The GetDPI Photography Forum
Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!
Determining nodal point of lens for panos
- Thread starter cerett
- Start date
stephengilbert
Active member
k-hawinkler
Well-known member
Don Libby
Well-known member
The RRS video shows how to find the nodal point using two straight objects then marking the point where no change occurs. You can set this up at any time prior to actually shooting the pano in the wild.
As far as how many are actually doing this is a difficult question to answer.
Most of my panoramas are from a series of flat stitching with the tech camera. I will occasionally try shooting a pano with my DF and in that case it's almost always spur of the moment handheld. When this happens I shoot left to right and attempt to keep the camera as level as possible overlapping about 50% for each image taken. I do this because I know from experience that once the files are joined together I'll have what I call a bowtie effect and will simply crop to suit what I want the finished image to look like. Not perfect however it'll do in a pinch when I don't have the time to shoot with the tech camera or it's a spur of the moment decision.
As far as how many are actually doing this is a difficult question to answer.
Most of my panoramas are from a series of flat stitching with the tech camera. I will occasionally try shooting a pano with my DF and in that case it's almost always spur of the moment handheld. When this happens I shoot left to right and attempt to keep the camera as level as possible overlapping about 50% for each image taken. I do this because I know from experience that once the files are joined together I'll have what I call a bowtie effect and will simply crop to suit what I want the finished image to look like. Not perfect however it'll do in a pinch when I don't have the time to shoot with the tech camera or it's a spur of the moment decision.
Thanks guys. As far as how to determine the nodal point, I think I understand that fairly well although have not done it yet. I have used my Canon 24mm TS-E with great success, but the nodal point was not an issue with that lens. I will try some images with my H5D with and without determining the nodal point and see what difference it makes.
AreBee
Member
Cerett,
It all depends on how rigorous you wish to be, and/or the risk you are prepared to accept of an error in the stitch (for example, are you at a location you will never return to, and therefore cannot afford for the panorama to stitch incorrectly?).
The rigorous approach is to:
1. Shoot the lens around its entrance pupil. This will eliminate parallax error between neighbouring images and enable the stitching software to overlap them exactly.
2. Place control points (which link common features on neighbouring images) manually. This will eliminate the possibility of misplaced control points placed by the stitching software.
Both are required in order to minimise the likelihood of an error in the stitch. With regard to point 1, the greater parallax error is, the greater will be the mismatch between common features in neighbouring images, and therefore the greater the risk that the stitching software's blending algorithm fails to conceal the error. With regard to point 2, it is not uncommon for stitching software to misplace control points. Inaccurate placement will return an inaccurate stitch, which again will increase the risk that the stitching software's blending algorithm fails to conceal the error.
There are many photographers who shoot panoramas that ignore point 1, 2, or 1 and 2, and are entirely satisfied with the stitch. There is no right or wrong - simply what the photographer considers correct for them.
Hope this helps.
It all depends on how rigorous you wish to be, and/or the risk you are prepared to accept of an error in the stitch (for example, are you at a location you will never return to, and therefore cannot afford for the panorama to stitch incorrectly?).
The rigorous approach is to:
1. Shoot the lens around its entrance pupil. This will eliminate parallax error between neighbouring images and enable the stitching software to overlap them exactly.
2. Place control points (which link common features on neighbouring images) manually. This will eliminate the possibility of misplaced control points placed by the stitching software.
Both are required in order to minimise the likelihood of an error in the stitch. With regard to point 1, the greater parallax error is, the greater will be the mismatch between common features in neighbouring images, and therefore the greater the risk that the stitching software's blending algorithm fails to conceal the error. With regard to point 2, it is not uncommon for stitching software to misplace control points. Inaccurate placement will return an inaccurate stitch, which again will increase the risk that the stitching software's blending algorithm fails to conceal the error.
There are many photographers who shoot panoramas that ignore point 1, 2, or 1 and 2, and are entirely satisfied with the stitch. There is no right or wrong - simply what the photographer considers correct for them.
Hope this helps.
stephengilbert
Active member
"I will try some images with my H5D with and without determining the nodal point and see what difference it makes."
This will involve more effort than simply determining the nodal point. Once you know the nodal point for a lens, what reason would there be to ignore it when shooting later?
