn't really make any sense to me.
See above response to Klaus: The center is in focus, the corners are less focused (with fisheye).

Assuming you are using the 'kit' zoom lens at 18mm focal length then your shooting a pattern is quite inadequate. This is the minimum suggested for 360x180 coverage with your setup as listed here:
http://www.vrwave.com/panoramic-lens-database/nikon/
Minimum (1.5x): N(adir), 10 images every 36° at -45° pitch, 10 images every 36° at 0° pitch, 10 images every 36° at +45° pitch, Z(enith)
The camera is rotated counterclockwise, and I am only shooting cylindrical because the zenith and nadir are unnecessary for my project. With the current lens attached, the Ninja only allows it a 60° angle toward zenith (since I am achieving an adequate upward viewing angle at 15, the 60 is more than enough).

.............

Your problems stem from a number of factors I feel:
1. The nature of the scene. It would be so much easier to shoot this scene using your fisheye lens.
I agree, it would be much easier. I already did it shooting in fisheye because I knew it would be easiest that way (since I had no experience). Since completing the first phase, I began exploring options for improving quality before moving onto the second phase. My boss and others are much happier with the non-fisheye results in terms of quality.

2. You chosen shooting pattern.
The smallest horizontal increment on my pano head is 15°, so that it my chosen increment of change. I will attempt -20°, 0, 20° for vertical shots

3. A potential issue with this 'kit' zoom lens of zoom and focus creep at higher +/- pitch angles. At least that what I found with mine on a Merlin robotic head.
This is possible. I don't even know what it means exactly.

.............

I really think you'd be better advised to stick with the fisheye - for shooting interior panos - and learn how to use it effectively. I suppose there's just a chance that you have a 'bad' one; there are reports of these lenses shipping with badly adjusted/calibrated focus settings.
This is a possibility. My chosen shooting pattern with the fisheye may have been inadequate. I will attempt a couple of fisheye shots now that I have a little more knowledge than I initially did and see if I can get the quality to the same level as I have been shooting without it.

..............

There is a Samyang branded version of this fisheye lens with a 'chip' to allow communication with the camera for metering:
http://www.dpreview.com/news/2011/11/25/samyang8nikon
This would be nice, but I am limited by the fact that I don't have the ability to purchase the equipment. Luckily, this forum is full of helpful people to help me make the most of what I have.

no avatar
mediavets
Moderator
 
Posts: 16415
Likes: 2 posts
Liked in: 130 posts
Joined: Wed Nov 14, 2007 2:12 pm
Location: Isleham, Cambridgeshire, UK.
Info

by mediavets » Tue May 22, 2012 7:14 pm

jeradg wrote:In my experience, and from my boss's perspective, I don't believe this is true. We have done side by side comparisons of the same rooms processed in the same way with APP and PTP. On the edges/corners of the viewable areas, there is still distortion.

The science of optics determines that a stitched pano image using the same type of projection and having the same Field of View will have the same 'distortions' regardless of the camera and lens used to shoot the constituent images.

Of course if your boss chooses to believe otherwise (does he also believe the earth is flat?) and you wish to keep your job...
Andrew Stephens
Many different Nodal Ninja and Agnos pano heads. Merlin/Panogear mount with Papywizard on Nokia Internet tablets.
Nikon D5100 and D40, Sigma 8mm f3.5 FE, Nikon 10.5mm FE, 35mm, 50mm, 18-55mm, 70-210mm. Promote control.

no avatar
jeradg
Member
 
Topic author
Posts: 99
Likes: 0 post
Liked in: 0 post
Joined: Tue Jan 17, 2012 6:31 pm
Info

by jeradg » Tue May 22, 2012 7:23 pm

klausesser wrote:
jeradg wrote:
klausesser wrote:There´s absolutely no difference in terms of distortion between a sphere shot with a fisheye and a sphere shot with, let´s say, 35mm or 85mm.

In my experience, and from my boss's perspective, I don't believe this is true. We have done side by side comparisons of the same rooms processed in the same way with APP and PTP. On the edges/corners of the viewable areas, there is still distortion.

believe me: it definitely IS true. :cool: You always have distortion in the edges - that´s inavoidable and depends of the settings of viewing-angle. If you use 60° or 70° instead of 90° any distortions in the corners become fewer.

