Switch to English Language Passer en langue française Omschakelen naar Nederlandse Taal Wechseln Sie zu deutschen Sprache Passa alla lingua italiana
Members: 68,677   Posts: 1,481,972   Online: 927

Making Centre Filter/s for LF wide angle lens/es

  1. OldBikerPete
    Hi guys.
    I must have too much time on my hands.
    Over the last few days I have written a program (in C on Linux) which creates a .BMP file which emulates an ND centre filter. It creates an image which is white at the corners and graduating to an appropriate density towards the centre by the 'cos cubed' rule.
    I have used the one I generated for my Nikkor-W 135mm lens (ND1.5) as a mask in Photoshop to correct a scan of a negative exposed using that lens. The difference is noticeable even with such a mild wide-angle as that. The factors for 90mm and 65mm wide-angles are ND2.3 and ND3.8, respectively.

    Anyone who wishes a copy of this program are welcome to it. PM me your eMail address and I'll send it to you.
    If there's enough interest I might even port it to Microsoft.

    At the moment I'm mulling over the physics involved in creating a centre filter out of negative material to be semi-permanantly fixed over the REAR element of wide-angle lenses.

    EDIT: I've nutted out how it would have to be done. Basically by photographing a mask generated by a program similar to this one, printed on A4 or larger paper. I'll have to get hold of some B&W negative material and play around with exposures and development times to get the necessary densities accurate.
  2. SMBooth
    SMBooth
    How clever are you..... I see a application for it on pinhole images as there suffer a great deal from fall off. A guy created a calculator to show light fall off in pinhole (http://www.durian.se/pinhole/evdrop.html) maybe it something in it you could use. Does your program have variables or is it fixed for each lens?
  3. OldBikerPete
    OldBikerPete
    The actual mathematics is only a tiny part of the program - a few lines of code. The rest is parameter input and checking and generating of the protocol of the output .BMP file.
    I'll look at that web site and see whether I can write a special pinhole version.
    To answer your question about inputs, the program asks for:

    lens focal length in mm
    image width in mm
    image height in mm
    mask width in pixels
    mask height in pixels
    name of output file.

    EDIT: It occurred to me that the apparent optical centre of the lens on the film side just might not be at the focal length distance from the film when the lens is focussed at infinity. So I have modified the program to accept two more inputs 'CoverageAngle' and 'ImageCircle' and to calculate the apparent optical centre from those. If either of these two new parameters are zero, the program ignores both and just uses the focal length.
    The parameters are now:

    lens focal length in mm
    coverage angle in degrees
    image circle diameter in mm
    image width in mm
    image height in mm
    mask width in pixels
    mask height in pixels
    name of output file.

    The difference for the Nikkor-SW 65mm f/4 lens is 62.5mm vs 65mm
  4. OldBikerPete
    OldBikerPete
    I've had a go at the pinhole version. It doesn't work. When implemented in double-precision C math, the math published at the above web-site has a discontinuity at about 22mm off-axis from the pinhole. I didn't worry about generating a mask at first, I just tried to duplicate their effects. What I got was a circle 44mm diameter going from white at the centre to black at the edge then immediately back to white grading to black at the corners of the 9x6 image. I used their 0.242mm, 55mm FL pinhole parameters.
    I'll keep picking away at it to see if I can fix the problem.

    It occurs to me that 1> vignetting in pinhole cameras is greater than in lens cameras - greater than can practicably compensated for in Photoshop 2>a slight reduction in resolution is unlikely to be a problem.

    So, why doesn't someone make a pinhole exposure in the target camera of a blank surface and develop the negative. That negative, with appropriate exposure and development to produce the correct densities becomes the right mask to sandwich with the film, correcting the vignetting at the source.

    Anyone want to give it a try?
  5. SMBooth
    SMBooth
    Pete, can you send the program to me so I can have a play?
    s.booth AT optusnet.com.au
Results 1 to 5 of 5


 

APUG PARTNERS EQUALLY FUNDING OUR COMMUNITY:



Contact Us  |  Support Us!  |  Advertise  |  Site Terms  |  Archive  —   Search  |  Mobile Device Access  |  RSS  |  Facebook  |  Linkedin