Luminous Landscape Forum
Equipment & Techniques => Cameras, Lenses and Shooting gear => Topic started by: Caviar on April 20, 2014, 03:40:11 am
-
I have read a lot about panning clamps vs leveling bases. And i understand the advantage of panning clamps as it is easier and more precise to set up for a horizontal pano. However, what equipment is required to make a pano with a tilt downward (or upward)? Lets say i'm on top of mount fuji and want to shoot a pano 30 degrees downward to the foot of the mountain all the way round.
If I browse the RRS site it seems you need the PG-02 Pro Omni pivot package. But that is quite some equipment. Wouldn't a monopod head of RRS be an easier solution on top of a panning clamp? I believe manfrotto also makes a tilt head. Perhaps together with a nodal slide if there are near objects. However, i never see it advertised as a solution for panos
Or to put it otherwise: what equipment qould you advise for a tilted pano with panning clamp? And why did or didn't you just go with a leveling base in this case, which is relatively cheap.
-
Hi,
I have used something like this:
(http://echophoto.dnsalias.net/ekr/images/PanoHead2.png)
Best regards
Erik
I have read a lot about panning clamps vs leveling bases. And i understand the advantage of panning clamps as it is easier and more precise to set up for a horizontal pano. However, what equipment is required to make a pano with a tilt downward (or upward)? Lets say i'm on top of mount fuji and want to shoot a pano 30 degrees downward to the foot of the mountain all the way round.
If I browse the RRS site it seems you need the PG-02 Pro Omni pivot package. But that is quite some equipment. Wouldn't a monopod head of RRS be an easier solution on top of a panning clamp? I believe manfrotto also makes a tilt head. Perhaps together with a nodal slide if there are near objects. However, i never see it advertised as a solution for panos
Or to put it otherwise: what equipment qould you advise for a tilted pano with panning clamp? And why did or didn't you just go with a leveling base in this case, which is relatively cheap.
-
I have read a lot about panning clamps vs leveling bases. And i understand the advantage of panning clamps as it is easier and more precise to set up for a horizontal pano. However, what equipment is required to make a pano with a tilt downward (or upward)?
Hi,
Depends on whether you want to make a single row pano, or a multi-row pano.
A single row Pano doesn't require much gear, just a Ballhead or similar with a panning clamp on the top, and a No-Parallax-Point bar to shift the camera back far enough to align the entry pupil of the lens above the center of the panning clamp. One must make sure that the optical axis is also aligned with the pivot point, which is accommodated by a well designed camera plate. An L-plate allows to quickly switch between landscape or portrait orientation while keeping things aligned and the center of gravity centered above the tripod. From the Really Right Stuff accessories that means a PCL-1 clamp + MPR-CL II rail (search for "Pano-Elem-Pro").
A multi-row pano requires much more additional equipment, and maybe you can get away without it. As long as you get your foreground stitched accurately, you may be able to blend the background elements without too much problems. But for a worry free setup, you'll need to invest a fair amount of money.
Lets say i'm on top of mount fuji and want to shoot a pano 30 degrees downward to the foot of the mountain all the way round.
It's not clear if you want a wide pano or a virtual 360 degree pano. The latter requires a lot more preparation, a multirow pano setup, and you need to shoot a separate Nadir shot to hide the tripod legs, and good software that allows to apply the corrections and masks.
If I browse the RRS site it seems you need the PG-02 Pro Omni pivot package. But that is quite some equipment.
If you want to shoot a difficult scenario, you'll need a lot of gear to pull it off without problems.
Wouldn't a monopod head of RRS be an easier solution on top of a panning clamp?
Easier, yes, but it will not work in all cases.
Or to put it otherwise: what equipment would you advise for a tilted pano with panning clamp? And why did or didn't you just go with a leveling base in this case, which is relatively cheap.
A leveling base is not going to solve the problem, and is not necessary because you can level more accurately in the stitching software. You do need a setup that allows to rotate the entrance pupil of the lens through a single point in 3D space, which is simple with a single row pano, and complex with multi-row panos.
Cheers,
Bart
-
Hi,
Depends on whether you want to make a single row pano, or a multi-row pano.
A single row Pano doesn't require much gear, just a Ballhead or similar with a panning clamp on the top, and a No-Parallax-Point bar to shift the camera back far enough to align the entry pupil of the lens above the center of the panning clamp. One must make sure that the optical axis is also aligned with the pivot point, which is accommodated by a well designed camera plate. An L-plate allows to quickly switch between landscape or portrait orientation while keeping things aligned and the center of gravity centered above the tripod. From the Really Right Stuff accessories that means a PCL-1 clamp + MPR-CL II rail (search for "Pano-Elem-Pro").
