Getting the mounting right [Archive] - Nodal Ninja Forum

PDA

View Full Version : Getting the mounting right



BaltimoreBob
04-14-2012, 05:17 AM
My mind always searches for concise general rules from which all details will flow.

Have read a multitude of articles about how to mount to avoid parallax and all have dealt with the whys and the how to's. I've concluded two general rules which I think will apply with both of my heads - the NN5 and the M1-L, and maybe others as well.

They are:
1. The position at which the lower rail clamps to the tripod mount is dependent solely upon the camera body and the manor and parts attaching it to the upper rail (nadir adaptor or not, camera plate, clamp attaching the body to the rail, where the vertical rail is attached to the lower rail, etc.). Any lens may be used to acquire this positioning.
2. The position at which the camera is stopped on the upper rail is governed solely by the lens and the focal lenght to be used for this panorama.

Please - tell me what I got wrong.

Thanks

Vincèn
04-14-2012, 05:26 AM
You are nearly perfectly right excepted on these details:
-> nadir adapter has no effect on settings ;)
-> on upper rail it's not focal length but more exactly the position of the nodal point of lens used and zoom level on that lens if it's a zoom lens !

Vincèn

John Houghton
04-14-2012, 06:00 AM
In the case of fisheye lenses particularly, another factor in the upper rail setting is the angle of yaw increment used in taking the successive shots. This is because the apparent position of the entrance pupil varies for light rays entering the camera at different angles to the lens axis. (As is now well understood, it is the movement of the entrance pupil that leads to parallax shifts rather than the nodal points, which can be a long way from the entrance pupil position). So if you shoot 6 shots around instead of 4, say, then there will be a small difference in the optimum setting.
.
John

BaltimoreBob
04-14-2012, 06:55 AM
You are nearly perfectly right excepted on these details:
-> nadir adapter has no effect on settings ;)
-> on upper rail it's not focal length but more exactly the position of the nodal point of lens used and zoom level on that lens if it's a zoom lens !

Vincèn
Vince'n
I think we are saying the same thing -
if I'm using, say Nikon's 70-200, the upper rail setting may be different at a focal length/zoom level of 70mm than it will be at 200mm...
Bob

BaltimoreBob
04-14-2012, 07:05 AM
In the case of fisheye lenses particularly, another factor in the upper rail setting is the angle of yaw increment used in taking the successive shots. This is because the apparent position of the entrance pupil varies for light rays entering the camera at different angles to the lens axis. (As is now well understood, it is the movement of the entrance pupil that leads to parallax shifts rather than the nodal points, which can be a long way from the entrance pupil position). So if you shoot 6 shots around instead of 4, say, then there will be a small difference in the optimum setting.
.
John
John
two things occur to me -
1. with the fisheye, greater overlap is most beneficial - say 60 degrees (6 shots) rather than 90 degrees (4 shots)
2. mounting the fisheye to position is done in the same way as any other lens - it is, agan, the yaw/overlap increment when rotating the camera that is paramont
Bob

hindenhaag
04-14-2012, 07:56 AM
Bob,
The URS = upper rail setting consist of two different measurements: entrance pupil of lens to lens flange - L2 in wikipedia entrance pupil database - and the distance from lens flange to tripod mount screw of the camera body you use - L1.

http://wiki.panotools.org/Entrance_Pupil_Database Lower rail setting is H plus the offset caused by camera plates and reference point of vertical rail to bottom of camera body. NN3 NN4 NN5 this is 13mm.

Bob, 70 - 200 is a special lens. The settings for D3 NN5 are @zoom 70 52mm, @zoom 200 135 with the upper rail turned forward by 180° from normal position showing to viewfinder, from zoom 105 it has to look to the opposite side move to the front of lower rail.

