Advanced Rotators for Nodal Ninjas

[Bill Bailey]

I wanted to make a quick post here as well for a recent update on the availability on the new advanced rotators.
Unfortunately we have had some manufacturing delays. Rest assured however we will not comprise quality of build for speed of production, Nick is very precision orientated. So a bit more time is needed to build up the next batch rotators.
While there is a very good chance we may see some inventory trickle in within the the 2-3 weeks we are posting an estimated ship date of Dec 1st. Not sure yet if we will be able to fill existing back orders - if not we will notify those that may be effected.

Once we are back into full swing with the rotators and Nodal Ninja configurations using these rotators we'll send out newsletters announcing the availability.
If you have not yet signed up for our newsletter please click here https://store.nodalninja.com/MailingList_subscribe.asp
Thank you for your patience.
Bill

[J Holmes]

Dear Nick,

At Bill Bailey's suggestion, I am posing my question for you about the R-D12 here.

I have been hoping to see a fast and compact click-stop device with fine enough angle settings for stitching for some time, which does not require changing disks. Thank you. Yours is the first.

My use for a rotator is to allow the base of a tripod head to rotate around the vertical axis, regardless of whether the camera is pointed up or down, so as to make repeatable camera positioning fast and easy. Largely this is so I can do focus blending of two or more exposures for each camera direction, without having to refocus for each frame (e.g. near/far, move camera, far/near, move camera, etc.), rather by doing it this way: shoot near, move camera, shoot near, move camera, shoot near, move camera, refocus to far, reverse direction, shoot far, move camera, shoot far, etc. This pattern has major ramifications for lens design. For example, Auto Focus lenses sometimes have very poor repeatability for manual focussing due to very short throws.

Only your R-D12 has a fine enough choice of rotation angles to work properly for the full line of lenses involved.

But unlike the Manfrotto 300N rotator, yours does not have a bolt at the top to attach the tripod head, rather you've designed it for attaching your nodal slides, etc., with the intent that people using pano techniques will always be shooting toward the horizon.

I would submit that pano techniques are good for a lot more than just panos, such as more traditional image making, but with variable aspect ratio not by cropping, and for increasing pixel count a lot, not to mention having the option of cylindrical renderings with wide scenes. And in any case when the horizon will be shown in the picture, rotating around the vertical axis is desirable. Therefore I need to use an R-D12 below the tripod head and thought you would like to either make a version with a 3/8" bolt on top, similar to that of the Manfrotto, or make an adapter plate for the existing R-D12 design.

I would prefer to not need to modify the R-D12 myself, as it would be very time consuming for me -- and then I'd have the only one :-) I'd rather your other customers could have one too.

Regards,

Joseph Holmes

[nick fan]

My use for a rotator is to allow the base of a tripod head to rotate around the vertical axis, regardless of whether the camera is pointed up or down, so as to make repeatable camera positioning fast and easy. Largely this is so I can do focus blending of two or more exposures for each camera direction, without having to refocus for each frame (e.g. near/far, move camera, far/near, move camera, etc.), rather by doing it this way: shoot near, move camera, shoot near, move camera, shoot near, move camera, refocus to far, reverse direction, shoot far, move camera, shoot far, etc. This pattern has major ramifications for lens design. For example, Auto Focus lenses sometimes have very poor repeatability for manual focussing due to very short throws.

You will still need stitching before focus/exposure blending. There is no way we can achieve per pixel precision for long focal length.

But unlike the Manfrotto 300N rotator, yours does not have a bolt at the top to attach the tripod head, rather you've designed it for attaching your nodal slides, etc., with the intent that people using pano techniques will always be shooting toward the horizon.

I can make an M6-3/8" adapter for your purpose.