Cheers,
Bart
Hi Bart,
Thanks for the explanation.
Let's forget about the multi row stuff for the moment. With the instruction above it works for a horizontal pan. But for a tlted pan you need at least a tilt head on top of the panning clamp, right? Something like the picture above. I was just wondering why this is not often advertised (except now :-) ) as a solution.
And wouldn't a leveling base be just as efficient and even cheaper (although harder to set up, because you have to turn the ballhead straight forward and not pivot it a bit to the left or right when you position it for tilt)?
-
Hi,
I have used something like this:
(http://echophoto.dnsalias.net/ekr/images/PanoHead2.png)
Best regards
Erik
What tilt head is it? And is it Arca Swiss compatible?
-
That looks like an older model of the Manfrotto monopod tilt head. With the right clamp it is Arca Swiss compatible.
-
I have read a lot about panning clamps vs leveling bases. And i understand the advantage of panning clamps as it is easier and more precise to set up for a horizontal pano. However, what equipment is required to make a pano with a tilt downward (or upward)? Lets say i'm on top of mount fuji and want to shoot a pano 30 degrees downward to the foot of the mountain all the way round.
The advantage of a panning clamp that goes on top of a tripod head is that it puts a rotator joint directly under the camera, This way the head becomes the leveling device ( if level is what you want) but really "level" is just another angle. Whatever angle you set the head to, the rotation plane remains constant for the degree of pitch (fore/aft) and roll (left/right) you set the head to. So in your example you tilt the head and then rotate the clamp. if you just rotated the head at it's base, unless the head is perfectly level, the camera will move in an arc.
If I browse the RRS site it seems you need the PG-02 Pro Omni pivot package. But that is quite some equipment. Wouldn't a monopod head of RRS be an easier solution on top of a panning clamp? I believe manfrotto also makes a tilt head. Perhaps together with a nodal slide if there are near objects. However, i never see it advertised as a solution for panos
I believe that this is not generally mentioned as a standard way of working because of the potential of parallax errors when there are significant near/far spatial relationships in the subject. It is a matter of geometry: in multi-row panoramic work such an set up drastically moves the nodal point or lens entrance pupil when setting a different tilt angle rather than pivoting and rotating around the n.p or entrance pupil. If there are no significant near/far relationships it is likely not an issue that the software cannot automatically handle and sometimes those parallax errors can be resolved through editing layers of the composite image, and sometimes they just can't be resolved after the fact. I suspect that most manufacturers of multi-row panoramic gear think that for most multi-row panoramic composites most folk prefer to resolve the issue when shooting.
Or to put it otherwise: what equipment qould you advise for a tilted pano with panning clamp? And why did or didn't you just go with a leveling base in this case, which is relatively cheap.
I have both panning clamps and a leveling base. This is the leveling base I use: http://www.bhphotovideo.com/bnh/controller/home?O=&sku=554092&gclid=CL6slqi_8L0CFSsQ7Aod0G8AkA&Q=&is=REG&A=details , a Manfrotto 338 QTVR Leveling Base. Mostly I use it between the tripod head and the panning clamp, especially for architectural photography. Even though I own and use very good tripod heads, having the leveling base between the tripod head and panning clamp is terrific when you need to make very precise adjustments.
You also mentioned the PG-02 Pro Omni pivot package. that is the one I currently use. I also have the older RRS multi-row kit and also worked with the Nodal Ninja for awhile as well. The great thing about RRS panoramic gear is that it is kind of like an erector set and not only do you only have to use the pieces a particular panorama requires it can also be used for other types of photography. The next link describes how RRS camera bars and PG-02 gear were used for a macro shot of a garden spider in her web late last summer: http://blog.reallyrightstuff.com/bts-ellisvener/
-
I suspect Colorado David is correct about it being a manfrotto tilt-only head. Add an Arca-Swiss based dovetail foot to it and it should fit in almost any clamp based on the Arca-Swiss standard.
-
I use this leveling base between a heavy video fluid head and a home-built jib. It is very stout. http://www.bhphotovideo.com/c/product/554093-REG/Manfrotto_438_438_Compact_Levelling_Head.html/prm/alsVwDtl
-
Hi Bart,
Thanks for the explanation.
Let's forget about the multi row stuff for the moment. With the instruction above it works for a horizontal pan. But for a tlted pan you need at least a tilt head on top of the panning clamp, right? Something like the picture above. I was just wondering why this is not often advertised (except now :-) ) as a solution.