590588589 D3 14-24 with changing Entrance pupil at different focus settings.

Heinz

BaltimoreBob
04-14-2012, 09:44 AM
Bob,

The URS = upper rail setting consist of two different measurements: entrance pupil of lens to lens flange - L2 in wikipedia entrance pupil database - and the distance from lens flange to tripod mount screw of the camera body you use - L1.

http://wiki.panotools.org/Entrance_Pupil_Database Lower rail setting is H plus the offset caused by camera plates and reference point of vertical rail to bottom of camera body. NN3 NN4 NN5 this is 13mm.

Bob, 70 - 200 is a special lens. The settings for D3 NN5 are @zoom 70 52mm, @zoom 200 135 with the upper rail turned forward by 180° from normal position showing to viewfinder, from zoom 105 it has to look to the opposite side move to the front of lower rail.


Heinz

Heinz - Hi

Re the lower rail setting - the following descriptions are of what I have physically done and do and are what I take to be what you have said:
... with NN5 I'm using the Live view of both my D3 and D300. I moving the vertical rail to center the Nodal Ninja logo in the center of the live view focus box - Bye the way that mounting system you came up with for me is a jewel - thanks!
... with the M1-L I'm again using the live view to center - the rotator has several East-West markings on each side of the clamping bars - one centered on the clamp knob which I take to be the center point marking. The lower rail has a gap for mounting screws running for most of its length - I estimate the center point of that gap. My vertical rail is mounted at the far end of the lower rail away from the bubble where it stays, as instructed; my camera is mounted on the upper rail and is pointing straight down. I place a bar above the rail the edge of which is set to the East-West center points on both sides of the Arca clamp. I then move the lower rail forward until the centered edge of that bar and the center of the screw gap are at the center of the live view focus box.

Thanks for the 70-200 settings for the NN5. So far I've only used it with my D3 on the M1-L. The setting I first arrived at is 15 for all focal lengths; I say "...first arrived..." as I need to re-check several times to make sure I can duplicate - my two pair of glasses leave me with gaps for some things and this is one of those things

Bob

hindenhaag
04-14-2012, 10:20 AM
Sorry, took a long time to upload the pics.

The way you find your LRS is a "first step". I always use Smooth method of circle saw teeth. You take shots around at -15° with f22 focussed to the corner of rotator. Then you go to PTGui, stitch, and move to the nadir point in .mov file to check for teeth. Cutting to the right, move to the right, cutting to the left, move to the left on lower rail. This method compensates Irregulars of equipment too. This means "displacement of sensor", differences in production of lenses etc.

http://www.easypano.com/forum/display_topic_threads.asp?ForumID=1&TopicID=4162

Are you using M1-L without Nadir Adapter? What lenses do you use? I'll send you the settings for D300, D3 M1-L, NN5.

Heinz

DemonDuck
04-14-2012, 10:32 AM
I check the parallax at the corners of the fisheye. That is to say, I tilt the camera down then look at the parallax of my alignment objects near the upper corners. I think that checking the parallax from upper right to lower left corner and from upper left to lower right corner might be even a better way because stitching errors due to parallax are frequent between middle row and nadir.

This for a full frame or 180 diagonal fisheye. Probably not useful for cropped or circular fisheye.

BaltimoreBob
04-14-2012, 11:37 AM
Sorry, took a long time to upload the pics.

The way you find your LRS is a "first step". I always use Smooth method of circle saw teeth. You take shots around at -15° with f22 focussed to the corner of rotator. Then you go to PTGui, stitch, and move to the nadir point in .mov file to check for teeth. Cutting to the right, move to the right, cutting to the left, move to the left on lower rail. This method compensates Irregulars of equipment too. This means "displacement of sensor", differences in production of lenses etc.

http://www.easypano.com/forum/display_topic_threads.asp?ForumID=1&TopicID=4162

Are you using M1-L without Nadir Adapter? What lenses do you use? I'll send you the settings for D300, D3 M1-L, NN5.