I would submit that pano techniques are good for a lot more than just panos, such as more traditional image making, but with variable aspect ratio not by cropping, and for increasing pixel count a lot, not to mention having the option of cylindrical renderings with wide scenes. And in any case when the horizon will be shown in the picture, rotating around the vertical axis is desirable.

what you described is just a process of making panos or mosaics in a case the scene is far away that no-parallax-point is not impoatant. NN5/5L is designed for making mosaics as well.


BTW, what do you like about my R-D12? You can use a Manfrotto 300N out of box.


Nick

[J Holmes]

Thank you Nick,

There is a way to make pixel-for-pixel alignment after the fact, automatically, for focus blending, which will work out very well when the alignment is close but not perfect -- the Align Selected Layers command in Photoshop CS 3 or CS 4. The aligned layers could then be focus blended in Helicon Focus or in Photoshop CS 4, if Adobe improves the operation of their new focus blending function, which currently works very poorly in my initial experience, and in that of a friend who tried it earlier. The result is masks where many regions are mis-selected, having chosen blurry information over sharp, in a spotty pattern.

But I appreciate your suggestion of stitching the near frames first, then stitching the far, then focus blending the two. I have not explored this very much, but I am sure that it can work. Indeed Helicon Focus does require that the image layers be aligned well to begin with, and that they be the exact same pixel dimensions. So I could stitch the two, adjust their pixel dimensions if necessary by quickly adding blank margins, align them in Photoshop with Align Selected Layers, crop to taste, then blend them in Helicon Focus. So the rotator would still help by greatly minimizing the differences between the near set and the far set, even if it's not pixel-for-pixel accurate.

There is another other good thing about using a rotator, which is that one could, camera permitting, leave the mirror locked up between shots, thus shortening the time between exposures since we don't have to wait as long for the camera to stop vibrating (without having to rotate blind, using the degree scale on the rotating base). The Mamiya AFD II/III won't do this, but some cameras have and more may in the future.

I assume that an adapter will require a spacer plate, similar in form to the one atop the Manfrotto (about 1/4" thick). Thanks very much for offering to make one. Hopefully you could make it one of your standard parts.

The thing I especially like about the R-D12 is that you have 7.5¬?, 10¬?, 12¬?, 15¬?, 18¬?, 20¬?, etc. These finer steps means that there is a setting which allows me to get the amount of overlap that I want with essentially every focal length I will use. Most rotators are intended for combining relatively short focal length lens images into an extremely wide angle of view captured. Imagine taking a picture that is anywhere between 25 and 90 degrees wide in total, but subdividing that into five to seven overlapping frames, and you can see that Manfrotto's options of 5¬?, 10¬?, 15¬?, 20¬?, 24¬?... are too coarse to give optimal overlap with the full range of longer lenses. Rather some focal lengths would work right, and some not. Ideally I would want your head to also have a 6¬? setting, but it's eminently workable the way it is now. The Manfrotto steps are too coarse at the long lens end of the scale. 5¬? to 10¬? being a 100% change in angle is not good... 10 to 15 being a 50% change isn't good either. Too coarse! 20% changes are great!

I will normally work with lenses that (in 35 mm equivalent) range from 106 mm down to 32 mm. I might also go up at high as 148 mm. That would be where I'd want the 6¬? setting, or perhaps 5¬?. Or maybe 6, 8, 10, 12, etc. I assume you don't have room for more detents with the current layout of the R-D12. I generally prefer a generous 50% overlap. When using Adobe's Photomerge, this gives the software more options as to where to draw the boundary between frames, which can help a lot to solve problems arising from swaying trees or moving waves. Often I would be happy enough with 30% overlap, but it's easy to make more frames in a single row, so I'd rather be sure, if there is time (the light is not failing fast).

When I'm using this general technique to use stitching to make images of traditionally narrow angles of view, I am not necessarily working without a foreground, so parallax is still an important issue in most cases and I use a nodal slide carefully. My optics are more like those of a 4x5 than those of a 35mm camera. Longer lenses with very little depth of field and more need for tilt or focus blending. But very high detail in the end.

