 |
|
|
|
|
|
 |
| ||||||||
 |
 
|
|
|
|
|
|
|
|
|
| ||||||||
|
|
| 
| User list | Rules | You are not logged in.
I am still trying to decide which pano head to buy to use with my Nikon D40 and Nikkor 10.5 fisheye.
I am attracted to the new Nodal Ninja 5 Lite on basis of features and price but I have been unable to find sets of sample images taken using the Nodal Ninja 3, or the new 5, with any sort of fisheye lens.
What I am trying to find out is whether the bottom rail or vertical rail appear in shots taken at zero degrees tilt with fisheye - eiither full circular or full-frame - when using either the Nodal Ninja 3 or 5 pano heads.
Obviously the bottom rail will be seen in 90-degree down shots.
How do others recommend handling the down (nadir) shots.
I note that Karl Harrison - Oxford Panoramas - shooting with a 360Precison head takes two 90-degree down shots rotating the head 180 degrees between the two shots.
But I find that I still have ghosting of the bottom rail after stitching with APP with Smartblend using his sample images shot with Canon 5D and 15mm Canon full circular fisheye.
I want to do a minimum of post-processing so I would prefer a head that did not show the bottom or vertical rails with fisheye lenses.
I think one can avoid the issue with an Agno's MrotatorTCS/TCP head because the bottom rail is offset to the rear on Agno's heads, as opposed to being in line with the axis of rotation which appears to be the case with most (all?) other brands of pano head.
So which head to choose to avoid additional post-processing?
What is the optimum workflow to handle the appearance of the bottom rail in 90-degree down shots for nadir? I don't mind seeing the small circular top of the rotator base and perhaps some of the tripod legs in a finished pano at this stage, but I do not want ghosted appearnace of the botton rail nor do I want issues arising from capturing the vertical rail in shots either.
Andrew
Last edited by mediavets (2008-02-08 14:53:31)
Offline
mediavets wrote:
I am still trying to decide which pano head to buy to use with my Nikon D40 and Nikkor 10.5 fisheye.
I am attracted to the new Nodal Ninja 5 Lite on basis of features and price but I have been unable to find sets of sample images taken using the Nodal Ninja 3, or the new 5, with any sort of fisheye lens.
What I am trying to find out is whether the bottom rail or vertical rail appear in shots taken at zero degrees tilt with fisheye - eiither full circular or full-frame - when using either the Nodal Ninja 3 or 5 pano heads.
Obviously the bottom rail will be seen in 90-degree down shots.
How do others recommend handling the down (nadir) shots.
Full circular - shots taken at zero degrees : bottom rail and small circular top of the rotator base are certainly visible due to 180 degree FOV. Work arround : direct the camera 5 or 10 degrees upward so that they are no more visible (benefit : the zenith coverage is much better.)
Rectangular image fisheye - shots taken at zero degrees : nothing visible when using a NN3.
Interestingly enough, for rectangular image fisheye and shots taken at 15 degrees toward the ground, the vertical rail is visible in the corner but APP/Smartblend always ignores it...
Here is a face cube from a pano I shot using NN3, a 16 mm equivalent rectangular fisheye and where the camera was 15° toward the ground (tripod legs were recessed.)
This size of "nadir hole" is easy to fill using APP and a hand-held nadir shot. My intention is to write a tutorial about nadir and zenith. Some tips : use APP "circular crop" with a small diameter (just enough to cover the hole) and use APP "local optimize" feature to avoid disturbing the whole pano because nadir shot was hand-held.)
Offline
GURL,
Thanks so much for that info. And for the sample cube face image.
What is the diagonal FOV of your 16mm fullframe fisheye on your camera? That is an Olympus E330 with a 4/3 if I recall correctly from your pother posts, so I'm not sure how results with the NN3 (or NN5)would compare with the 1.6 crop sensor of my D40 using the Nikkor 10.5mm. Would I be more or less likley to see bits of the bottom and vertical rails in portrait orientation full-frame fisheye shots do you think?
Why is the bootom rail not visible in your stitched cube face image?
I imagine it was visible in each of the original shots before stitching?
So does APP/Smartblend ignore it?
I think the bottom rail would not appear in full circular images when using an Agno's T series head - at zero degrees - even with 180 degrees vertical FOV because their design offsets the bottom rail behind the centre of rotation. I wonder why Nodal Ninja did not adopt that idea for the NN5.