This will involve more effort than simply determining the nodal point. Once you know the nodal point for a lens, what reason would there be to ignore it when shooting later?
hasselblad actually provides the location of the nodal point for it's lenses on it's information site. (location with respect to film/sensor plane)
what i have done is to mark my RRS nodal slide for each relevant V-series lens i have so i know where to position the slide with respect to pano rotation.
by the way, the nodal point is not the entrance pupil, but the virtual point about which rotaion can occur without parallax.
i have found Autopano giga to give better results than photoshop for merging without bowties.
and i shoot right to left (if i can remember), because then capture 1 places the files in the viewer in the natural order (last shot will be to the left)
what i have done is to mark my RRS nodal slide for each relevant V-series lens i have so i know where to position the slide with respect to pano rotation.
by the way, the nodal point is not the entrance pupil, but the virtual point about which rotaion can occur without parallax.
i have found Autopano giga to give better results than photoshop for merging without bowties.
and i shoot right to left (if i can remember), because then capture 1 places the files in the viewer in the natural order (last shot will be to the left)
Shashin
Well-known member
I have shot a lot of landscape panos without swinging around the nodal point. I do a lot handheld also. Those have worked really well. The only time it becomes an issue is if you have a complex scene with foreground and background elements, like shooting in the middle of a forest. Then swinging the camera through the nodal point is important. I simply carry a sliding macro rail I can mount on the top of my P0 head when I need to use the nodal point.
Shashin
Well-known member
To travel light. Pano heads and setups can be large and heavy.
Thank you for the tip. This is what I got for my HC 80mm:
"ENTRANCE PUPIL POSITION: 79 mm in front of the film plane (at infinite focus setting). The entrance pupil position is the correct position of the axis of rotation when making a panorama image by combining individual images of a scene."
Now, can you explain, based on this, where to position my camera on the nodal rail?
No reason other than curiosity.
i use the RRS pano clamp and the RRS nodal rail.
i clamped the camera to the nodal rail and measured from the film plane index mark the stated amount and placed a mark on the side of the nodal rail (for each lens).
then slide the nodal rail in the pano clamp aligning the mark you just made with the center of rotation mark for the pano clamp
i clamped the camera to the nodal rail and measured from the film plane index mark the stated amount and placed a mark on the side of the nodal rail (for each lens).
then slide the nodal rail in the pano clamp aligning the mark you just made with the center of rotation mark for the pano clamp
Does anyone know if Mamiya/Phase One has a similar information table? That would be very useful.
I'm not usually inclined to do stitching but I recently did a three panel stitch with an inaccurate "nodal point" (i.e. I placed the aperture plane approximately at the center of rotation on the tripod). FWIW basically the only place where I could see a defect was a slight mis-aligning of bricks at a couple of spots (attached) and even then it wasn't terribly objectionable. The attached image is an 83 pixel wide crop of a 14,508 pixel image. So less than 1/3 inch in a final 300dpi print.
I've done pano's with and without using the nodal point. The "best" method will depend on what subject matter you are shooting, how much time you have to set up your shot, the gear you have with you, what focal length you're shooting, etc. I use different techniques depending on the mood I'm in. Also things like weather come into play...if it's really cold or raining for example, I usually ditch the nodal rail and use whatever technique is simplest. I am of the opinion it's good to practice panos handheld, simply with panning, and also with a nodal rail and know how to get the best results with each technique in your own hands.
The RRS video is a good one and anyone trying to determine the nodal point could probably do it quicker just by following their instructions rather than trying to decipher info from the manufacturer.
Lately I've just been using the RRS panning head without the rail (simply because I'm lazy) and my results for nature/landscape (where subject matter is at infinity for the most part) are acceptable to me. I find I travel with the rail and then never end up using it.
The RRS video is a good one and anyone trying to determine the nodal point could probably do it quicker just by following their instructions rather than trying to decipher info from the manufacturer.
Lately I've just been using the RRS panning head without the rail (simply because I'm lazy) and my results for nature/landscape (where subject matter is at infinity for the most part) are acceptable to me. I find I travel with the rail and then never end up using it.
Shashin
Well-known member
The aperture is usually somewhere between the two nodal points (S and S') in a lens, but never at them. But you can simply find the nodal point by swing the camera and seeing if the horizontal distance between a near object and a far object changes. If it does, you are not at the right position. Adjust the camera until that parallax shift no longer occurs.
jotloob
Subscriber Member
A bit more comfortable than sliding the rail is a geared rail in addition .
AFAIK RRS has something similar in their product portfolio .
I have a little table stored on my iPHONE with all NPP for my HASSELBLAD lenses and use a little folding meterstick to do the proper adjustment .
Image taken before my SWC QR-plate modification and changing the QR-clamp to RRS gear .
