Panning clamp - making tilted panos

[Ellis Vener]

Bill, yes that is a good illustration of what I meant. That keeps the nodal point/entrance pupil in the rotation axis as you rotate (yaw) the camera. For a single row stitched panorama you don't even need the tilt head on top of a standard tripod head.

[Caviar]

Ellis,

No, I haven't tried that combination. Shown below is a crude edit of my original setup to what I imagine you are recommending. What is the advantage of your recommended setup.

Thanks for the help,

Bill

Maybe I'm missing something, but with this result you have a completely different effect. If you would now make a 360 degree pan, your camera points downwards at 180 degrees. You initial setup would keep the tilt the same. For small panos it wouldn't matter too much, but for all the way around this does not seem a good setup.

[Caviar]

Hi Bill,

That's exactly what I do, with a ballhead instead of a swivel monopod head, and a rail with an integrated clamp instead of the bar with dual clamps, if I set out to do single row panos. It makes for a low profile, low center of gravity setup that's very compact.

Cheers,
Bart

P.S. I've attached a quick grabshot of the Ballhead with PCL-1 on top, tilted down, with the MPR-CL II ready to accept the camera L-plate that's permanently on my camera. I usually shoot with the camera in portrait orientation for some added vertical angle of view, and stitch for the wider angle of view

Hi Bart,

Same comment as above. You can't use this setup for 360 degrees panos, right?

Another question though: is that gorillapod version strong enough for a full size DSLR? I have the "normal" one and find it too weak for my 5D mkii plus lens.

[Alistair]

Hi Bart,

Same comment as above. You can't use this setup for 360 degrees panos, right?

In my view it will be fine for 360 degrees because there is (presumably) a rotating base at the bottom of the ball head which is used for rotation rather than the rotating element at the top of the ball head. The angle of the slide rail is somewhat extreme but again I am presuming this is for demonstration purposes.
In fact the rotating element at the top of the ball head is superfluous to me and could be replaced by a plain clamp. The set-up is in fact the one you described earlier in this thread except that it adds a sliding rail to get lens nodal point over the pivot point of the support setup. As you point out earlier, it would be important in this set-up that you only tilt the ball head on one axis for multi-row.
As I pointed out earlier you would need to level the tripod by adjusting its legs so you would really need a levelling bubble on your tripod.
I have a full set of rails and bases for multi-row panos but have used the set-up described above on a number of occasions when I want to travel light. All you really need is a DMP-200, a ball head with rotating base and a tripod with a levelling bubble.
This may well be a different approach to the one intended by Bill and Bart in posting the image but there is more that one way to skin the cat!

[Peter McLennan]

At the risk of infuriating everyone, I just stitched an eight image, downward tilting, ninety degree pano from my 14 mm Samyang of Grand Canyon, shot last week.

Hand held.

[bjanes]

The complexity of a proper panoramic setup is in the levelling and in aligning the camera on both axis around the nodal point of the lens. Some set-ups also include an indexing component to give consistency of degrees of movement between shots but this is a little excessive in my view. Sunwayfoto make very high quality arca swiss compatible components (rails, clamps, rotating bases, levelling bases etc.) which allows one to put a good system together for not TOO exorbitant a sum. The various components can be re-purposed in a number of different projects such as macro rail etc. so they can be useful general additions to your kit.

A single DMP-200 used with a ball head is all you need for parallax-corrected single row stitches. http://www.sunwayfoto.com/e_goodsmulu.aspx?cid=39

If you find you like doing panoramas  you can over time add the components that allow parallax-corrected multi-row.

The author states that the sunwayfoto products are very high quality. The DMP200 is half the price of the RSS equivalent and appears that it would be a very good addition to my pano setup, since my nodal slider with the double clamps is too high in profile. Has anyone else has experience with it and would it be compatible with my RSS gear? A reviewer on the B&H site said that it is a loose fit with the RSS lever clamp but is fine with a screw clamp.

Bill

[Alistair]

The author states that the sunwayfoto products are very high quality. The DMP200 is half the price of the RSS equivalent and appears that it would be a very good addition to my pano setup, since my nodal slider with the double clamps is too high in profile. Has anyone else has experience with it and would it be compatible with my RSS gear? A reviewer on the B&H site said that it is a loose fit with the RSS lever clamp but is fine with a screw clamp.

