Topic: Some Confusion Over Tripod Leveling for Panoramas (Read 17898 times)

JimAscher · « **on:** February 15, 2012, 11:10:43 am »

When I shoot my panoramas I place a leveling head UNDER the ball head. Thus when I've leveled my tripod with the leveling head I'm still able to make vertical tilt adjustments with the ball head. However, I note that RRS advises that their PCL-1 Panning Clamp be placed ON TOP of the ball head. (http://reallyrightstuff.com/ProductDesc.aspx?code=Pano-Elem-Pkg&type=3&eq=&desc=Pano-Elements-Package%3a-For-single-row&key=it). This set-up, however, would seem to me to preclude the option for making vertical tilt adjustments with the camera. What am I missing here? Is there a presumption that for "true" panoramas, no vertical tilt adjustment is recommended or appropriate?

Ellis Vener · « **Reply #1 on:** February 15, 2012, 12:54:40 pm »

if you put the PCL-1 or similar on top of the head you are using the tripod head as a leveling device - if you level the top of the head.

If you level the base of the head and then tilt the head, even with a nodal slide, you are moving the nodal point / entrance pupil of the lens away from being vertically centered over the rotation point as you rotate.

Depending on your subject matter and the near to far subject relationships involved this may or may not matter. Distant landscapes with few to no foreground subjects would be a case where your software can likely compensate for the parallax.

My experience from doing it both ways is that as a general rule it makes for less work if I keep the nodal point / entrance pupil centered on a common vertical and horizontal axis. Which means using a riser with a tilt mechanism (like the RRS PG-02 VA or similar) above a leveled head.

Although I use L-bracket plates on my cameras, for some vertically oriented single frame shots I like having a PCL-1 as the head topper for another reason: I tilt the head over and get the camera horizon level and then use the PCL-1 to set the vertical angle.

Bart_van_der_Wolf · « **Reply #2 on:** February 15, 2012, 02:46:50 pm »

Quote from: Ellis Vener on February 15, 2012, 12:54:40 pm

If you level the base of the head and then tilt the head, even with a nodal slide, you are moving the nodal point / entrance pupil of the lens away from being vertically centered over the rotation point as you rotate.

Hi Ellis,

Exactly correct. Rotating at the base of the head risks that the entrance pupil does not stay stationary unless it is also exactly leveled. Another benefit is that Yaw rotation at the top keeps the entrance pupil stationary even when the top is not level. That allows to shoot subjects where the horizon is not exactly center image.

Cheers,
Bart

JimAscher · « **Reply #3 on:** February 15, 2012, 03:49:58 pm »

Quote from: Ellis Vener on February 15, 2012, 12:54:40 pm

...Although I use L-bracket plates on my cameras, for some vertically oriented single frame shots I like having a PCL-1 as the head topper for another reason: I tilt the head over and get the camera horizon level and then use the PCL-1 to set the vertical angle.

Ellis: Much useful information from you and Bart. My previous panorama photography with the leveling base under the ball head has worked out all right I guess because the subject matter has always been at infinity focus. Now that I've ordered myself a nodal slide and am intending to calculate the nodal point for panorama work with both near and far objects, I hadn't anticipated that a vertical lens tilt might have this effect.

I am, though, Ellis, not certain I entirely understand the above quote from part of your message. Are you saying that instead of using an L-bracket for vertical framing, one might instead tilt the ball head (after the leveler placed on top of it has leveled the ball head) ninety-degrees to achieve vertical framing, and by which the nodal point will not be altered by a subsequent vertical tilt? I"m certain I've got this recommendation of yours all wrong, but I don't know how. Thanks. Jim

Ellis Vener · « **Reply #4 on:** February 15, 2012, 04:03:36 pm »

Quote from: JimAscher on February 15, 2012, 03:49:58 pm

I am, though, Ellis, not certain I entirely understand the above quote from part of your message. Are you saying that instead of using an L-bracket for vertical framing, one might instead tilt the ball head (after the leveler placed on top of it has leveled the ball head) ninety-degrees to achieve vertical framing, and by which the nodal point will not be altered by a subsequent vertical tilt? I"m certain I've got this recommendation of yours all wrong, but I don't know how. Thanks. Jim

Not quite. I was speaking of either simple single frame photos where I want to be able to make a vertical composition (portrait orientation) with the horizon level and then find tune the vertical framing. Doing this also works for for vertical panoramas.