Remember: its up to you to choose the angle which provides the best over-all quality. 90° can´t do that - so use a lower angle setting in the editor or in the xml. That´s axactly like a superwide lens: in the outer areas distortions became extreme.
A lens with 90° distorts very heavily - and so does a viewer setting of 90°. The distortions of a 70° lens are much lower - and so it is with a viewersetting to 70°.

I did tours which have 70° instead of the default 90° constantly because the client wanted it less extreme.

See it here: http://www.sonne-frankenberg.de/de/13/home.aspx

jeradg wrote:You are correct about the resolution. This is an additional benefit, and actually the reason I switched away from the fisheye lens (because I read a post of yours a couple weeks ago pointing this out). With better resolution I can more precisely edit the image before adjusting the size of the final pano for the tour.

The resolution becomes higher - of course: you make more images . . :cool: - but in terms of distortions there´s no difference: a sphere is a sphere and remains a sphere (as a matter of fact you always see 6 cubefaces projected spherically) . . . . no matter wether you shoot is with a fisheye or a 300mm lens . . ;):cool: The amount of distortion which you see depends on how you set the viewing angle (FOV - field of view) in the xml.

best, Klaus

You're right about FOV, it does significantly affect distortion. I spent quite a bit of time doing line straightening in APP, particularly with the shots taken with fisheye, and it still appears to bend objects with the same FOV. Attached are 2 images of the same room.



no avatar
jeradg
Member
 
Topic author
Posts: 99
Likes: 0 post
Liked in: 0 post
Joined: Tue Jan 17, 2012 6:31 pm
Info

by jeradg » Tue May 22, 2012 7:27 pm

Here is the other... it was too large. Putting them in tours with the same FOV makes the differences obvious. As I said in an earlier post, it could be an inadequate number of shots with the fisheye?



User avatar
klausesser
Member
 
Posts: 8836
Likes: 5 posts
Liked in: 64 posts
Joined: Mon May 22, 2006 12:18 am
Location: Duesseldorf, Germany
Info

by klausesser » Tue May 22, 2012 7:33 pm

jeradg wrote:. . and I am only shooting cylindrical . .

and here´s the issue: a cylinder has a vertically straightened plane while it´s horizontally round. So the image/s are vertically plane-projected - and therefore heavy distorted. Shoot spherical and use a spherical projection and it gets much better. This has nothing to do with the kind of lens yoi use. It´s only related to the kind of projection.
But if you and your boss think to know more of the basics than a professional photographer like me and a long-time panorama-crack like Andrew: no problem at all . . ;):cool:

good luck, Klaus
Last edited by klausesser on Tue May 22, 2012 7:38 pm, edited 1 time in total.
Simplicity is the keynote of all true elegance. Coco Chanel

no avatar
jeradg
Member
 
Topic author
Posts: 99
Likes: 0 post
Liked in: 0 post
Joined: Tue Jan 17, 2012 6:31 pm
Info

by jeradg » Tue May 22, 2012 7:35 pm

mediavets wrote:
jeradg wrote:In my experience, and from my boss's perspective, I don't believe this is true. We have done side by side comparisons of the same rooms processed in the same way with APP and PTP. On the edges/corners of the viewable areas, there is still distortion.

The science of optics determines that a stitched pano image using the same type of projection and having the same Field of View will have the same 'distortions' regardless of the camera and lens used to shoot the constituent images.

Of course if your boss chooses to believe otherwise (does he also believe the earth is flat?) and you wish to keep your job...