Bill

Bill, it is arca-swiss compatible so yes, should be compatible with RRS. I have a mix of Wimberley, RRS, Markins, Sunwayfoto and Kirk and they can all be fastened securely to each other. I do not have any lever clamps though so have not encountered the issue you mention above. But the issue does not make sense to me as the profile of the rails and jaws are the same regardless of whether clamp or screw. It sounds like the lever needed adjusting a little.
I first started buying Sunwayfoto L brackets as I got sick of the additional expense for each body upgrade and now have a box full of old and expensive L brackets! I really liked the price/quality and now buy all Sunwayfoto. I would say the quality is same as Wimberley and about 90% of RRS and Kirk which can feel a little over-engineered at times to me but I will leave it to others to comment with their experience.

[Bart_van_der_Wolf]

Hi Bart,

Same comment as above. You can't use this setup for 360 degrees panos, right?

Hi,

Correct, for a 360 pano you'd need to use a multi-row setup which also allows to rotate up and down (=Pitch) through the NPP, not only left and right (=Yaw). If you don't use a multi-row setup, the NPP will move between the various pitch angles you may use. You may be able to do it with a (circular) fisheye (and calibrate the NPP for a single pitch angle), but that requires a bit of extra work and you also want to add a Nadir image to hide the tripod.

Another question though: is that gorillapod version strong enough for a full size DSLR? I have the "normal" one and find it too weak for my 5D mkii plus lens.

It's the GorillaPod Focus, designed for heavier equipment (up to 5 kg). Of course you still need it to grip onto something  really tight to avoid movement in the NPP.

Cheers,
Bart

[Alistair]

Hi,

Correct, for a 360 pano you'd need to use a multi-row setup which also allows to rotate up and down (=Pitch) through the NPP, not only left and right (=Yaw). If you don't use a multi-row setup, the NPP will move between the various pitch angles you may use. You may be able to do it with a (circular) fisheye (and calibrate the NPP for a single pitch angle), but that requires a bit of extra work and you also want to add a Nadir image to hide the tripod.


Cheers,
Bart

Bart, but why not use the rotating base of the ball head for yaw and the ball head itself for pitch? Provided the tripod is level and the ball head was only tilted on one axis (use the virtual horizon on-camera)  I think it would work fine for 360 degrees. The nodal point of the lens is very close to the pivot point of both pitch and yaw (maybe an inch high for pitch but close enough).

[Ellis Vener]

At the risk of infuriating everyone, I just stitched an eight image, downward tilting, ninety degree pano from my 14 mm Samyang of Grand Canyon, shot last week.

Hand held.

That neither surprises or infuriates me, whether you shot with a 'full frame 36mp or an APS-C 6mp camera,Even if you printed at full resolution.

Why doesn't it get under my skin? Simple:

1) You were using a very short focal length lens - meaning any possible near/far subject relationships were likely infinitesimally small unless you had  subject matter very close to the lens. Since you say you photographed with the camera pointing  down into the Grand Canyon, I doubt the latter was the case.

2) The improvements in stitching software, whether you use PTGui, Auto Pan Pro, or Photoshop CS6/CC have been major, even for straight out of the box result with no fine tunign of the stitching parameters.

What I have been trying to point out with regard to rotators and other specialized support gear is that:

A) The more complex the physical geometry (near/far spatial relationships) the more need there is for precision instruments like rotators and nodal sliders

B) For a grand vista panoramic landscapes with wide angle to normal, maybe with even short telephoto lens (up to the 100~135mm range) specialized camera support hardware, including even a simple tripod, may not be necessary at all - but it will make your life simpler,  your results more repeatable, open up photo possibilites that handholding a camera precludes. Certainly at a minimum, a tripod will help you get a higher degree of fine detail resolution.

C) Where specialized support gear like the RRS PG-02 Omni really comes into its own is with spatially complex scenes, long exposures, or long lenses.

D) whether the rotation plane for the camera camera is level  or tilted up or down or at an angle to the horizon, doesn't really matter to the camera or the stitching software as long as you have plenty of overlap between adjacent frames.

[Bart_van_der_Wolf]

Bart, but why not use the rotating base of the ball head for yaw and the ball head itself for pitch?