Relatively quickly, more and more photographers are learning about the benefits of using stitching, and more will want to use it for improving quality, as opposed to making traditional panoramas and QTVR images, etc.

Thanks again!


Joseph Holmes

[nick fan]

Hi Joseph,

There is a way to make pixel-for-pixel alignment after the fact, automatically, for focus blending, which will work out very well when the alignment is close but not perfect -- the Align Selected Layers command in Photoshop CS 3 or CS 4. The aligned layers could then be focus blended in Helicon Focus or in Photoshop CS 4, if Adobe improves the operation of their new focus blending function, which currently works very poorly in my initial experience, and in that of a friend who tried it earlier. The result is masks where many regions are mis-selected, having chosen blurry information over sharp, in a spotty pattern.

But I appreciate your suggestion of stitching the near frames first, then stitching the far, then focus blending the two. I have not explored this very much, but I am sure that it can work. Indeed Helicon Focus does require that the image layers be aligned well to begin with, and that they be the exact same pixel dimensions. So I could stitch the two, adjust their pixel dimensions if necessary by quickly adding blank margins, align them in Photoshop with Align Selected Layers, crop to taste, then blend them in Helicon Focus. So the rotator would still help by greatly minimizing the differences between the near set and the far set, even if it's not pixel-for-pixel accurate.

Have you heard of Max Lyons, a pioneer and expert in high resolution mosaics. He has the PTAssembler which will stitch the panos.
http://www.tawbaware.com/ptasmblr.htm
Lately, he has release a focus and/or exposure blender (Tufuse) much better than programs you mentioned. Tufuse is a command line program which can be invoked by PTAssembler.
http://www.tawbaware.com/tufuse.htm
Or you can have a Tufuse with GUI, the TufusePro
http://www.tawbaware.com/tufusepro.htm
There is also a forum for the programs and mosaic making in general.
http://www.tawbaware.com/forum2/index.php

I assume that an adapter will require a spacer plate, similar in form to the one atop the Manfrotto (about 1/4" thick). Thanks very much for offering to make one. Hopefully you could make it one of your standard parts.

if you mount a ball head on top of the rotator, you don't need the top rotating plane in 300N for initial angle adjustment. I will try to make a small screw with M6 at one end and 3/8" at the other. I will let you know if it works.

The thing I especially like about the R-D12 is that you have 7.5¬?, 10¬?, 12¬?, 15¬?, 18¬?, 20¬?, etc. These finer steps means that there is a setting which allows me to get the amount of overlap that I want with essentially every focal length I will use. Most rotators are intended for combining relatively short focal length lens images into an extremely wide angle of view captured. Imagine taking a picture that is anywhere between 25 and 90 degrees wide in total, but subdividing that into five to seven overlapping frames, and you can see that Manfrotto's options of 5¬?, 10¬?, 15¬?, 20¬?, 24¬?... are too coarse to give optimal overlap with the full range of longer lenses. Rather some focal lengths would work right, and some not. Ideally I would want your head to also have a 6¬? setting, but it's eminently workable the way it is now. The Manfrotto steps are too coarse at the long lens end of the scale. 5¬? to 10¬? being a 100% change in angle is not good... 10 to 15 being a 50% change isn't good either. Too coarse! 20% changes are great!

Oh, that's what I was thinking.

When I'm using this general technique to use stitching to make images of traditionally narrow angles of view, I am not necessarily working without a foreground, so parallax is still an important issue in most cases and I use a nodal slide carefully. My optics are more like those of a 4x5 than those of a 35mm camera. Longer lenses with very little depth of field and more need for tilt or focus blending. But very high detail in the end.

To completely avoid parallax, you need to rotate about NPP in horizontal and vertical axis of rotation. You will need a pano head stronger than NN5 for your gear. We will have products suited for your application in the future.