I understand the science of optics is ultimately behind the situation, but the hardware/software (and its quality and manipulation - including user manipulation) determine outcomes as well. Like I said, I will go reattempt some fisheye shooting for experimentation, but my experience thus far has been that no amount of line straightening in APP has been as successful as shooting without the fisheye.

no avatar
jeradg
Member
 
Topic author
Posts: 99
Likes: 0 post
Liked in: 0 post
Joined: Tue Jan 17, 2012 6:31 pm
Info

by jeradg » Tue May 22, 2012 7:36 pm

klausesser wrote:
jeradg wrote:. . and I am only shooting cylindrical . .

and here´s the issue: a cylinder has a vertical straight plane while it´s horizontally round. So the image/s are vertically plane-projected - and therefore heavy distorted. Shoot spherical and use a spherical projection and it gets much better. This has nothing to do with the kind of lens yoi use. It´s only related to the kind of projection.
But if you and your boss think to know more of the basics than a professional photographer like me and a long-time panorama-crack like Andrew: no problem at all . . ;):cool:

good luck, Klaus

So including the zenith/nadir will decrease the distortion for the entire pano?
Edit: Never claimed to know more. I come here because of your willingness to share your knowledge/experiences with the inexperienced users such as myself. I am only reporting what my eyes are telling me I am seeing as a product of my own work.
Last edited by jeradg on Tue May 22, 2012 7:40 pm, edited 1 time in total.

no avatar
mediavets
Moderator
 
Posts: 16415
Likes: 2 posts
Liked in: 130 posts
Joined: Wed Nov 14, 2007 2:12 pm
Location: Isleham, Cambridgeshire, UK.
Info

by mediavets » Tue May 22, 2012 7:40 pm

jeradg wrote:What aperture did you use and how were you focusing the lens?
5.0, using area auto focus and the camera's built in hdr in fine jpeg format.

No, no, no...f5 is not good with that lens. And it's a fully manual lens it has to be manually focussed you cannot use autofocus; in fact it's never desiranble to use autofocus when shooting panos, the conventional approach to pano shooting is manual everything (white balance, shutter speed, aperture, focus) and for top quality to shoot RAW. 'Built-in HDR'? - do you mean AEB, automatic exposure bracketing?

If shooting panos then the fisheye distortion is accounted for when projecting the stitched pano image; so your comment about distortion doesn't really make any sense to me.
See above response to Klaus: The center is in focus, the corners are less focused (with fisheye).

I think that's because you didn't focus the lens correctly. Fisheye lenses have massive depth of field if focussed correctly; and your lens has a good reputation for sharpness.

Assuming you are using the 'kit' zoom lens at 18mm focal length then your shooting a pattern is quite inadequate. This is the minimum suggested for 360x180 coverage with your setup as listed here:
http://www.vrwave.com/panoramic-lens-database/nikon/
Minimum (1.5x): N(adir), 10 images every 36° at -45° pitch, 10 images every 36° at 0° pitch, 10 images every 36° at +45° pitch, Z(enith)
The camera is rotated counterclockwise, and I am only shooting cylindrical because the zenith and nadir are unnecessary for my project. With the current lens attached, the Ninja only allows it a 60° angle toward zenith (since I am achieving an adequate upward viewing angle at 15, the 60 is more than enough).

Oh dear, if you only need a pano with a restricted Vertical Field Of View (BTW Panotour/Panotour Pro does not accept images with a cylindrical projection as input) then a single row shot at zero pitch with the camera in portrait orientation using your fisheye lens; 6-around, shooting every 60 degrees on the yaw axis should do the job easily. te goal is to have about 25-30% overlap between images - more is likley to cause problems when stitching, less and you may not be able to link adjacent images.

I really think you'd be better advised to stick with the fisheye - for shooting interior panos - and learn how to use it effectively. I suppose there's just a chance that you have a 'bad' one; there are reports of these lenses shipping with badly adjusted/calibrated focus settings.
This is a possibility. My chosen shooting pattern with the fisheye may have been inadequate. I will attempt a couple of fisheye shots now that I have a little more knowledge than I initially did and see if I can get the quality to the same level as I have been shooting without it.

Let me see if I understand your situation...you haven't a clue what your doing, you didn't choose the equipment, you've had no training but your boss expects you to produce top quality results?

I've had bosses like that in the past; I think I still bear the psychological scars.

What I don't understand is why you seem so reluctant to accept the good advice being offered by people with considerable experience - Klaus is a professional photographer!