JimAscher · « **Reply #5 on:** February 15, 2012, 04:54:23 pm »

Quote from: Ellis Vener on February 15, 2012, 04:03:36 pm

Not quite. I was speaking of either simple single frame photos where I want to be able to make a vertical composition (portrait orientation) with the horizon level and then find tune the vertical framing. Doing this also works for for vertical panoramas.

Ellis: I regret that you've even got me more confused than before. But I don't want you to expend any more of your time trying to get it through my (obviously) dense brain. What I do think that I understand is that it's apparently not possible to manage a vertical tilt in panorama photography and also maintain a previously calculated and set nodal point -- that the vertical tilting upsets the accuracy of the nodal calculations. One has the choice of either shooting panoramas at infinity focus (with no near and far objects in the framing) or setting the nodal point with the camera leveled with a leveler at the top of the ball head, then shooting straight ahead, with no vertical tilt (i.e., no lowering or raising of the horizon (or anything else)). Sounds pretty dire to me. Many thanks again. Jim

marcmccalmont · « **Reply #6 on:** February 16, 2012, 04:17:45 am »

Invest in a leveling tripod and a panning head then you can pretty much do anything you want
Marc

JimAscher · « **Reply #7 on:** February 16, 2012, 06:38:17 am »

Quote from: marcmccalmont on February 16, 2012, 04:17:45 am

Invest in a leveling tripod and a panning head then you can pretty much do anything you want
Marc

Marc: Thanks for the reassurance. I already have both. Jim

Ellis Vener · « **Reply #8 on:** February 16, 2012, 08:44:15 am »

Quote from: JimAscher on February 15, 2012, 04:54:23 pm

Ellis: I regret that you've even got me more confused than before. But I don't want you to expend any more of your time trying to get it through my (obviously) dense brain. What I do think that I understand is that it's apparently not possible to manage a vertical tilt in panorama photography and also maintain a previously calculated and set nodal point -- that the vertical tilting upsets the accuracy of the nodal calculations. One has the choice of either shooting panoramas at infinity focus (with no near and far objects in the framing) or setting the nodal point with the camera leveled with a leveler at the top of the ball head, then shooting straight ahead, with no vertical tilt (i.e., no lowering or raising of the horizon (or anything else)). Sounds pretty dire to me. Many thanks again. Jim

1) the nodal point is set by using the slide in the panning clamp to place the nodal point /entrance pupil in line with the axis of rotation. Whether you have the rotation plane oriented horizontally or vertically or somewhere in between makes no difference: the mechanics of how it works are still the same. That is another reason why you want a panning mechanism like the Really Right Stuff PCL-1 on top of your tripod head, and not use the rotating base of your tripod head to pan with*.

Here is how you can prove this for yourself without spendign a penny: Hold your left hand (or your right, no difference) up in front of you and extend your index finder. Place your other hand , open and palm down - this will be the plane you are rotating - on top of the vertical digit it doesn't matter where the finger makes contact with the finger - the finger is the rotational axis . Rotate the hand that is on top from left to right. Without removing that hand, tilt the finger to the horizontal. As long as you don't change the point where the axis intersects the plane of rotation if the nodal point/ entrance pupil for the lens stays in the same line as the axis you are rotating the camera and lens around you eliminate the possibility of parallax ( http://en.wikipedia.org/wiki/Parallax) occurring.

* yes that can work but only if the top of the head is perfectly parallel with the base of the tripod head, or if all of the significant subjects in your pan are effectively at an infinite distance from your lens, and in some cases if you are willing to spend a lot of time in front of your computer screen editing multiple layers of your panorama.

JimAscher · « **Reply #9 on:** February 16, 2012, 11:16:26 am »

Quote from: Ellis Vener on February 16, 2012, 08:44:15 am

...That is another reason why you want a panning mechanism like the Really Right Stuff PCL-1 on top of your tripod head, and not use the rotating base of your tripod head to pan with*.

* yes that can work but only if the top of the head is perfectly parallel with the base of the tripod head, or if all of the significant subjects in your pan are effectively at an infinite distance from your lens....

Ellis: Thanks for your further expenditure of time and effort to get the concept across to me. There is definite a glimmer of understanding developing!

I have an Acratech Leveling Base (http://acratech.net/product.php?productid=3) which, as I believe I stated at the beginning of this thread, I use under my ball head as my rotational point. With my camera on top of the ball head I photograph panoramas occasionally using some vertical tilt, but just about always set for infinity focus. The results to date have been fine.

