Nikon APS sensor + Nikon 16mm Fisheye + NN3

Welcome to Nodal Ninja Forum

+ Reply to Thread
Results 1 to 8 of 8

Thread: Nikon APS sensor + Nikon 16mm Fisheye + NN3

Share/Bookmark
  1. Nikon APS sensor + Nikon 16mm Fisheye + NN3

    #1

    I have a question from a colleague who has recently purchased an NN3 to use with a Nikon outfit with an APS sensor and Nikon 16mm fisheye lens and who needs guidance on the number of shots to take to create a full spherical panorama.

    I have used Frank's calculator at:
    www.frankvanderpol.nl/fov_pan_calc.htm
    and made my own calculations, but do not have a 16mm fisheye to test them with, so am asking for anyone who has used this combination if they would like to pass on their experiences please?

    Unfortunately he is some 4000Km from me so I cannot just pop round and have a look.
  2. #2
    Users Country Flag
    Join Date: Aug 2008
    Location: Netherlands
    Posts: 1,734

    Hi Hugh,

    Took an indoor and outdoor shot with 8 and 10 pics around with D7000 and 16mm/f.28. Following Frank's calculation 8 shots around have an overlap of 21%, 10 shots around have an overlap of 37%.
    You can expect, because 21% is on the low side, I had more difficulties in stitching 8 shots indoors. I had to add CP's manually in several pairs to get a stitch. 10 shots around stitched much better. I believe I had to add Cp's in one pair, because I missed to take the 10th shot!! End of day after a long day's work. Shit happens.

    The outdoor panos stitched at once, 8 shots as well as 10 shots without adding any CP manually. Of course you have more information outdoor.

    Conclusion:

    Outdoor, in case of lot of information, 8 shots. In case of less information 10 shots.

    Indoor, like real estate, I would go for 10 shots around to be safe and to have less work to stitch.

    You have to remove CA in the first step before doing anything else. Without removing it, you get nice pink and green lines. I used DxO to remove the CA, Vignetting, but I did not change distortion. PTGui does a much better job. I will upload the pics and the projects files to ge.tt, will take some time. I'll try to do it this night but may be tomorrow evening.

    So long my conclusions,
    regards,
    Heinz
    Last edited by hindenhaag; 08-13-2013 at 02:19 PM.
  3. #3
    Users Country Flag
    Join Date: Aug 2008
    Location: Netherlands
    Posts: 1,734

    Hi Hugh,

    Here's the link to the pics: http://ge.tt/39hvGvo/v/0

    Heinz
  4. #4

    Hello Heinz,

    Than you for your quality response with useful information.
    I had come up with 8 shots round as a reasonable solution, but that 10 or 12 would give a better solution, so your information has reassured me that I was in the right field.
    I think that it is best for my colleague to go for 10 as I feel he needs a "routine" until he becomes more experienced and my NN3 came with 8 and 10 detent rings, but not 12.

    Thanks, Hugh.
  5. #5
    Users Country Flag
    Join Date: Aug 2008
    Location: Netherlands
    Posts: 1,734

    Hello Hugh, pleasure to help you.

    I've added all the pics so he can try himself stitching them and see where the problems are by comparing a new stitch and the PTGui project's file results.
    Mine came with 8 - 10 ring as well.

    Kind regards,
    Heinz
  6. #6

    Heinz, I'm curious to know why your interior shots included in the zip file show such huge variations in brightness when the camera was in manual mode and the exif data for all the images is identical. Something in the RAW conversion?

    John
  7. #7
    Users Country Flag
    Join Date: Aug 2008
    Location: Netherlands
    Posts: 1,734

    John, I'll check it. Thx.

    Heinz
  8. #8

    Hello Heinz,

    Your data is a great help and much appreciated as I don't have a 16mm fisheye to experiment with.

    I am recommending 10 shots round to my colleague to give more scope for Control Point generation.
    (I like lots of overlap!).

    Quote Originally Posted by John Houghton View Post
    Heinz, I'm curious to know why your interior shots included in the zip file show such huge variations in brightness when the camera was in manual mode and the exif data for all the images is identical. Something in the RAW conversion?
    John
    Maybe something to do with DxO?

    Thanks again Heinz,
    Best regards, Hugh.
+ Reply to Thread

Posting Permissions

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