Heinz
The "smooth method" - I've done all of that -
M1-L without Nador Adaptor - for the moment - I'm going to order one and I know that changes the positioning of the vertical rail.
What lenses do I use:
I've pretty much been reserving my D300 with the NN5 for 360/180 using the Nikon 10.5 unshaved fisheye
Use my D3 with the M1-L and have been experimenting with various lenses - so far have used the 70-200, 80-400, 28-300, 24-70. Going to put my 70-300 and my 14-24 to use, thinking about my 16mm Fisheye also - all Nikon lenses by the way.
I use the 300 with the fisheye because of the crop factor; looking to find, for me, a good long-lens pano lens to use to take advantage of the D3's detail capturing ability. Once I settle on my long lens(es) I may use the D3 for 360/180 also. Yea, I know - I could do it all with one camera, but... I've got two so I'm using both, hopefully to their best advantage; so much to learn.
Heinz, I'd greatly appreciate the settings. I do have your settings from your post of 9/2/11
As always, thanks
Bob

BaltimoreBob
04-14-2012, 11:43 AM
I check the parallax at the corners of the fisheye. That is to say, I tilt the camera down then look at the parallax of my alignment objects near the upper corners. I think that checking the parallax from upper right to lower left corner and from upper left to lower right corner might be even a better way because stitching errors due to parallax are frequent between middle row and nadir.

This for a full frame or 180 diagonal fisheye. Probably not useful for cropped or circular fisheye.

Haven't tried that method - I will certainly do so - thanks Demon.
The fisheye I currently us is the Nikon 10.5 which is not the full-frame fisheye, unshaved, with the Nikon D300, which is a "cropped" not full- frame camera. Why wouldn't they work together?
Bob

John Houghton
04-14-2012, 03:03 PM
Bob, The Nikon 10.5mm is known as a fullframe or diagonal fisheye on a D300. The lens is designed for cropped sensor cameras.
.
John

BaltimoreBob
04-14-2012, 06:10 PM
So they will work together in the fashion that DemonDuck described
Appreciate
Bob

Hugh
04-15-2012, 10:02 AM
Hello Bob,

I use a D300 with the 10.5mm fisheye and it works fine.
This combination uses all the sensor and is 180 accross the diagonal, as John states.
If I am not interested in the zenith or nadir I shoot 8 around (45 intervals) with the lens axis horizontal, otherwise I shoot 8 around at -10 plus 4 at +45, which still gives me a small "hole" at the nadir, but if I need a "full" panorama I can either take a "patch" shot or use a tripod cap.
The numbers are degrees, but I am still new to the iPad and have not found how to add the degree symbol.

Best regards, Hugh.

DennisS
04-15-2012, 03:10 PM
with NN5 I'm using the Live view of both my D3 and D300. I moving the vertical rail to center the Nodal Ninja logo in the center of the live view focus box
This is only a starting position. There are way too many variables that come into play. A properly calibrated pano head does not necessarily position the camera so you can see the "knob" in the center of your viewfinder.

I think way too much is being made here. It is quite simple. Follow Smooth's and John's calibration exercises and you will be able to produce panoramas that will stich just fine no matter what camera/pano head combo you use.

Hugh
04-16-2012, 12:34 AM
Hello Bob,

Forgot to mention ...
My 10.5mm is not shaved and is used withe lens hood still in place without any intrusion into the image.
Also, I use the gold ring as my reference for setting up on the top rail and do not take much notice of the numbers, so line up the gold ring with the vertical axis of rotation by eye.

http://hugha.co.uk/NodalPoint/Index.htm

http://hugha.co.uk/Acrobat%20PDF%20Files/Spherical-Panoramas-for-HDS-Point-Clouds.pdf
Psge 7

BaltimoreBob
04-16-2012, 07:20 AM
Hi Hugh
Appreciate the input - thanks
I shoot one row with 6 images, one Zenith and two Nadir - all seems to be working fine now. I don't use the lens rather but somethign similar


Hello Bob,

I use a D300 with the 10.5mm fisheye and it works fine.
This combination uses all the sensor and is 180 accross the diagonal, as John states.
If I am not interested in the zenith or nadir I shoot 8 around (45 intervals) with the lens axis horizontal, otherwise I shoot 8 around at -10 plus 4 at +45, which still gives me a small "hole" at the nadir, but if I need a "full" panorama I can either take a "patch" shot or use a tripod cap.
The numbers are degrees, but I am still new to the iPad and have not found how to add the degree symbol.