It certainly is not visible - nor is vertical rail because there isn't one - with their MrotatorC head for the Sigma 8mm fisheye..
And of course their is no bottom or vertical rail if one uses their RingT head - although that offers no vertical rotation unless mounted on a T series bracket and that is an expensive combination.
Andrew
Offline
Here's the response I got from Hans Nyberg on Panoguide - it may be of interest to others:
"To avoid interference from the rotator or bottom rail you just shoot the 6 around at -10 or 12 degrees.
This means you just shoot 6 around + the zenith.
There is still enough overlap for the zenith image and you do not shoot any nadir at -90 degrees.
around -10 degrees gets you everything you can see with a fullframe fisheye.
Just tilt down until you can see the rotator at the bottom.
The missing part for the nadir is very small and can in most cases be done by cloning but you can of course also shoot a nadir handheld or using other methods for example using viewpoint in PTGui Pro."
Andrew
Offline
mediavets wrote:
What is the diagonal FOV of your 16mm fullframe fisheye on your camera?
180 degrees (a 8 mm "4/3" lens is equivalent to a 16 mm "FF" lens.)
mediavets wrote:
I'm not sure how results with the NN3 (or NN5)would compare with the 1.6 crop sensor of my D40 using the Nikkor 10.5mm. Would I be more or less likley to see bits of the bottom and vertical rails in portrait orientation full-frame fisheye shots do you think?
10.5 x 1.6 = 16 mm equiv. (Nikon) - - - 8 x 2 = 16 mm equiv. (Olympus) - - - 16 x 1 = 16 mm "true" (a full-frame Canon 1D and a 16 mm "rectangular" fisheye for example)
Well, both "Nikon D40 + 10.5mm" and "E-330 + 8mm" combos have a 180° diagonal FOV. Olympus have a somewhat more squarish width/height ratio: less of bottom arm but more of vertical arm should be visible when using Olympus fisheye.
The main point: lot of peoples use Nikon APS-C cameras, Nikon 10.5 mm fisheye and NN3 and are happy with them. For any problem between NN5 and this 10.5 mm fisheye you should scan www.panoguide.com but I highly doubt Nickfan (NNx maker) would launch a pano-head which don't work well with the most appraised fisheye for spherical panos...
mediavets wrote:
Why is the bootom rail not visible in your stitched cube face image?
I imagine it was visible in each of the original shots before stitching?
So does APP/Smartblend ignore it?
It was visible in the bottom right corner of each downward shots but APP and Smartblend are wise enough to avoid using fisheye shot corners when this is possible. I shoot 6 views for each series (a "series" have an identical pitch value, downward or upward) because I found 4 views is not enough (sometimes 4 views works, sometimes it don't) and because large CF cards are now very inexpensive (it would be quite easy to defend the idea that the less you are using image corners the better is the pano.)
mediavets wrote:
I think the bottom rail would not appear in full circular images when using an Agno's T series head - at zero degrees - even with 180 degrees vertical FOV because their design offsets the bottom rail behind the centre of rotation. I wonder why Nodal Ninja did not adopt that idea for the NN5.
It certainly is not visible - nor is vertical rail because there isn't one - with their MrotatorC head for the Sigma 8mm fisheye..
And of course their is no bottom or vertical rail if one uses their RingT head - although that offers no vertical rotation unless mounted on a T series bracket and that is an expensive combination.
I studied Agnos pano heads "on the papers" but never had a chance to use them. There are some Agnos users on the Panoguide forum: don't be afraid to ask...
The most interesting results I saw from Agnos users : panos where a pole is used. Besides that, most of the stuff they now sell is new: it looks more "straightforward" than it was.
Offline
GURL wrote:
The main point: lot of peoples use Nikon APS-C cameras, Nikon 10.5 mm fisheye and NN3 and are happy with them. For any problem between NN5 and this 10.5 mm fisheye you should scan www.panoguide.com but I highly doubt Nickfan (NNx maker) would launch a pano-head which don't work well with the most appraised fisheye for spherical panos....
Thanks GURL for your very helpful comments and advice.
I have ordered the Nodal Ninja 5 Lite.
I am aware that the NN5 is 'overkill' for my current D40 camera/10.5mm fisheye lens - an NN3 would be fine for that - but the NN5 provides greater scope for other camera/lens combinations (who knows I may win the lottery) and is not too much heavier than the NN3 (although more expensive) and about the same price but still lighter than any equivalent Agno's head.
Andrew
Offline