And wouldn't a leveling base be just as efficient and even cheaper (although harder to set up, because you have to turn the ballhead straight forward and not pivot it a bit to the left or right when you position it for tilt)?
The setup in the picture above would work fine for most cases. It uses the ball head as a levelling base and adds a rotating clamp and for multi-rows a tilt head on top. Its problem is that it does not allow the lens to pivot around its nodal point and is therefore prone to producing parallax. In reality this is not a problem unless there are strong vertical elements (and horizontal in the case of multi-row) in the fore or mid ground of an image and/or if you are using a longer lens. And yes, if you can tilt a ball head on one axis only and so you can dispense with the tilt head for multi-row stitches. Of course you would still be without parallax correction. If you have a levelling bubble on your tripod (or can level your tripod legs some other way) then you do not need the additional rotating clamp shown above. You can level your tripod and then use the base of the ball head for rotation.
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.
-
Hi Bart,
Thanks for the explanation.
Let's forget about the multi row stuff for the moment. With the instruction above it works for a horizontal pan. But for a tlted pan you need at least a tilt head on top of the panning clamp, right? Something like the picture above. I was just wondering why this is not often advertised (except now :-) ) as a solution.
Hi,
Because it doesn't keep the entrance pupil stationary when you rotate, it is not the solution that gives the required result.
A better solution is a ballhead with a rotating clamp on top. The head will allow to point anywhere you want, also level if needed, and the rotation will never cause parallax issues (assuming the use of a NPP bar). Always rotate at the top through the NPP if you want to avoid parallax issues. Using the correct setup allows to work very fast, which can also help avoid other issues.
Cheers,
Bart
-
Hi,
Because it doesn't keep the entrance pupil stationary when you rotate, it is not the solution that gives the required result.
A better solution is a ballhead with a rotating clamp on top. The head will allow to point anywhere you want, also level if needed, and the rotation will never cause parallax issues (assuming the use of a NPP bar). Always rotate at the top through the NPP if you want to avoid parallax issues. Using the correct setup allows to work very fast, which can also help avoid other issues.
Cheers,
Bart
Hi Bart,
Could you elaborate? From the picture with the manfrotto tilt head, the camera L-plate is just inserted right into the tilt head without correction for the no-parallax (or nodal) point. I see that and with a NPP bar that could have been solved I guess. But is DOES already use the panning clamp on top of the head as you suggested, right?
And there is one thing I don't get about your method: suppose I mount the camera straight on the panning clamp. Suppose I use a NPP bar to find the right NPP point. If I now tilt the ballhead downward and start panning the clamp, I would not get the desired result. My camera will initially point downward, but after 180 degrees of panning it will point upward into the sky.
Regards
Menno
-
The advantage of a panning clamp that goes on top of a tripod head is that it puts a rotator joint directly under the camera, This way the head becomes the leveling device ( if level is what you want) but really "level" is just another angle. Whatever angle you set the head to, the rotation plane remains constant for the degree of pitch (fore/aft) and roll (left/right) you set the head to. So in your example you tilt the head and then rotate the clamp. if you just rotated the head at it's base, unless the head is perfectly level, the camera will move in an arc.
I believe that this is not generally mentioned as a standard way of working because of the potential of parallax errors when there are significant near/far spatial relationships in the subject. It is a matter of geometry: in multi-row panoramic work such an set up drastically moves the nodal point or lens entrance pupil when setting a different tilt angle rather than pivoting and rotating around the n.p or entrance pupil. If there are no significant near/far relationships it is likely not an issue that the software cannot automatically handle and sometimes those parallax errors can be resolved through editing layers of the composite image, and sometimes they just can't be resolved after the fact. I suspect that most manufacturers of multi-row panoramic gear think that for most multi-row panoramic composites most folk prefer to resolve the issue when shooting.
Thanks Ellis,
I think I get it. If you would have figured out the nodal point for horizontal position with the "manfrotto tilt head" option, the nodal point will shift if you tilt the head, because the rotation point is a little below the lens. It's probably not too bad if you only tilt within 20 degrees.
So the PG-02 omni pivot is the only correct way to make sure the nodal point stays at the same location when tilting.
(http://www.reallyrightstuff.com/core/media/media.nl?id=4122&c=3705077&h=501f7f226a283998e655).
But that is a bit of an "all the way" solution and I could imagine using a cheap tilt head if you want to shoot a single row pano with a fixed tilt. Am I right?
