Side to side adjustment needed on R10 (and R1)
Results 1 to 14 of 14

Thread: Side to side adjustment needed on R10 (and R1)

Share/Bookmark
  1. Side to side adjustment needed on R10 (and R1)

    #1
    Users Country Flag
    Join Date: Sep 2009
    Location: Los Angeles, CA
    Posts: 359

    I have been using an R1 pano head at 5 degrees tilt for years now. Looking into my spare parts box I found an R10 fixed at 5 degrees. I got a screaming deal a few years ago off of eBay, so I could not turn it down. Today I decided to see what it would take to put the R10 into service.
    .
    Going through the calibration exercise I found that the fore/aft setting was exactly the same for both the R1 and R10 pano heads. Good job thinking ahead Nick. (Nikon D800, Tokina 12-17, 5 degrees up tilt = 21mm)
    .
    Problem is that my 6" calibration disk I put between the rotator and tripod head was coming out saw toothed in the stitched pano.
    .
    Mind you pictures taken with the R10 slightly off center stitch ok, but when you are using a monopod you need to eliminate as much error in your set up as you can since you will be inducing NPP errors while taking pictures.
    .
    Now that I have been doing this for a few years I knew what adjustment needed to be made in order to have the disk come out nice and round.
    .
    I went out to my machine shop for yet another round of "modify the pano head". I turned the precision drilled holes into precision slots. I also elongated the counter sinks.
    .
    Now I have about 2mm side to side adjustment. I was able to dial in the pano head to the point where my calibration disk is nice and round, not pointy or lumpy.
    .
    Nick, please modify your R10 clamp to include slots instead of holes. No matter how well you make your pano heads (no complaints here) there is still room to improve and get the camera precisely at the NPP.
    Last edited by DennisS; 03-17-2014 at 06:24 PM.
  2. #2
    Users Country Flag
    Join Date: Sep 2009
    Location: Los Angeles, CA
    Posts: 359

    Finally got around to taking pictures of my modified R10.
    .
    Here are my mods to the R10. They allow for greater repeatability (when you take apart and reassemble) and more accurate calibration.
    .
    The base gets two pins. This allows the vertical arm to be turned around for either upward or downward tilt. The vertical arm will always go back into the same position it was prior to disassembly.
    .

    .
    Next the vertical arm receives a "keyway" that allows for very precise side to side adjustment. The slot is for the spring under the plunge pin.
    .

    .
    The base of the clamp gets a slot to mate with the vertical arm. The hole in the center is for the spring loaded plunge pin (which is left out of the original R10). The bolt holes are slotted to allow for side to side adjustment.
    .

    .
    Top view of the clamp.
    .

    .
    Top view of the assembly.
    .

    .
    Nobody needs any of these mods in order to create great panos with a stock unit. For my use, these mods allow for more accuracy and greater usability of an already great pano head.
  3. #3
    Users Country Flag
    Join Date: Sep 2009
    Location: Los Angeles, CA
    Posts: 359

    Distance scale engraved on the bottom.
    .
  4. #4
    Users Country Flag
    Join Date: Apr 2008
    Location: Hong Kong
    Posts: 2,462

    nice mod.



    Fanotec
    We listen. We try harder.
  5. #5

    Any chance of something like this being offered officially Nick? I too find myself having to move the R1/R10 sideways! I've mounted mine on a short rail and placed a clamp on the rotator! It works but is a little inelegant!
  6. #6
    Users Country Flag
    Join Date: Sep 2009
    Location: Los Angeles, CA
    Posts: 359

    Nick, are you listening? People have contacted me off list about this mod. Remember the nadir adapter and NN3 ring clamp adapter?
  7. #7

    I second the motion. My panos using the R1 also are coming out in a way that indicates a need for side-to-side adjustment (saw-tooth patterns on the tripod mount at the center of the nadir). I don't have Dennis' machine shop capability. Is there any other way to make these adjustments? I am having trouble understanding what Supertrogg is doing with a rail and clamp. I'd take inelegant if I could eliminate these nagging stitching errors.
  8. #8
    Users Country Flag
    Join Date: Sep 2009
    Location: Los Angeles, CA
    Posts: 359

    It sometimes takes a while for good ideas to go into production. My prototype Nadir Adapter was rejected by all the NN dealers. 2 years later Nick produced it.
    .
    I cut windows in my lens ring clamps. At that time Nick told me it would not be cost effective to offer the service to cut windows. We now have windows.
    .
    Side to side adjustment is needed since the ring clamps do not hold the lens centered. I have 3 or 4 different model lens clamps and each one is off center one way or another, varying by different amounts. I have to say that stitching is so much easer now that I can dial in the left/right setting.
    .
    The issue is supply/demand. The pano community has not voiced enough demand to nudge Nick to include this modification in the R1/R10 pano head. If enough people ask for it, Nick will do it. After all, their slogan is "We listen. We try harder" Get all your pano buddies to email Nick or their local dealer asking for this mod and if enough respond, you will see it come to life. It may take a while but the effort would be worth it.
  9. #9
    Users Country Flag
    Join Date: Apr 2008
    Location: Hong Kong
    Posts: 2,462

    Quote Originally Posted by stuartf287 View Post
    I second the motion. My panos using the R1 also are coming out in a way that indicates a need for side-to-side adjustment (saw-tooth patterns on the tripod mount at the center of the nadir). I don't have Dennis' machine shop capability. Is there any other way to make these adjustments? I am having trouble understanding what Supertrogg is doing with a rail and clamp. I'd take inelegant if I could eliminate these nagging stitching errors.
    You can shim a piece of paper underneath the lens ring insert to correct for off axis misalignment. But the misalignment should be very small and any good stitcher can blend the error out. Also stitching error can be due to poor lens parameter optimization.
    Nick