Hi Alistair,

Because when you rotate around the ball head base (Yaw), the entrance pupil will rotate in a circle which causes horizontal parallax, unless you use a NNP bar or rail to align the entrance pupil with the rotation axis. When you tilt (Pitch) the ball head (rotation axis runs through the ball), you will move the entrance pupil forward/backward and down (and maybe sideways, because a ball head allows too much freedom to position accurately) which will create vertical and some horizontal and depth parallax.  

Provided the tripod is level and the ball head was only tilted on one axis (use the virtual horizon on-camera)  I think it would work fine for 360 degrees. The nodal point of the lens is very close to the pivot point of both pitch and yaw (maybe an inch high for pitch but close enough).

Whether 'very close' is 'close enough', depends on the amount of foreground/background detail and how detailed it is. It also varies with focal length, shorter is more forgiving. The worst kind of subject to get right with 'almost' no NPP shift are occlusions, where e.g. branches or gates or poles or windows are in the foreground and in focus, and the background also has sharp detail. You are fully at the mercy of the stitcher quality (both additional camera shift capability and blending are required to stand a chance) if anything useful comes from that. 'Murphy' will make sure that that 'once in a lifetime shot' will have unsolvable errors.

There are only two possible exceptions to get it right without a full multi-row set-up. One is by calibrating the NNP over the leveled base center with a single, fixed, unique, tilted position of the ball head. As long as the tilt is not changed, alignment will remain the same, correct. It is of course rather difficult to exactly nail that tilt angle, so repeating that setup will again require calibration (which takes time, and it is much more efficient if only required to do once). The other exception is when the subject is exactly in a single flat plane (no parallax possible in a flat plane), AKA a viewplane, and your stitching software is capable of calculating a camera position (entrance pupil) displacement.

Cheers,
Bart

[Alistair]

Hi Alistair,

Because when you rotate around the ball head base (Yaw), the entrance pupil will rotate in a circle which causes horizontal parallax, unless you use a NNP bar or rail to align the entrance pupil with the rotation axis.

Yes of course. But surely that is why you have the MPR-CL II there on top of the ball head.

When you tilt (Pitch) the ball head (rotation axis runs through the ball), you will move the entrance pupil forward/backward and down (and maybe sideways, because a ball head allows too much freedom to position accurately) which will create vertical and some horizontal and depth parallax.  

Agreed. But variance will be very limited as the NPP is directly above the yaw pivot point and very close to the pitch pivot point. Sideways movement can be managed via the camera's virtual horizon.

Whether 'very close' is 'close enough', depends on the amount of foreground/background detail and how detailed it is. It also varies with focal length, shorter is more forgiving. The worst kind of subject to get right with 'almost' no NPP shift are occlusions, where e.g. branches or gates or poles or windows are in the foreground and in focus, and the background also has sharp detail. You are fully at the mercy of the stitcher quality (both additional camera shift capability and blending are required to stand a chance) if anything useful comes from that. 'Murphy' will make sure that that 'once in a lifetime shot' will have unsolvable errors.

There are only two possible exceptions to get it right without a full multi-row set-up. One is by calibrating the NNP over the leveled base center with a single, fixed, unique, tilted position of the ball head. As long as the tilt is not changed, alignment will remain the same, correct. It is of course rather difficult to exactly nail that tilt angle, so repeating that setup will again require calibration (which takes time, and it is much more efficient if only required to do once). The other exception is when the subject is exactly in a single flat plane (no parallax possible in a flat plane), AKA a viewplane, and your stitching software is capable of calculating a camera position (entrance pupil) displacement.

Cheers,
Bart

Agreed again. But the OP seemed to be looking to avoid expenditure and I feel your ball head and nodal slide solution will get him underway with some chance of success until he feels like digging into his offsprings' inheritance. ;-)

[Caviar]

Thanks for all the explanations and discussion on this 'math' topic. I have pretty clear picture now and understand the pros and cons of the various solutions. Thanks!

[Ellis Vener]

it's geometry, not mathematics but I am glad you are getting the feel for it. It is always good to have a deeper understanding of what is going on even if it seems to have no immediately revelance to what you are doing right now. The discussion has certainly helped me.

So here's where I am at with my understanding of the relationship between tripod heads (including leveling bases), rotators, and nodal displacement slides. My understanding is based on a triad of thought experiment, modeling, and most crucially empirical testing by shooting actual photos.