I agree with the manufacturers by the way of rather having it fixed when shooting instead of afterwards. I already have a backlog of photos to be edited. So everything that can be done in the field is far more enjoyable than in the evening behind the PC screen. :)
Regards
Menno
-
That looks like an older model of the Manfrotto monopod tilt head. With the right clamp it is Arca Swiss compatible.
Yes, I think that is correct. I have the RSS single row pano setup and have attached a Manfrotto monopod tilt head to enable a single row pano with tilt up or down. Erik was using a similar setup, but without the nodal slider component. If the subject distance is large, as is often the case with landscapes, parallax is not a significant problem. I would be interested in Bart's comments if he is still following this thread.
Shown below are shots of the single row pano setup with and without the tilt head.
Regards,
Bill
-
So the PG-02 omni pivot is the only correct way to make sure the nodal point stays at the same location when tilting.
It isn't the only way but it is the most robust way that I have found offered by any manufacturer.
But that is a bit of an "all the way" solution and I could imagine using a cheap tilt head if you want to shoot a single row pano with a fixed tilt. Am I right?
If you are shooting a single row panoramic with the camera/lens axis axis tilted up or down I don't think you need to add a tilt head on top of a tripod head with a panning clamp, just tilt the head and then use the panning clamp to pan. I am willing to be persuaded that adding a tilt head to my usual set up will make for a better way of working however.
...everything that can be done in the field is far more enjoyable than in the evening behind the PC screen.
Could not have said it better myself.
Most of panoramic work is not landscape oriented, I primarily shoot industrial and other architectural interiors, like this: https://plus.google.com/u/0/109729682469414639106/posts
-
Yes, I think that is correct. I have the RSS single row pano setup and have attached a Manfrotto monopod tilt head to enable a single row pano with tilt up or down. Erik was using a similar setup, but without the nodal slider component.
That is interesting Bill. have you tried with the nodal rail on top of the tilt head instead of below it?
-
If the subject distance is large, as is often the case with landscapes, parallax is not a significant problem. I would be interested in Bart's comments if he is still following this thread.
Hi Bill,
That's exactly correct, one can only correct for parallax after the fact at 1 plane in the distance (either foreground, or background, not both at the same time). Since at a distance the parallax in pixels is smaller, due to magnification factor, we are often able to blend the distant features in a smart fashion, hard to immediately spot the errors by eye. Only by rotation through the NPP there will be no parallax, and thus no errors, at any distance. Therefore, adding the rotating clamp on top is mandatory.
What is entirely possible though, is to tilt the camera's optical axis to a non-level angle (e.g. down the mountain), and do a single row stitch (with NPP rotation) at that angle. In the Pano Stitcher software, one then uses the same Pitch angle for the entire row, and get a perfect parallax free stitch, where only the excess space need to be cropped.
Attached is an example of that, where I shot 3 images at an angle of some 30 degrees down, but rotated the camera through the NPP (rotating clamp on top of the ballhead, with MPR-CL II in that clamp to align the NPP), so no parallax issues in neither foreground nor background. The first attachment is just the rough stitch centered in view.
The second attachment is the same stitch but after moving the horizon up by adding a Pitch angle to all three images to remove keystoning in the background (a leveling plate would not have been able to achieve such an angle), which 'levels the image' although it will introduce perspective distortion in the floor as if the shot was taken level with a huge image circle and wide angle shift.
And the third attachment is the result after cropping, tonemapping, and a bit of cloning to fill in some details.
Cheers,
Bart
-
That is interesting Bill. have you tried with the nodal rail on top of the tilt head instead of below it?
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
-
Hi Bill,
That's exactly correct, one can only correct for parallax after the fact at 1 plane in the distance (either foreground, or background, not both at the same time). Since at a distance the parallax in pixels is smaller, due to magnification factor, we are often able to blend the distant features in a smart fashion, hard to immediately spot the errors by eye. Only by rotation through the NPP there will be no parallax, and thus no errors, at any distance. Therefore, adding the rotating clamp on top is mandatory.
Bart,
Thanks for taking the time to reply. With regard to your statement in bold, I ask on top of what? Is that what Ellis is recommending?
Bill
-
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.
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
-
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.
-
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.
-
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.
-
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!
-
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.
-
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
-
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.
-
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
-
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).
-
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, 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
-
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. ;-)
-
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!
-
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.
-
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
-
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
-
That is an interesting tripod. Who makes it?
Hi Bill,
It's a GorillaPod 'Focus' (http://joby.com/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