Nick

[J Holmes]

Hi Nick,

Thanks again. I have indeed just two days ago been looking over Max Lyons pages and Tufuse, and it looks promising, but I'm using a Mac and not yet a Mac that can boot into Windows. I'm glad to hear that you find it to work better than Helicon Focus, which often works quite well now. At least the Mac version of Helicon Focus is about to catch up with the Windows version, making retouching composite images much easier. It will allow direct cloning from any of the source images into the combined image with an easy interface. TuFuse Pro sounds really good.

I do have a multi-row pano setup already, which I've used quite a bit with my Canon 5D, but I've yet to use the multi-row capability with my Mamiya. I'm hoping that if I run into the need to stitch a vertical column from three or more horizontal P45+ frames, that using the single row setup will work out OK despite up to an inch or so of parallax error, on account of Photomerge's almost magical ability to find the safe seams between poorly aligned images (CS3 and later only). Previous testing suggests this might even work when as close as five feet or so.

The plate atop the Manfrotto isn't a rotator, I believe, rather just a spacer disk to accommodate the hex nut in the middle of the bolt. See the zoomed-in picture here:

http://www.manfrotto.com/webdav/site...5695d5cd_L.jpg

An adapter bolt absent the hex nut in its middle would naturally tend to come off in the bottom of the tripod head when unscrewing the tripod head from the top of the rotator. So the way Manfrotto did it seems to be the natural solution, though I can't see the details, like how they connected the 1/4" thick disc to their rotator.

So should I go ahead and order the R-D12 now, or might you decide that it would be best to make an alternate version which is intended for use between a tripod and tripod head? (as opposed to an adapter)

Thanks for your help.


Joe

[nick fan]

So should I go ahead and order the R-D12 now, or might you decide that it would be best to make an alternate version which is intended for use between a tripod and tripod head? (as opposed to an adapter)

Wait until I test the adapter.


Nick

[Castillonis]

DOF / Focus blending on Mac
You can probably run tfuse.exe on a OS X Leopard using Apple's Boot camp to run windows XP or Vista.
http://www.apple.com/macosx/features/bootcamp.html
You can run it using vmware if you do or do not have the Leopard release.
http://www.vmware.com/products/fusion/

Gui Wrappers for tufuse.exe
There are two gui wrappers for tufuse.exe. Max originally wanted someone else to write a gui and someone ended up writing tufusion.exe at about the same time. You should go grab it on Max Lyons forum while it still is available. Hint you need to grab another library as well as tufuse.exe for it to work. Dont' forget to download the library that it mentions.
http://www.tawbaware.com/forum2/viewtopic.php?t=4966

[waters]

Nick:
I received my D8 rotator a couple of days ago and wanted to congratulate you on a quantum leap! Though of course heavier than the MkII, there is no comparison. No tensioning knobs to loosen, just a beautifully functional piece of equipment.

Thanks,
Tim Eastman

[han]

Hello Nick,

I bought the "Nodal Ninja 5 Panorama kop + RD12 Rotator" last month,
and now I read that there is a RD16 Rotator.

I want to shoot highres mosaics with a long lens, so I really could use the extra clicks.

2 questions

-Will the RD16 Rotator be sold seperately?
-Will there be a "simple" rotator only for highres mosaics in the future?
I only need the 20 degrees and smaller clicks.


thanks,

Han
Amsterdam

[nick fan]

Hello Nick,

I bought the "Nodal Ninja 5 Panorama kop + RD12 Rotator" last month,
and now I read that there is a RD16 Rotator.

I want to shoot highres mosaics with a long lens, so I really could use the extra clicks.

2 questions

-Will the RD16 Rotator be sold seperately?
-Will there be a "simple" rotator only for highres mosaics in the future?
I only need the 20 degrees and smaller clicks.


thanks,

Han
Amsterdam

Hi Han,

You can contact your reseller to see if you can upgrade the rotator. No many people need the finer intervals, R-D12 is still fine for many people. We will accept its return. If resellers refuse to do so, Bill and I will be happy to help you out.
yes, we sell rotators separately. Otherwise our old customers will be very angry. :-)
I have no plan to release another rotator similar to R-D8 and R-D16. R-D16 will satisfy all your need. There is not much cost/weight saving to abandon the coarser interval options. That said I may design a new rotator when I discover functions not satisfied by the R-D8/16. :-) I hope you are not asking me to stop developing products better satisfying needs of customers.
BTW, what specific needs do you have for the rotator? What interval options you plan to use?