If, in accord with your conceptualization, I were to instead place my Arcatech Leveling Base on TOP of my ball head, instead of underneath it, and the nodal slide and camera on top of that, wouldn't this then be the equivalent of using a RRS PCL-1 (and nodal slide) on top of the ball head? With my camera on top of the Acratech base and the base leveled and the nodal point axis appropriately determined by "sliding" the nodal slide in the Arcatech base, it is my present understanding based on your explanation that I could then introduce some vertical tilt while maintaining the integrity of the rotational nodal point axis. Or am I still getting it all wrong? I trust I have worded this sufficiently precisely, but I have my doubts, for which I apologize if I am causing any confusion. Again, thanks. Jim

Ellis Vener · « **Reply #10 on:** February 16, 2012, 12:10:52 pm »

In theory, yes. Really Right Stuff has some goid tutorials on the mechanics involved in shooting panoramas including one on how to determine where the nodal point / entrance pupil for a lens is.

JimAscher · « **Reply #11 on:** February 16, 2012, 12:24:41 pm »

Quote from: Ellis Vener on February 16, 2012, 12:10:52 pm

In theory, yes. Really Right Stuff has some goid tutorials on the mechanics involved in shooting panoramas including one on how to determine where the nodal point / entrance pupil for a lens is.

Thanks, Ellis. I will check them out. Jim

OldRoy · « **Reply #12 on:** February 17, 2012, 05:50:13 am »

I shoot VR panoramas both commercially and for amusement (a personal project). So my query is primarily derived from this experience, using fish-eyes, although I also shoot and stitch landscape panos with various subjects using non-FE primes.

I have an NN 3 pano head. It's under-engineered but usable and I keep a record of accurate setup positions for NPPs of the lenses I habitually use. I seldom bother too much about levelling at the shooting stage given that, particularly with VR panos, you almost invariably have to level at some point during the stitching.

The emphasis on levelling - I long ago abandoned lugging my Manfrotto levelling plate around with me - has always puzzled me. Given that every pixel is remapped in a stitched pano, why worry about accurate levelling?

EDITED: Here's another useful tutorial on finding the NPP. The site has numerous valuable tutorials for people beginning with stitched panos, although the emphasis is on VR.
http://www.johnhpanos.com/epcalib.htm
Roy

Bart_van_der_Wolf · « **Reply #13 on:** February 17, 2012, 08:21:18 am »

Quote from: OldRoy on February 17, 2012, 05:50:13 am

The emphasis on levelling - I long ago abandoned lugging my Manfrotto levelling plate around with me - has always puzzled me. Given that every pixel is remapped in a stitched pano, why worry about accurate levelling?

Hi,

I agree that the role of leveling is often misunderstood in the context of pano stitching. However, it does matter when it's done incorrectly (e.g. with rotation at the base of a leveled head), and/or if one wants to avoid having to crop unnecessary amounts from the top and bottom of wide views, or resort to content aware fill. One recurring stitching question in forums is about horizons that are not straight but follow a wave pattern, BTW easily fixed in pano software by using the correct pitch, and where better leveling would have saved more of the original image after software leveling and cropping.

So, yes it helps to get a proper leveling when you do not shoot 360 degree VR images, but no it is not necessary to overdo it, unless final cropping after software leveling must be kept to an absolute minimum.

Cheers,
Bart

JimAscher · « **Reply #14 on:** February 17, 2012, 08:59:56 am »

Quote from: OldRoy on February 17, 2012, 05:50:13 am

...The emphasis on levelling - I long ago abandoned lugging my Manfrotto levelling plate around with me - has always puzzled me....

Roy

My leveling with my Acratech Leveling Base mounted under my tripod head is so easy, with the leveling base itself of no significant additional size or weight for me, that its use seems more than justified by the panos it enables me to produce, which require almost no cropping at the top or bottom, other than for aesthetic reasons.

Ellis Vener · « **Reply #15 on:** February 17, 2012, 11:52:47 am »

Quote from: JimAscher on February 17, 2012, 08:59:56 am

My leveling with my Acratech Leveling Base mounted under my tripod head is so easy, with the leveling base itself of no significant additional size or weight for me, that its use seems more than justified by the panos it enables me to produce, which require almost no cropping at the top or bottom, other than for aesthetic reasons.