    Fanotec
    We listen. We try harder.
  10. #10

    Thanks. I was wondering whether something like that would work. I'll try it soon and post the results.
  11. #11
    Users Country Flag
    Join Date: Sep 2009
    Location: Los Angeles, CA
    Posts: 359

    If shims are good enough for the R1, why provide side to side adjustment for all the other pano heads like the M1? Because there is no way to manufacture equipment 100% "hard coded". You need adjustments fore/aft and left/right.
    .
    (Start of soap box rant. Skip to then end if you like.)
    There is a reason I do NOT offer stitching advice when it comes to lens paramters are or how to optimize for shift, tilt or yaw. Since my Nodal Ninja equipment is so well calibrated I never have stitching issues so I never mess with anything other than control points and masking. I do not have a clue how to make other adjustments in the software.
    .
    I put a lens in the ring and mounted it horizontally. I then put a a dial indicator at the top. After rotating 180 degrees I found variances up to .050". that means I would have to put a .025" shim on one side to compensate. If something happened to the shim during my travels I would be out of luck. If the shift was towards the adjustable jaw on the clamp, you will never be able to shim out the error.
    .
    Why resort to shims when there is no reason not to get the pictures as perfectly centered around the NPP as possible right out of the camera. You are already set up to machine the required surfaces. Adding a couple dozen lines of G code to cut the slots and such is not a big deal. Putting a shim under the foot will tilt the camera off verticle axis and possibly cause the panorama to stitch with a tilted horizon. Yet another issue that could have been avoided if the camera was plumb and centered.
    (End of soap box rant)
    .
    I agree that stitching software can compensate for alignment errors, but why not get the pictures as perfect as possible right out of the camera in the first place? Professional equipment should not require shims.
  12. #12
    Users Country Flag
    Join Date: Apr 2008
    Location: Hong Kong
    Posts: 2,462

    Quote Originally Posted by DennisS View Post
    If shims are good enough for the R1, why provide side to side adjustment for all the other pano heads like the M1? Because there is no way to manufacture equipment 100% "hard coded". You need adjustments fore/aft and left/right.
    .
    (Start of soap box rant. Skip to then end if you like.)
    There is a reason I do NOT offer stitching advice when it comes to lens paramters are or how to optimize for shift, tilt or yaw. Since my Nodal Ninja equipment is so well calibrated I never have stitching issues so I never mess with anything other than control points and masking. I do not have a clue how to make other adjustments in the software.
    .
    I put a lens in the ring and mounted it horizontally. I then put a a dial indicator at the top. After rotating 180 degrees I found variances up to .050". that means I would have to put a .025" shim on one side to compensate. If something happened to the shim during my travels I would be out of luck. If the shift was towards the adjustable jaw on the clamp, you will never be able to shim out the error.
    .
    Why resort to shims when there is no reason not to get the pictures as perfectly centered around the NPP as possible right out of the camera. You are already set up to machine the required surfaces. Adding a couple dozen lines of G code to cut the slots and such is not a big deal. Putting a shim under the foot will tilt the camera off verticle axis and possibly cause the panorama to stitch with a tilted horizon. Yet another issue that could have been avoided if the camera was plumb and centered.
    (End of soap box rant)
    .
    I agree that stitching software can compensate for alignment errors, but why not get the pictures as perfect as possible right out of the camera in the first place? Professional equipment should not require shims.
    As we know, there is no single NPP for fisheye lenses. correcting this 0.025" just causes more trouble than benefits. I have consulted many experts in the field about which NPP to choose. Different people have different preferences. Some people will put more allowance at zenith and nadir. Some people put more emphasis on the horizon. You know typically we shot single row at 90 deg intervals for Sigma 8mm. This means the interval is 180 deg vertically and the NPP is very different from that of 90 deg. So any NPP is just a compromise. The consensus is a difference of 1mm is perfectly acceptable with modern stitcher.

    Nick
    Last edited by nick fan; 10-10-2014 at 02:09 AM.



    Fanotec
    We listen. We try harder.
  13. #13

    To see what effect a lower rail offset has on the stitching, I ran a series of tests with a Canon 8-15mm fisheye at 12mm on a 5DII (equivalent to an 8mm fisheye on an APS-C camera). 4 shots around on a NN5 in my kitchen, with nearest objects about 0.7m away. I shot panoramas at 1mm intervals on the lower rail setting: from -2mm to +8mm relative to the assumed optimum position (near enough). Good stitches (no visible glitches) were obtained with offsets of up to 2-3mm (av cp distances up to 1.0). Glitches began to show at 4mm and higher (av cp dist >1.2). PTGui's optimizer average control point distances were plotted for the various offset distances:



    I also shot a panorama from an identical position with the same lens in a ring mount on an R1 head. The setting appeared to be spot on, giving an average control point distance less than any of the values obtained in the 5DII tests.

    Of course, this data relates to just one lens in one situation, albeit a typical one, posted here FWIW.

    John
    Last edited by John Houghton; 10-10-2014 at 06:56 AM.
  14. #14
    Users Country Flag
    Join Date: Apr 2008
    Location: Hong Kong
    Posts: 2,462

    As usual, John's answer and demonstration are concise and conclusive. :-)

    Thx for the test.

    Nick



    Fanotec
    We listen. We try harder.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •