Multi Shot, Single Row (not really panoramic) Setup Help Requested

[dwswager]

Can some answer for me or point me to a discussion of setting up for a Single Row, Multi Shot Sequence with the camera pointed down?

Answer I'm looking for is NOT go buy a $1000 RRS Pano Package!

Neglecting for parallax issues, to shoot level, I know all I need is a panning capability above my Arca B1 Clamp and below the camera that is level.  I'm assuming that to shoot with the camera pointed downward, I would still need to pan from a level base and the axis of the camera perpendicular to the direction I'm shooting (side to side of the camera as it sits on the rig) would have to be level while the axis in the direction of shooting can be tilted.  Basically, I would level the base of the ballhead, level the camera above it, then tilt the camera downward in that one axis only and use the level panning base of the ballhead to rotate the camera.

[Ellis Vener]

For single row stitched work all you need is a panning clamp and nodal slide. And then you tilt the head 90 degrees so the camera is pointing down.

[Bart_van_der_Wolf]

Can some answer for me or point me to a discussion of setting up for a Single Row, Multi Shot Sequence with the camera pointed down?

Answer I'm looking for is NOT go buy a $1000 RRS Pano Package!

Hi,

In addition to Ellis' answer, if your subject is flat, it can be done with software alone (in PTGUI it is called Viewpoint correction).

If your subject is 3-Dimensional, and you want to avoid parallax, you have to keep the entrance pupil of the lens stationary and shift the sensor plane. All other cases will create parallax, so anything goes. Using long focal lengths and a large distance to the subject matter will reduce the amount of parallax.

So it depends on the subject, and whether parallax should be avoided or not.

Cheers,
Bart

[dwswager]

For single row stitched work all you need is a panning clamp and nodal slide. And then you tilt the head 90 degrees so the camera is pointing down.

I have a macro focusing rail that I'm using as a nodal slider.  That isn't the issue.  And pointing the camera down 90 degrees isn't a solution as I want to point it maybe 10, 20, or 30 degrees below the horizon.  I don't believe I can use just a panning clamp as the panning clamp would hold the camera which will not be level and when I pan it, it would rotate in an arc (down in the middle) relative to plane of the image.  I'm fairly certain the panning base must be level.

If I look at the gimballed super setups the rotating gimbals at the apex of the tripod is always leveled.  The camera mounted on the arm can be pointed down, but that point of rotation has to always be level.

[dwswager]

Trying to recreate this type of camera position without all the equipment until I figure it out.  Then I will start buying more equipment.  Looking at the image above from Denton Images, The clamp above the ballhead must be level and the clamp holding the camera must be perpendicular to the clamp above the ballhead.  However, the camera is pointing down.

Neglecting parallax and the whole nodal point issue for now.  If I mount the camera just as it is in the photo in the clamp above the ball head (pointing down, but level side to side along the top edge, level the tripod apex under the ballhead and rotate the panning base of the ballhead, that should do it...right?  Then I would need the nodal slider to move the camera back and get the nodal point of the lens in line with the point of horizontal rotation.  Obviously, this is all easier to setup with all the gear pictured above.  However, I always like to learn first and then buy what I need to do it easy.

[Bart_van_der_Wolf]

I have a macro focusing rail that I'm using as a nodal slider.  That isn't the issue.  And pointing the camera down 90 degrees isn't a solution as I want to point it maybe 10, 20, or 30 degrees below the horizon.  I don't believe I can use just a panning clamp as the panning clamp would hold the camera which will not be level and when I pan it, it would rotate in an arc (down in the middle) relative to plane of the image.  I'm fairly certain the panning base must be level.

Hi,

Now that it is becoming a bit clearer what you are trying to achieve, it may be simpler to do than you think. However, it does require some components that play nice together. This is the basic requirement.

There are cheaper clones (e.g. Hejnar-Photo) on the market, but the required functionality is what you see there. Most flexible systems use Arca style camera plates, clamps, bars. That will allow expansion over time by only adding a few components, or replacing them for more sturdy ones for heavier gear requirements.

The rotation axis/axes should be located at the center of the entrance pupil position, the No-Parallax slider with camera at the end should be mounted on that rotator (which in turn should be mounted on top of a ballhead or similar positioning device).

If I look at the gimballed super setups the rotating gimbals at the apex of the tripod is always leveled.  The camera mounted on the arm can be pointed down, but that point of rotation has to always be level.

No, that's not a requirement. One can tilt the camera up or down at any degree. As long as the single row rotation (e.g. panning) components are linked to the camera plate, one just expands the Field of View by rotating. The pano stitching software then should automatically (or be manually instructed to) use a horizon that is not running through the center of the images. A downwards pitch will require a horizon setting that's above the center of the images, just where the actual horizon is.

Cheers,
Bart

[Bart_van_der_Wolf]

Trying to recreate this type of camera position without all the equipment until I figure it out.  Then I will start buying more equipment.

Hi,

You are needlessly complicating things by working backwards from a multirow setup. Start simple, and you will be able to expand over time.

A single row setup consists of a few basic components that can be reused if the right components are selected. The basic thought remains: keep the axes of rotation aligned with the center of the entrance pupil of the lens. Leveling of the whole setup is not required (!) for stitching.

Step 1: make sure that the lateral offset is (or can be) positioned such that the optical axis is positioned exactly aligned with the center of the panning base that's connected to the camera. A Camera L-plate (Arca style) is useful, because it allows some lateral shift, and if well designed it is easy to switch between landscape and portrait orientation.
Step 2: Use a No-Parallax bar or slider (or macro-rail) to get enough fore-aft offset to also align the center of the entrance pupil with the center of the panning base.

You now have 2-Dimensional alignment of the entrance pupil with the axis of rotation provided by the panning clamp. Any rotation will keep the center of the entrance pupil stationary, thus no parallax will exist. You're done.

Pitching the panning base to a non-level angle will not introduce problems, as long as the Pano stitching software can position the horizon at the required position. It can't be stressed enough, leveling is not a requirement for a perfect stitch, as long as the software is decent enough to reposition the theoretical/model horizon.

This is as simple as it gets, and with some experience one can even do some multi-row stitching if the subject is not too critical.

Cheers,
Bart

[bjanes]

Can some answer for me or point me to a discussion of setting up for a Single Row, Multi Shot Sequence with the camera pointed down?

Answer I'm looking for is NOT go buy a $1000 RRS Pano Package!

Neglecting for parallax issues, to shoot level, I know all I need is a panning capability above my Arca B1 Clamp and below the camera that is level.  I'm assuming that to shoot with the camera pointed downward, I would still need to pan from a level base and the axis of the camera perpendicular to the direction I'm shooting (side to side of the camera as it sits on the rig) would have to be level while the axis in the direction of shooting can be tilted.  Basically, I would level the base of the ballhead, level the camera above it, then tilt the camera downward in that one axis only and use the level panning base of the ballhead to rotate the camera.

Here is a makeshift setup utilizing the RSS basic pano setup for single row and adapted for multi-row by attaching a monopod head that allows the camera to be tilted up or down. Not that elegant, but it works.

Bill

[Ellis Vener]

Trying to recreate this type of camera position without all the equipment until I figure it out.  Then I will start buying more equipment.  Looking at the image above from Denton Images, The clamp above the ballhead must be level and the clamp holding the camera must be perpendicular to the clamp above the ballhead.  However, the camera is pointing down.

Neglecting parallax and the whole nodal point issue for now.  If I mount the camera just as it is in the photo in the clamp above the ball head (pointing down, but level side to side along the top edge, level the tripod apex under the ballhead and rotate the panning base of the ballhead, that should do it...right?  Then I would need the nodal slider to move the camera back and get the nodal point of the lens in line with the point of horizontal rotation.  Obviously, this is all easier to setup with all the gear pictured above.  However, I always like to learn first and then buy what I need to do it easy.

Three terms:

Pitch - fore and aft tilt angle.
Roll - left to right tilt angle.
Yaw - "compass" angle at which the lens is pointing

The purpose of displacement rails - a single "nodal slide" for single row stitching,  or a nodal slide + a horizontal and vertical displacement bar for multi-row stitching - is to eliminate parallax errors.
 
The purpose of having a panning mechanism on top of the tripod head is to ensure that  no matter what angle  the heads platform is tilted to , the  the arc  described by the panning motion will be a plane - a straight line - and not a curved one.

Since the base of the head is separated by the from the camera platform by either a ball or separate joints setting pitch and roll angle -  it  doesn't matter if  the base of the head is level any more than whether ground or floor the tripod is resting on is level.  This holds true as long as the position of the head and tripod are fixed...meaning  don't use the lower rotator!*

With a rotating clamp at the camera platform level, once you set the camera platform at any angle you define the rotation plane for the camera.

Now if you are planning on shooting a mosaic of say nine frames to create a higher resolution view of an object then you will need a more complicated set up where you have two rotators - one for yaw (side to side angle and the nodal point  of the lens will need to be centered where the yaw axis and the pitch axis intersect.  That is the purpose of the more complicated panoramic rigs like the RRS PG-02 and Ultiamte pro, or the rig you pictured.

*If the field of the subjects is sufficiently far away - at photographic infinity for the lens and camera you are using -  this doesn't matter, especially with contemporary stitching software. It's where you have significant differences  in near/ far distances -relative to the subject - that it does.

[dwswager]

Since the base of the head is separated by the from the camera platform by either a ball or separate joints setting pitch and roll angle -  it  doesn't matter if  the base of the head is level any more than whether ground or floor the tripod is resting on is level.  This holds true as long as the position of the head and tripod are fixed...meaning  don't use the lower rotator!*

With a rotating clamp at the camera platform level, once you set the camera platform at any angle you define the rotation plane for the camera.

But if I PITCH the panning clamp downward so the camera is no longer level the camera body will ROLL as the panning clamp rotates through the panning motion and the panning will be in an arc and not a straight line.  Thinking Tilt-A-Whirl.

Thanks for all the suggestions, think I got this licked!

Givens:

• A horizontal panning plane of motion must be level to the horizon and a vertical panning plane of motion must be perpendicular to the horizon and therefore the horizontal panning plane, otherwise you introduce ROLL into the camera body.
• The nodal point of the lens/camera combination must be centered about the axis or rotation in both the horizontal or vertical planes to eliminate parallax errors.  If the distance between the near and far objects in the image is small compared to the distance from the camera then parallax errors will be minimized.

Single Row - Camera Level

Equipment necessary: Panning Clamp and Nodal slider.  All that is necessary is to level the panning clamp and move the nodal slider with the camera and lens such that the nodal point is in line with the axis of rotation of the panning clamp.  Therefore, we introduce no Pitch or Roll.  Simple in theory, but, of course, easy enough to screw up in the field!

Single Row - Camera not level

Now we introduce Pitch into the camera (No Roll allowed).  Equipment necessary: Panning Clamp, Nodal Slider and a way to pitch the camera forward without introducing roll.  Thinking a tilting monopod head would work.



We level the panning clamp, clap the monopod head into it so it is level.  Mount the nodal slider into the monopod head.  When we pitch the camera forward, we are also going to pitch the nodal point forward away from the axis of rotation such that for each pitch angle, we would need to readjust the nodal slider to re-position the nodal point to the axis of rotation of the panning clamp.  This is not a problem for single row, but to do this for multi row would be a chore.  This is why the $1000 rigs exist: to center the nodal point such that all camera movements are centered around that single reference point!   To do multi row with this cheapy setup would require re positioning the nodal point for each row.

[Ellis Vener]

"But if I PITCH the panning clamp downward so the camera is no longer level the camera body will ROLL as the panning clamp rotates through the panning motion and the panning will be in an arc and not a straight line."

The arc will be a straight line but if what you mean is that it will not be level to the horizon you are correct.

[Ellis Vener]

Have you seen this?  https://www.youtube.com/watch?v=bKAW2P7ajWA

You might find the segment from  0:55 to 1:21 interesting

[dwswager]

Here is a makeshift setup utilizing the RSS basic pano setup for single row and adapted for multi-row by attaching a monopod head that allows the camera to be tilted up or down. Not that elegant, but it works.

Bill

Yeah, thanks.  That is the thought I had.  BTW, wonder if the Nodal slide should be in the panning clamp or in the monopod clamp?  New thing to investigate.

[dwswager]

"But if I PITCH the panning clamp downward so the camera is no longer level the camera body will ROLL as the panning clamp rotates through the panning motion and the panning will be in an arc and not a straight line."

The arc will be a straight line but if what you mean is that it will not be level to the horizon you are correct.

I guess what I'm trying to say, and doing a very poor job of it, is that the FIELD of VIEW won't be level throughout the shot sequence.  BTW, will check out the video.  Thanks!

[Bart_van_der_Wolf]

I guess what I'm trying to say, and doing a very poor job of it, is that the FIELD of VIEW won't be level throughout the shot sequence.

Indeed, you can think of it as a ribbon of images taken from the center of a hollow sphere with the world being projected from the center on the inside of that sphere. The camera stays level in its own coordinate system but the sphere seems to be tilted at an angle, just like a star filled celestial sky.

The software will then bring both coordinate systems into alignment, and the world horizon will be leveled and straight.

Cheers,
Bart

[dwswager]

Have you seen this?  https://www.youtube.com/watch?v=bKAW2P7ajWA

You might find the segment from  0:55 to 1:21 interesting

Still trying to wrap my head around this.  Notice the 1st pan they do is from the bottom panning base that is level.  The 2nd appears to be on the off level panning part on top of the head.  This is in direct conflict of everything I have read or seen.

This video from RRS for their $915 Multi-Row Pano Elements Package (1:05), the last step in the setup video is level the base at the tripod apex which is the place of horizontal rotation and assures the vertical rotation is in a plane perpendicular to the horizontal.



Guess I'm going to have to play and see for myself.

[dwswager]

Here is a makeshift setup utilizing the RSS basic pano setup for single row and adapted for multi-row by attaching a monopod head that allows the camera to be tilted up or down. Not that elegant, but it works.

Bill

Do you account for the shift in nodal point when you tilt up or down of horizontal or is it just "good enough"?  As in engineer, I'm already thinking of the geometry to calculate the shift. 

I'm gonna build this setup to play with.  Still debating on the RRS version or using the Sunwayfoto DH-03i panning clamp (with backward degree scale), 240mm Rail, and bidirectional clamp and adding the Sunwayfoto monopod head DT-01 that has built in dovetail on the base.

[dwswager]

Here is a makeshift setup utilizing the RSS basic pano setup for single row and adapted for multi-row by attaching a monopod head that allows the camera to be tilted up or down. Not that elegant, but it works.

Bill

Bill,

Don't know if you have tried swapping the nodal slide and the monopod tilt head.  If you mount the tilt head centered into the panning clamp so the axis of rotation goes through the center of the clamp, it now just requires knowing that single nodal point on the slider.  At 0° tilt you just set as normal or skip the tilt head all together.  At any other tilt angle all you need to do is position that same known nodal point for the camera/lens combo at the center axis of the tilt head.  About as much precision as any other method in the field. 

[allegretto]

Two rather experienced Pano shooters and the OP still doesn't believe them… try this;

Of course when the plate is not level the translation across the field relative to a perfect horizontal line will appear curved. Because it is. But it is still the pano of that latitude (think of it that way). It is the flat projection that makes it appear curved. It is straight in 3-D space

Consider that the actual event is in Raymond Geometry (sp?) that is 3 dimensional, not two. So long as the sensor is perpendicular to the subject and you are centered through the nodal point you are "on the line" and will get a proper pan of that meridian. Similarly you will have to adjust focus unless the subject is similarly curved to the path of the translation.