There is a Samyang branded version of this fisheye lens with a 'chip' to allow communication with the camera for metering:
http://www.dpreview.com/news/2011/11/25/samyang8nikon
This would be nice, but I am limited by the fact that I don't have the ability to purchase the equipment. Luckily, this forum is full of helpful people to help me make the most of what I have.

Who did choose the equipment and why did they choose this setup?

A fully manual fisheye lens is not really an ideal choice for a novice photographer. The Nikkor 10.5mm fisheye would have been easier for you to handle.
Last edited by mediavets on Tue May 22, 2012 7:43 pm, edited 1 time in total.
Andrew Stephens
Many different Nodal Ninja and Agnos pano heads. Merlin/Panogear mount with Papywizard on Nokia Internet tablets.
Nikon D5100 and D40, Sigma 8mm f3.5 FE, Nikon 10.5mm FE, 35mm, 50mm, 18-55mm, 70-210mm. Promote control.

User avatar
klausesser
Member
 
Posts: 8836
Likes: 5 posts
Liked in: 64 posts
Joined: Mon May 22, 2006 12:18 am
Location: Duesseldorf, Germany
Info

by klausesser » Tue May 22, 2012 7:41 pm

jeradg wrote:
klausesser wrote:
jeradg wrote:. . and I am only shooting cylindrical . .

and here´s the issue: a cylinder has a vertical straight plane while it´s horizontally round. So the image/s are vertically plane-projected - and therefore heavy distorted. Shoot spherical and use a spherical projection and it gets much better. This has nothing to do with the kind of lens yoi use. It´s only related to the kind of projection.
But if you and your boss think to know more of the basics than a professional photographer like me and a long-time panorama-crack like Andrew: no problem at all . . ;):cool:

good luck, Klaus

So including the zenith/nadir will decrease the distortion for the entire pano?

If you don´t use a cylindrical but spherical projection: yes. But - like said before: it depends on the FOV you coose: take 60° or 70° instead of 90° if you feel the distortions are too much.

best, Klaus
Simplicity is the keynote of all true elegance. Coco Chanel

no avatar
jeradg
Member
 
Topic author
Posts: 99
Likes: 0 post
Liked in: 0 post
Joined: Tue Jan 17, 2012 6:31 pm
Info

by jeradg » Tue May 22, 2012 8:34 pm

Just to be clear, my shots and set up with the standard lens were experimental, but ended up better and easier than the use of the fisheye even though it took longer to shoot.

No, no, no...f5 is not good with that lens. And it's a fully manual lens it has to be manually focussed you cannot use autofocus; in fact it's never desiranble to use autofocus when shooting panos, the conventional approach to pano shooting is manual everything (white balance, shutter speed, aperture, focus) and for top quality to shoot RAW. 'Built-in HDR'? - do you mean AEB, automatic exposure bracketing?
Menu --> Shooting Menu --> HDR = how I was selecting HDR. It would take one at the selected shutter speed and another longer speed and use internal software to create the HDR jpeg. This setting was not compatible with RAW. I have been using everything manual except for the focus, and that HDR setting choosing the 2nd shutter.

I think that's because you didn't focus the lens correctly. Fisheye lenses have massive depth of field if focussed correctly; and your lens has a good reputation for sharpness.
This is probably true. Should I leave the same focus setting throughout the shoot or change it based on distance from objects/walls?


Oh dear, if you only need a pano with a restricted Vertical Field Of View (BTW Panotour/Panotour Pro does not accept images with a cylindrical projection as input) then a single row shot at zero pitch with the camera in portrait orientation using your fisheye lens; 6-around, shooting every 60 degrees on the yaw axis should do the job easily. te goal is to have about 25-30% overlap between images - more is likley to cause problems when stitching, less and you may not be able to link adjacent images.
I actually checked the coverage/overlap of the pictures with objects to ensure approx 1/3 - with the standard. When I first started, using the fisheye, I was shooting 4 (every 90°), which is probably a large part of the problem. I stuck with shooting 4 horizontal, 1 zenith, 1 nadir because I was getting well stitched results in APP. I would then delete the nadir because it had the tripod in it (and I would be cropping anyways).