How are you making sure the platform you attach the camera to is level? You can level the base of a tripod all you want, but it does nothing for you if the base of the camera is not level. I nfact all you are doign is adding an extra joint to your set up. I have a leveling platform ( the big Manfrotto) and I use it - but instead of putting a tripod head on top of it I put an RRS PCL-1 rotating clamp there.

JimAscher · « **Reply #16 on:** February 17, 2012, 12:28:02 pm »

Quote from: Ellis Vener on February 17, 2012, 11:52:47 am

How are you making sure the platform you attach the camera to is level? You can level the base of a tripod all you want, but it does nothing for you if the base of the camera is not level. I nfact all you are doign is adding an extra joint to your set up. I have a leveling platform ( the big Manfrotto) and I use it - but instead of putting a tripod head on top of it I put an RRS PCL-1 rotating clamp there.

Ellis:

1. My comment above re my use of the Arctech Leveling Base was based on my panorama photography efforts to date when parallax has not been a consideration, or necessity. However, my concerns now for parallax adjustment is what has motivated my initiation of this thread.

2. The excellent RRS tutorial you pointed me to includes the following:

"Level Your Equipment

"One Step Leveling

"If you use our PCL-1 panning clamp, this step is a snap. Loosen the ball and shift the clamp until the spirit level indicates its level, and in one step, you've leveled both the camera (the lateral axis) and the panning base (the axis of rotation).

"Two-Step Leveling

"Alternatively, level your tripod and camera separately. When you level the tripod, what you're really trying to get level is the panning base of your ballhead. For leveling the tripod, a leveling base or a leveling center column makes the first step quick and easy. And to level the camera? Use the spirit level built into one of our clamps, or use a double-bubble level mounted in the camera's hot shoe."

Ellis, it's the two-step leveling that pertains in my case. But I am still exploring the issue of having an option for introducing some vertical tilt, for which you have been very helpful.

Ellis Vener · « **Reply #17 on:** February 17, 2012, 12:51:25 pm »

If you don't want a rotation plane that is horizontally level, the answer is simple with the PCL-1 on top of your your tripod head. Just tilt the head.

and if you are worried about parallax, add a nodal slide.

the rig I generally use is, from top to bottom

1) A heavy duty tripod (A Gitzo 410C)

2) an equally heavy duty head (either a Foba ASMIA (http://www.bhphotovideo.com/c/product/247458-REG/Foba_31_0118_ASMIA_Double_Pan_Tilt.html) or an Arca-Swiss B1 Monoball if I am travelling light.

3) An RRS PCL-1 Panning clamp.

4) An RRS CB-10 or CB-18 camera bar ( http://reallyrightstuff.com/ProductDesc.aspx?code=CB-10 and http://reallyrightstuff.com/ProductDesc.aspx?code=CB-18)

5) A PG-02VA fitted with a B2-Pro II clamp

6) An RRS MPR-CL-II or for very long focal length lenses qith their own tripod mounts, the RRS CB-10 with RRS FAS sliding clamp.

I batch process the raw frames with Lightroom 4 and stitch the exported TIFFs with PTGui Pro 9.x. Any clean up beyond PTGui Pro's capability and any other necessary post processing I do in Photoshop CS5.

JonathanRimmel · « **Reply #18 on:** February 17, 2012, 01:12:21 pm »

Now let me see if I understand all of this correctly. If I want to do panoramas, it is best to have a leveling plate. But what is going to produce better results, putting it on top of the ball head, or below? I always thought below would work best, but now I am not so sure. I was thinking of getting one for my next tripod.

Ellis Vener · « **Reply #19 on:** February 17, 2012, 01:24:32 pm »

Quote from: JonathanRimmel on February 17, 2012, 01:12:21 pm

Now let me see if I understand all of this correctly. If I want to do panoramas, it is best to have a leveling plate. But what is going to produce better results, putting it on top of the ball head, or below? I always thought below would work best, but now I am not so sure. I was thinking of getting one for my next tripod.

The RRS PCL-1 is not a leveling device, it is a rotational device.

Your tripod head is a leveling (and "level" is just a tilt that is perpendicular to a line running through the center of the planet) device and you want your rotational device ( the PCL-1) on top of either your tripod head or on top of your dedicated leveling device (which really is just a tripod head with a limited range of movement).

I think I have now explained the mechanics of this as clearly and in as simple of terms as I think is humanly possible.

Author Topic: Some Confusion Over Tripod Leveling for Panoramas (Read 17898 times)