1) When doing single row panoramas, best practice is to keep the nodal point or entrance pupil of the lens in the axis of rotation. This holds true where the plane of rotation is at any angle of tilt , including 0˚ ((AKA "level"). 

2) As long as the N.P. stays in that rotational axis it makes no difference if the rotator is below the tripod head or on top of it.

3) A nodal displacement rail facilitates that alignment.

4) When shooting a multirow panoramic, ideally a second rotator is needed to create a rotational axis perpendicular to the first rotator. This is necessary to keep the N.P. in the primary rotational axis when you change the pitch angle to shoot the additional rows.

5) How necessary it is to maintain the N.P. to the rotational axis alignment and get good high quality results is dependent  on three factors:

A) Focal length of the lens as in most cases this governs the size of the subject to the size at which it is rendered on the sensor (see C.)

B) The difference in near-to-far subject distances in the subject field, relative to the distance of the camera to the subject field.

C) The degree of fine detail at the size the photograph is reproduced at (and there are a slews of perceptual factors inside that, but basically the smaller the reproduction and the greater the viewer's distance to the reproduction, the less ultra-fine detail matters.

6) Software: Contemporary stitching programs like PTGui Pro, Auto Pano Pro, and Photomerge in PsCS6 and PsCC are very good at correcting  many common user errors during photography, and in some (but not all) cases can do a "good enough" job that specialized camera mounting gear may not be necessary. However it certainly will make life a lot easier.  Where specialized rigs like the Really Right Stuff  PG-02 package http://www.reallyrightstuff.com/s.nl/it.A/id.8872/.f or the GigaPan EPIC http://www.bhphotovideo.com/c/product/689699-REG/Giga_Pan_EPIC_PRO_EPIC_Pro_Robotic_Camera.html are essential  is when the subject field is spatially complex and/or the reproduction size is large.

My practice when stitching is to have the stitching software output the panoramic composite as a "blended plus layers" PSD format document.  Having the layers option allows me to go in and edit the contents of the photo especially important if some of the objects in the scene moved during the process of photographing the scene.

[bjanes]

Bill, it is arca-swiss compatible so yes, should be compatible with RRS. I have a mix of Wimberley, RRS, Markins, Sunwayfoto and Kirk and they can all be fastened securely to each other. I do not have any lever clamps though so have not encountered the issue you mention above. But the issue does not make sense to me as the profile of the rails and jaws are the same regardless of whether clamp or screw. It sounds like the lever needed adjusting a little.
I first started buying Sunwayfoto L brackets as I got sick of the additional expense for each body upgrade and now have a box full of old and expensive L brackets! I really liked the price/quality and now buy all Sunwayfoto. I would say the quality is same as Wimberley and about 90% of RRS and Kirk which can feel a little over-engineered at times to me but I will leave it to others to comment with their experience.

I bought the Sunwayfoto DMP 200 nodal slider and am returning it as it is very loose in my RSS lever release clamp with which it is unusable. AFAIK, the lever clamp is not adjustable. However, the DMP 200 did work with my RSS screw clamps. The finish and engraved numbers are not quite up to RSS standards.

Regards,

Bill

[bjanes]

Hi Bill,

That's exactly what I do, with a ballhead instead of a swivel monopod head, and a rail with an integrated clamp instead of the bar with dual clamps, if I set out to do single row panos. It makes for a low profile, low center of gravity setup that's very compact.

Cheers,
Bart

P.S. I've attached a quick grabshot of the Ballhead with PCL-1 on top, tilted down, with the MPR-CL II ready to accept the camera L-plate that's permanentlytt on my camera. I usually shoot with the camera in portrait orientation for some added vertical angle of view, and stitch for the wider angle of view

Bart,

That is an interesting tripod. Who makes it?

Bill

[Bart_van_der_Wolf]

That is an interesting tripod. Who makes it?

Hi Bill,

It's a GorillaPod 'Focus' (=the largest model), made by 'Joby.com'. It can come in handy when you need to temporarily attach the camera to, i.e. wrap the pod's legs around, e.g. a branch, lamppost, gate, a car seat for an interior, etc. Also allows to get low to the ground without getting mud/sand or scratches on the camera, or rest it against a wall. I had my RRS BH-55 mounted on top, but smaller ball heads may also suffice, depending on the weight that needs to be supported.

Cheers,
Bart