Best regards, Hugh.

BaltimoreBob
04-16-2012, 07:29 AM
Dennis
Thanks - but I don't think I agree - "...too much..."
I started this thread by starting by stating what my understanding had arrived at as two broad general rules for use with any camera and any lens seeking confirmation or correction, and flesh-in. Vincen offered confirmation, others flesh-in. I was not seeking instruction and your comments feed into the "understanding".


This is only a starting position. There are way too many variables that come into play. A properly calibrated pano head does not necessarily position the camera so you can see the "knob" in the center of your viewfinder.

I think way too much is being made here. It is quite simple. Follow Smooth's and John's calibration exercises and you will be able to produce panoramas that will stich just fine no matter what camera/pano head combo you use.

DennisS
04-16-2012, 07:52 AM
Bob,

In your quest to understand what is going on when one shoots a panorama, developing a series of rules is a good thing. You have to have a firm grasp on the basics so you know what to adjust when things go wrong.

Over the years PTGui has gotten so much better at stitching images that are not quite aligned. You can be millimeters off and still get a good stitch.

Not everyone has measuring equipment available in order to measure the bottom of the camera to the center of the lens.

After using several camera/lens/pano head combinations I have found that there is 1 rule that can be applied to all setups: Use Smooth's method of calibration. His method tests the entire system, not just your hardware but your stitching software also. Using someone else's numbers and pointing the camera down at the rotator are very good places to start. In the end you still have to verify your rig (and workflow) is calibrated.
When are we going to start posting your panoramas? I for one would like to see what you can produce.

Dennis

BaltimoreBob
04-17-2012, 08:15 AM
Bob,

In your quest to understand what is going on when one shoots a panorama, developing a series of rules is a good thing. You have to have a firm grasp on the basics so you know what to adjust when things go wrong.

Over the years PTGui has gotten so much better at stitching images that are not quite aligned. You can be millimeters off and still get a good stitch.

Not everyone has measuring equipment available in order to measure the bottom of the camera to the center of the lens.

After using several camera/lens/pano head combinations I have found that there is 1 rule that can be applied to all setups: Use Smooth's method of calibration. His method tests the entire system, not just your hardware but your stitching software also. Using someone else's numbers and pointing the camera down at the rotator are very good places to start. In the end you still have to verify your rig (and workflow) is calibrated.
When are we going to start posting your panoramas? I for one would like to see what you can produce.

Dennis

Dennis
Have read so much sometimes get overwhelmed - "Smooth's method" if I recall is shooting straight down and then stitching - if a smoothe circle, all OK, if saw-teeth adjust. Did that to confirm what I was mechanically doing - came out OK. Now I'm confident that I can cart my stuff into the field and set up accurately. I do use PTGui Pro...
I reaching to the point where I'm getting enough confidence in what I'm doing start to finish enough so that what I won't look like it was made by a complete dolt in comparison to others.

DennisS
04-17-2012, 09:43 AM
Bob,
To be clear, Smooth's method is to tilt down enough in order to get the rotator into the picture. I use a 6" disk mounted between the rotator and tripod head. Straight down will not work as well as tilting down, say 30 degrees or so depending on the lens used. You will have to play with the down tilt so that you get enough of the rotator in the picture.
There is so much to read. You can get burried in the details. That is why I try to keep it very simple. Don't worry about what camera/lens/pano head you are using. Just do the calibration and you will be fine.
Now that you think you are calibrated, go out and shoot something, stitch it and show us your results. It took me 6 months of trial and error before I finally got a good workflow figured out. Back when I started there was not much information and even less hardware available. Talk is cheap while a single pano is worth a thousand words.
Dennis