Nick

[DemonDuck]

Another wacky idea from DemonDuck -- in addition to the cross hairs on the top of the rotator so one can more easily do the preliminary alignment to find the nodal point. I would also like click stops on the tilt rotator. This is why -- if you are shooting more than one row it is better to shoot all the images for one vertical slice together. If you shoot the bottom row all the way around then the top (or middle) row and you don't have camera that stores images fast, then by the time you shoot the next row, the light can change enough to make your pano difficult to stitch and/or just not good.

If you can shoot the vertical strip all together, the light is more constant for that strip. Does that make sense?

The light can change very rapidly during sunset and sunrise. Or if there are clouds on a windy day. Or if people are in more than one row. I think having click stops on the tilt rotator would make things easier. Without click stops on the tilt rotator, trying to shoot vertical strips is not practical since you really can't expect good alignment horizontally.

DD

[nick fan]

Another wacky idea from DemonDuck -- in addition to the cross hairs on the top of the rotator so one can more easily do the preliminary alignment to find the nodal point. I would also like click stops on the tilt rotator. This is why -- if you are shooting more than one row it is better to shoot all the images for one vertical slice together. If you shoot the bottom row all the way around then the top (or middle) row and you don't have camera that stores images fast, then by the time you shoot the next row, the light can change enough to make your pano difficult to stitch and/or just not good.

If you can shoot the vertical strip all together, the light is more constant for that strip. Does that make sense?

The light can change very rapidly during sunset and sunrise. Or if there are clouds on a windy day. Or if people are in more than one row. I think having click stops on the tilt rotator would make things easier. Without click stops on the tilt rotator, trying to shoot vertical strips is not practical since you really can't expect good alignment horizontally.

DD

Hi DD,

Thanks for sharing. Actually, with some practice you can get the right feeling of clicks in the tilt rotator in NN5. Probably I can make some tweaks in the tension knob to better achieve this function.


Nick

[han]

Hi Han,

You can contact your reseller to see if you can upgrade the rotator. No many people need the finer intervals, R-D12 is still fine for many people. We will accept its return. If resellers refuse to do so, Bill and I will be happy to help you out.
yes, we sell rotators separately. Otherwise our old customers will be very angry. :-)
I have no plan to release another rotator similar to R-D8 and R-D16. R-D16 will satisfy all your need. There is not much cost/weight saving to abandon the coarser interval options. That said I may design a new rotator when I discover functions not satisfied by the R-D8/16. :-) I hope you are not asking me to stop developing products better satisfying needs of customers.
BTW, what specific needs do you have for the rotator? What interval options you plan to use?


Nick

Hello Nick,

I read the RD16 is suitable for lenses ranging to 140mm (Full frame)

I want to make mosaics with a 100 mm lens on a canon 40D so I need 3 deg intervals(aprox). So as you
are not planning to produce a rotator with such small intervals the only option I have is buying a fullframe camera. :-(

One further question I have is this:

Now I own the NN5 + RD12 Rotator. and I am thinking of buying a NN3 + RD 16
rotator. Is it possible to mount the RD-16 under the NN5 and mount the
RD12 under the NN3 ??


greetings,

Han
Amsterdam

[nick fan]

One further question I have is this:

Now I own the NN5 + RD12 Rotator. and I am thinking of buying a NN3 + RD 16
rotator. Is it possible to mount the RD-16 under the NN5 and mount the
RD12 under the NN3 ??

yes, you can swap the R-D12 and R-D16 rotators.


Nick