Let me see if I understand your situation...you haven't a clue what your doing, you didn't choose the equipment, you've had no training but your boss expects you to produce top quality results?
All true. Luckily, we have defined top quality results as making a best effort with our in-house production of a tour that we can have continued control over as our facilities change. 5 months @ <10 hrs/week = a lot of progress and even more learning. The overall project cost including my pay is still less than we were quoted by some companies, all of which would require travel expenses since there are no tour companies local. Despite all my issues, which you now see are numerous, my results have still been better than some "professional" companies.

What I don't understand is why you seem so reluctant to accept the good advice being offered by people with considerable experience - Klaus is a professional photographer!
I am accepting the advice. You're overlooking my thankful statements and jumping to the meat of my problems (which is helping to speed the resolution of them and my learning).

Who did choose the equipment and why did they choose this setup?
A fully manual fisheye lens is not really an ideal choice for a novice photographer. The Nikkor 10.5mm fisheye would have been easier for you to handle.
The camera is one we already had, and the fisheye lens was suggested online as being ideal for ease of use and indoor panotours. My boss found the actual lens, and I found PTP/APP. I think I have identified many of my shooting problems through this one thread, and wish the discourse could have happened 4 months ago. I think that 4 months ago I wouldn't have been able to find the settings on the camera to change or known enough about everything to understand it all.


So last questions before I run off to attempt what I've learned so far today: this lens has focal range from 0.3m to 3m to infinite; should I be adjusting it each shot or leaving it constant for an entire set for a pano? The hallway pano is the closest-quarters shot I have to do, and the room with the wood flooring/couches is the largest. Just in looking at a couple of focus-adjusted shots I just took, I can't see that there is a huge difference in each image. Which direction would you think I should lean (0.3 - infinite) in each of the 2 shots mentioned?

Klaus, I misspoke, I said cylindrical because I meant the FOV is limited to 360 x 65 (with standard lens) and 360 x 130 (with fisheye). The projection is still spherical (I just checked the .pano file).

no avatar
jeradg
Member
 
Topic author
Posts: 99
Likes: 0 post
Liked in: 0 post
Joined: Tue Jan 17, 2012 6:31 pm
Info

by jeradg » Tue May 22, 2012 8:35 pm

And back to the original topic... Is there any way to spread out the images and increase the width within APP?

no avatar
jeradg
Member
 
Topic author
Posts: 99
Likes: 0 post
Liked in: 0 post
Joined: Tue Jan 17, 2012 6:31 pm
Info

by jeradg » Tue May 22, 2012 9:05 pm

Are these the correct image property settings for my setup?

I don't know how to manually set the aperture in APP, but I took these with 3.5.



User avatar
klausesser
Member
 
Posts: 8836
Likes: 5 posts
Liked in: 64 posts
Joined: Mon May 22, 2006 12:18 am
Location: Duesseldorf, Germany
Info

by klausesser » Tue May 22, 2012 9:50 pm

jeradg wrote:Are these the correct image property settings for my setup?

I don't know how to manually set the aperture in APP, but I took these with 3.5.

I don´t know the camera and the lens - but i bet the focal-length is not 1,04mm . . . :cool:

best, Klaus
Simplicity is the keynote of all true elegance. Coco Chanel

no avatar
mediavets
Moderator
 
Posts: 16415
Likes: 2 posts
Liked in: 130 posts
Joined: Wed Nov 14, 2007 2:12 pm
Location: Isleham, Cambridgeshire, UK.
Info

by mediavets » Tue May 22, 2012 11:27 pm

jeradg wrote:Are these the correct image property settings for my setup?

No, settings should be like this:






Andrew Stephens
Many different Nodal Ninja and Agnos pano heads. Merlin/Panogear mount with Papywizard on Nokia Internet tablets.
Nikon D5100 and D40, Sigma 8mm f3.5 FE, Nikon 10.5mm FE, 35mm, 50mm, 18-55mm, 70-210mm. Promote control.

Next

Who is online

Users browsing this forum: No registered users and 2 guests